The ol' 5e problem: how to deal with Small bonuses

[Grod_The_Giant]

I really want to like 5e, I promise. It's simple but still feels like D&D. And I have no complaints about bounded accuracy, at least in concept. But... the combination of small bonuses and large dice... It's just not working for me. The bonuses even for a specialist are so overshadowed by the die roll it's just... It hurts. In play it honestly felt like the only thing that mattered was what came up on the die- there was no point to having proficiencies or ability modifiers of anything.

Now, admittedly this was at level 1, and the DM wasn't great, but still- I doubt another +1 or +2 would matter too much. So I ask you, my more experienced friends: how do you make your PCs feel competent, feel like their character creation choices matter?

So far, I've thought about:
Way fewer rolls is probably a given; the whole "only when failure is interesting
Letting people with proficiency take 10 most of the time would probably help a lot out of combat, at least.
Give anyone with proficiency Advantage as written, and change the normal Advantage mechanic to something else (+1d6?)
Replace the d20 altogether- 2d10, 3d6... 1d10 and reduce all DCs by 5...

I dunno. What do you guys do to make the game feel less stupidly swingy?

[Flashy]

I think the easiest fix for this is to start at 3rd level rather than first. Here's why I say this.


The bonuses even for a specialist are so overshadowed by the die roll it's just... It hurts.

Now, admittedly this was at level 1, and the DM wasn't great, but still- I doubt another +1 or +2 would matter too much.

In 5e, regardless of your class, you aren't playing a specialist at 1st level. You're playing an initiate, and one who hasn't gotten the majority of their basic features yet. Bards don't have Expertise or Jack of All Trades, Clerics can't Channel Divinity, Paladins can't smite, Wizards don't have a subclass. 1-3 are essentially tutorial levels. I suspect the feel you want will be much better represented by a game starting at 3rd level. Everyone has their archetype, spellcasters are getting six spell slots rather than two, and it becomes possible to specialize.

[Lines]

I really want to like 5e, I promise. It's simple but still feels like D&D. And I have no complaints about bounded accuracy, at least in concept. But... the combination of small bonuses and large dice... It's just not working for me. The bonuses even for a specialist are so overshadowed by the die roll it's just... It hurts. In play it honestly felt like the only thing that mattered was what came up on the die- there was no point to having proficiencies or ability modifiers of anything.

Now, admittedly this was at level 1, and the DM wasn't great, but still- I doubt another +1 or +2 would matter too much. So I ask you, my more experienced friends: how do you make your PCs feel competent, feel like their character creation choices matter?

So far, I've thought about:
Way fewer rolls is probably a given; the whole "only when failure is interesting
Letting people with proficiency take 10 most of the time would probably help a lot out of combat, at least.
Give anyone with proficiency Advantage as written, and change the normal Advantage mechanic to something else (+1d6?)
Replace the d20 altogether- 2d10, 3d6... 1d10 and reduce all DCs by 5...

I dunno. What do you guys do to make the game feel less stupidly swingy?

I tried the last one and my players hated it because they didn't get to roll d20s. Seems mathematically sound though.

[pwykersotz]

Being swingy is kind of the name of this game. In general you have three categories of action. Auto-fail, auto-succeed, or swingy.

I like the 2d10 and 3d6 methods to force a little more of an average, I used those for a player who rolled notoriously low (he has for the last 10 years, it was time for a change). I'd say start with 2d10 because it doesn't change the math as much. 3d6 changes quite a lot.

That said, if you really want to get on board with this edition as it is, you'll want to stop thinking of the end DC as your competence. You shouldn't be rolling for the mundane, only the fantastic. So some circumstance is preventing you from achieving your full potential, and the GM is accounting for it.

[PeteNutButter]

Where exactly are you seeing this problem? Is it attack rolls, skills, saves, or all of the above? Each +1 is 5%, so that a character even at lvl 1 has likely +5 (2 from proficiency, and 3 from 16 in the appropriate ability) at whatever the character is good at. That makes the character 25% more likely to succeed than a base 10 non-proficient. 25% can still be very swingy with a d20.

That isn't that much, but by high level it gets closer to +10 to 15. That's such a massive difference that it begins to trivialize the die roll. I have problems with 5e too, but 3.5 had the same issue at low lvls. And it got so out of hand at higher lvls with fighter types getting more than 20 or 30 to their attack rolls, and the monsters ACs had to keep pace, making the nonfighters only hit on a natural 20 (ignoring touch ac).

The reality is both editions have a sweet spot and they are about the same somewhere in the mid-levels. It's a characteristic problem of the d20 in general, being so swingy. If it really bothers you, consider homebrewing the d20 into 3d6. Bell curves do a magical thing to bounded accuracy.

EDIT: boss called > delayed post > super sniped

[obryn]

The downside to curved RNGs is that they make extraordinary success less likely.

So if, for example, your Fighter gets grabbed by a Giant, your prospects become really, really bleak. Or, even more likely, it gets even rockier when an enemy spellcaster is targeting one of your weak saves.

This isn't to say it's a bad idea, at all. It's just often overlooked when people are considering 2d10 or 3d6.

[Ninja_Prawn]

Way fewer rolls is probably a given; the whole "only when failure is interesting
Letting people with proficiency take 10 most of the time would probably help a lot out of combat, at least.

The first one should be a given with a decent DM, and the second (i.e. increase the number of passive checks) is definitely something that more people should do more often.

And like Flashy said, a level 1 character *is* pretty ordinary. I recently built a level 10 AT with a +19 bonus to lockpicking checks. With Mage Hand Legerdemain, she could do it remotely and invisibly. She could basically laugh at any attempt to lock anything. But that kind of power is restricted to higher level characters who've got a few magic items under their belt. If you're starting from level 1, you have to earn it.

[PeteNutButter]

The downside to curved RNGs is that they make extraordinary success less likely.

So if, for example, your Fighter gets grabbed by a Giant, your prospects become really, really bleak. Or, even more likely, it gets even rockier when an enemy spellcaster is targeting one of your weak saves.
This isn't to say it's a bad idea, at all. It's just often overlooked when people are considering 2d10 or 3d6.

It definitely ups the realism factor, and I'd suggest not going with the 3d6 unless you understand basic statistics. A high AC/DC becomes increasingly harder to hit with each point. Maybe having crit's start at 16 or 17 can alleviate some of this.

EDIT: heh said AC/DC :smallcool:

[Segev]

If you like the notion of the 3d6's bell curve, but want to keep the range of 1-20, you can make it 3d6+1d4, but treat the 1 and 2 on the d4 as -1 and -2, and the 3 and for as +1 and +2. The resulting probability curve looks like this (http://anydice.com/program/7c86).


Another option, for skills, might be to re-introduce "take 10" as a rule. When not stressed, you can just say you rolled a 10 on the d20. It doesn't invalidate the Rogue's talent feature; they can roll and then take 10 if they got less than that on the d20.

The other thing you can do is angle for Advantage. It will generally inflate your roll by 3-5 points on average (though it won't change your cap).

[Satinavian]

If you like the notion of the 3d6's bell curve, but want to keep the range of 1-20, you can make it 3d6+1d4, but treat the 1 and 2 on the d4 as -1 and -2, and the 3 and for as +1 and +2. The resulting probability curve looks like this (http://anydice.com/program/7c86).I would prefer 2d8+d6-2 for the range. Which is the same as 2d8+d6 and +2 to all DCs.

[N810]

Or 2d6 + 1d8 (range of 3-20) :smallwink:

[Kurald Galain]

I dunno. What do you guys do to make the game feel less stupidly swingy?

That's by design. But the easiest fix would be to double all proficiency bonuses, and increase DCs by 3.

[eastmabl]

Expand the concept of passive checks. Unless you're actively trying to do something that's difficult or where there's always a chance of failure (e.g. natural 1 on attacks), assume that the player is making a passive check.

If you assume a bonus of +0 or greater, that means you're handling most easy tasks without a problem. For low level specialist with +5 bonus to a skill, you're making most medium checks without breaking a sweat.

As players reach higher levels, they'll start to hit hard DCs with the passive skill checks.

***

Alternatively, if your pain point is that the dice rolls matter too much, read up on "failing forward." It takes the "success or failure" dichotomy of d20 games and puts it into the rubbish bin.

Basically, instead of complete successes or complete failures, some failures will become successes with a cost. In 5e, my rule of thumb is DC check minus 5.

[Tanarii]

Way fewer rolls is probably a given; the whole "only when failure is interesting That should already be a given. A better way to phrase it is "only when the failure has consequences".

This is my favorite write up on how to adjudicate actions, by Angry DM. It's a bit long, but that's because he's delving into the underlying theory. In particular 3a & 3b deals with "should I make a check", but the whole thing is a good read:
http://angrydm.com/2013/04/adjudicate-actions-like-a-boss/

[Grod_The_Giant]

Individual replies below, but overall impressions:

As the issue seems to be one of probability, there seem to be two solutions: mess with the dice, or increase the bonuses. The former messes with feel, the latter requires changing a lot of numbers. Of the two, I kind of like the idea of tying a more consistent die roll to proficiency. Say that when proficient with an ability check, you can roll 3d6 or 1d10+10 instead of a d20. Except maybe in combat? How does that sound?


I think the easiest fix for this is to start at 3rd level rather than first. Here's why I say this.
My problem is with probability though-- not class features. I don't need a level 1 wizard to be a great mage, but I'd expect him to have better Arcana rolls than the Barbarian more than a quarter of the time.


Where exactly are you seeing this problem? Is it attack rolls, skills, saves, or all of the above? Each +1 is 5%, so that a character even at lvl 1 has likely +5 (2 from proficiency, and 3 from 16 in the appropriate ability) at whatever the character is good at. That makes the character 25% more likely to succeed than a base 10 non-proficient. 25% can still be very swingy with a d20.
This particular DM made skills the most offensive, I think (he made the gnome make two separate checks to get on and off a barstool :smallfurious:), but the issue is general. 25% just isn't enough, I don't think.


That isn't that much, but by high level it gets closer to +10 to 15. That's such a massive difference that it begins to trivialize the die roll. I have problems with 5e too, but 3.5 had the same issue at low lvls. And it got so out of hand at higher lvls with fighter types getting more than 20 or 30 to their attack rolls, and the monsters ACs had to keep pace, making the nonfighters only hit on a natural 20 (ignoring touch ac).
I'd argue that 50% advantage over a schmuck (+10) is reasonable bonus for a competent character, and 100% (+20, excluding crits) is solid for a master. I like Bounded Accuracy keeping DCs in reach for everyone; I just think that the numbers got set waaaay too low for a die as big as a d20.


And like Flashy said, a level 1 character *is* pretty ordinary. I recently built a level 10 AT with a +19 bonus to lockpicking checks. With Mage Hand Legerdemain, she could do it remotely and invisibly. She could basically laugh at any attempt to lock anything. But that kind of power is restricted to higher level characters who've got a few magic items under their belt. If you're starting from level 1, you have to earn it.
I don't want power, I just want some semblance of consistency. A specialist should be able to reliably do level-appropriate things he specializes in; advancement should mean that he gets to do bigger and better things.


That's by design. But the easiest fix would be to double all proficiency bonuses, and increase DCs by 3.
Hmm... what would you do about AC? The same +3 increase?


Alternatively, if your pain point is that the dice rolls matter too much, read up on "failing forward."

That should already be a given. A better way to phrase it is "only when the failure has consequences".
I'm familiar with the concepts, yeah.

[Oramac]

This particular DM made skills the most offensive, I think (he made the gnome make two separate checks to get on and off a barstool :smallfurious:), but the issue is general. 25% just isn't enough, I don't think.

o_O

Whiskey Tango Foxtrot. Your problem is the DM, not the game. That's ridiculous. For that, I wouldn't even make the Gnome make one check unless he was 3 sheets to the wind hammered.

[GlenSmash!]

o_O

Whiskey Tango Foxtrot. Your problem is the DM, not the game. That's ridiculous. For that, I wouldn't even make the Gnome make one check unless he was 3 sheets to the wind hammered.

Even if he was three sheets to the wind hammered I probably would not call for a check unless there was a pressing need to get on and off that stool in a timely fashion.

To the OP. I'm tempted to give advice on how I adjudicate, but it seems like it won't matter. You already seem to know what to do, but you aren't the DM.

So maybe you could pass on some advice?

To address the wizard vs barbarian on the Arcana check: The DM determines the DC for an ability check where the outcome is uncertain. If I decide the DC for a wizard to recall some bit of arcane knowledge is INT (Arcana) DC 10 I'm perfectly within my rights to set the DC as INT(Arcana) DC 20 when the Barbarian attempts to recall the same bit of knowledge. Or even have the Barbarian flat out fail without any dice roll at all. I don't need to mess with Proficiency, or Advantage, or turning +1s to +3s to make the game work.

[Regulas]

To begin with 5e relies a bit more on DM fiat or rulings. For example even if your barbarian made an Arcana role, assuming that he got a 20 what should he find out? It's arcane magic! Because that's probably the limit of what he knows about magic, unless maybe he's seen that magic in an earlier session specifically.


It sounds like the problem is that your DM is just allowing people to do things their characters shouldn't be able to or is failing to consider what your characters individual traits are when ruling things.

The reason the accuracy is so bounded even for skills is specifically so that the die is relevant (cause when you have a +20 it really isn't at all). That doesn't mean that a roll of a 3+4 for 7 should be a total failure you should still be told obvious basics, you just won't know the critical info the DC required.

[Grod_The_Giant]

Even if he was three sheets to the wind hammered I probably would not call for a check unless there was a pressing need to get on and off that stool in a timely fashion.
Yeah, he was... not ideal. I don't think I'm going back to that game. But I AM running one of my own tomorrow, so...


To the OP. I'm tempted to give advice on how I adjudicate, but it seems like it won't matter. You already seem to know what to do, but you aren't the DM.
No please, I'd like to hear.


To address the wizard vs barbarian on the Arcana check: The DM determines the DC for an ability check where the outcome is uncertain. If I decide the DC for a wizard to recall some bit of arcane knowledge is INT (Arcana) DC 10 I'm perfectly within my rights to set the DC as INT(Arcana) DC 20 when the Barbarian attempts to recall the same bit of knowledge. Or even have the Barbarian flat out fail without any dice roll at all. I don't need to mess with Proficiency, or Advantage, or turning +1s to +3s to make the game work.

To begin with 5e relies a bit more on DM fiat or rulings. For example even if your barbarian made an Arcana role, assuming that he got a 20 what should he find out? It's arcane magic! Because that's probably the limit of what he knows about magic, unless maybe he's seen that magic in an earlier session specifically.
I mean, you can fiat everything like that, but... part of the point of having a game with established rules is that you don't--or shouldn't--need to do that sort of fudging. I don't want to have to already decide whether or not my players succeed--not in the sort of dramatic moment when you're supposed to be actually rolling. I shouldn't need to say "well, it's a hard check for a barbarian, but an easy one for a wizard;" that's not why we have character sheets. When a check needs to be made, there's supposed to be an objective DC: DC 13 to avoid the wizard's spell, DC 11 to stab the goblin, DC 15 to climb the smooth wall. The game tries to present rules for "can I do this;" if it requires the DM to consistently overrule it, that's a major problem. 5e has the rules it needs for that sort of thing, but the math seems to be letting it down.

[Regulas]

I mean, you can fiat everything like that, but... part of the point of having a game with established rules is that you don't--or shouldn't--need to do that sort of fudging. I don't want to have to already decide whether or not my players succeed--not in the sort of dramatic moment when you're supposed to be actually rolling. I shouldn't need to say "well, it's a hard check for a barbarian, but an easy one for a wizard;" that's not why we have character sheets. When a check needs to be made, there's supposed to be an objective DC: DC 13 to avoid the wizard's spell, DC 11 to stab the goblin, DC 15 to climb the smooth wall. The game tries to present rules for "can I do this;" if it requires the DM to consistently overrule it, that's a major problem. 5e has the rules it needs for that sort of thing, but the math seems to be letting it down.

Well the DM shouldn't be telling the players necessarily that you failed because your a Barbarian, however on the flip side I don't think there should be a need to tell the Barbarian every time that you don't know anything on the subject, once really should be enough. But yes it should be pretty obvious without needing more then a second thought for the DM to know that a Barbarian can't tell what spell that is no matter what he roles.

And I can understand it would take some getting used to, 3.5 edition pretty much is the "rules for everything" whereas 5th edition is explicitly aiming to simply be "general framework", with a stronger emphasis on roleplaying and decision making.

I would argue the math isn't meant to be balanced, it's meant to provide a diverse array of posibilitesl

[Biggstick]

I mean, you can fiat everything like that, but... part of the point of having a game with established rules is that you don't--or shouldn't--need to do that sort of fudging. I don't want to have to already decide whether or not my players succeed--not in the sort of dramatic moment when you're supposed to be actually rolling. I shouldn't need to say "well, it's a hard check for a barbarian, but an easy one for a wizard;" that's not why we have character sheets. When a check needs to be made, there's supposed to be an objective DC: DC 13 to avoid the wizard's spell, DC 11 to stab the goblin, DC 15 to climb the smooth wall. The game tries to present rules for "can I do this;" if it requires the DM to consistently overrule it, that's a major problem. 5e has the rules it needs for that sort of thing, but the math seems to be letting it down.


Not having ever DM'ed a game, I don't "know" the actual preparation that a DM puts into creating everything, including out of combat and in combat checks. What you're bringing up though sounds like that the checks are something that you've already set-up anyways; the rules didn't decide your particular DC's for your very specific instance of a Religion or History or Athletics check, you did. You've already gone so far as to create a unique check as it is, adding a note that the wizard would have a lower DC to make should be easy if the information is already there (or a lower DC for the barbarian to climb said wall without assistance).

[Grod_The_Giant]

But yes it should be pretty obvious without needing more then a second thought for the DM to know that a Barbarian can't tell what spell that is no matter what he roles.

And I can understand it would take some getting used to, 3.5 edition pretty much is the "rules for everything" whereas 5th edition is explicitly aiming to simply be "general framework", with a stronger emphasis on roleplaying and decision making.

I would argue the math isn't meant to be balanced, it's meant to provide a diverse array of posibilitesl
Why should it be? The wizard is only 30% better at Arcana than the Barbarian; it's right there on their sheets. They both have a reasonable chance of making the DC 15 check. I can't find anything in the books that says "ignore the math when you don't like it." There are rules to the game, rules which we paid a great deal of money for someone to write. Saying "ignore the rules" isn't helping. Saying "5e is more about roleplaying" isn't an excuse; I've played games like Fate that pay more than lip service to that idea and they still take the time to make sure the math works.

[Regulas]

Why should it be? The wizard is only 30% better at Arcana than the Barbarian; it's right there on their sheets. They both have a reasonable chance of making the DC 15 check. I can't find anything in the books that says "ignore the math when you don't like it." There are rules to the game, rules which we paid a great deal of money for someone to write. Saying "ignore the rules" isn't helping. Saying "5e is more about roleplaying" isn't an excuse; I've played games like Fate that pay more than lip service to that idea and they still take the time to make sure the math works.

I guess the difference is that to me the notion that the barbarian who hasn't by background studied arcana can't succeed at arcana is so obvious that it doesn't even take a thought. In my eyes it's not a rule because it's so obvious and self evident as to be an axiom, and the rule was excluded because it would just be a waste of space on the page.

[Tanarii]

I guess the difference is that to me the notion that the barbarian who hasn't by background studied arcana can't succeed at arcana is so obvious that it doesn't even take a thought. In my eyes it's not a rule because it's so obvious and self evident as to be an axiom, and the rule was excluded because it would just be a waste of space on the page.Or to put it another way, the designers wanted the rules to be flexible enough for the DM to decide when a check was called for and what success & failure looked like. As opposed to trying to hard-code every possible DC for all characters.

What a given character might recall based on a 'knowledge/lore' skill type check *can* be different based on their background, at the DM's discretion. Because it's not hard-coded in. Or the DM can decide that all PCs have the same chance to do everything (identical DCs) for any given task, again at their discretion.

I'll note two things:
1) The skill system is meant to be quick and dirty. It's intentionally designed to be fast and easy.
2) ALL d20 rolls in the game are intentionally designed so that PCs will remain not only within striking distance of each other, AND not to have the bonuses outweight the chance of the die roll. It's very intentional that a 1st level untrained no-ability score character and a trained max proficient max ability score character are only 60% chance of success apart, and that 60% is less than the entire span of the die being checked in the first place (a d20). Because the Designers wanted all PCs to be approximately in range of each other, and for the die roll to always be meaningful.

The latter is a reaction to how much people complained about bonuses to die rolls not only being hugely far apart between trained and untrained in 3e, but also totally overshadowing the d20 die roll.

[pwykersotz]

Or to put it another way, the designers wanted the rules to be flexible enough for the DM to decide when a check was called for and what success & failure looked like. As opposed to trying to hard-code every possible DC for all characters.

What a given character might recall based on a 'knowledge/lore' skill type check *can* be different based on their background, at the DM's discretion. Because it's not hard-coded in. Or the DM can decide that all PCs have the same chance to do everything (identical DCs) for any given task, again at their discretion.

I'll note two things:
1) The skill system is meant to be quick and dirty. It's intentionally designed to be fast and easy.
2) ALL d20 rolls in the game are intentionally designed so that PCs will remain not only within striking distance of each other, AND not to have the bonuses outweight the chance of the die roll. It's very intentional that a 1st level untrained no-ability score character and a trained max proficient max ability score character are only 60% chance of success apart, and that 60% is less than the entire span of the die being checked in the first place (a d20). Because the Designers wanted all PCs to be approximately in range of each other, and for the die roll to always be meaningful.

The latter is a reaction to how much people complained about bonuses to die rolls not only being hugely far apart between trained and untrained in 3e, but also totally overshadowing the d20 die roll.

^This is accurate. If you're looking at specialists being skilled enough not to fail, you just have to embrace that mindset as the DM decides whether a check is worth calling for. It's no longer the nature of the beast to use the skill system to extrapolate common activities, it's used specifically for extraordinary circumstances.

[georgie_leech]

Look at it this way. In 3.x, Circumstance Bonuses and Penalties were given out to reflect different circumstances at the DM's discretion. Like, say a Rogue is trying to sweet talk a noble into getting them an invitation some important event. This Rogue happens to have known the noble previously, dresses to impress, and sweetens the deal by giving them a gift that past experience has taught that the noble likes. The DM decides this is worth a +5 to their diplomacy check.

But say it was a brutish Barbarian dressed in animal hides, scuffed and dirty from travel with the foul odor of not having bathed in weeks clinging to him. The Barbarian angrily demands an invitation. The DM decides this is a -5 to the Check.

Putting that together, there's a +10 difference between the Circumstance Modifiers between the two against the same DC. In 5e, the DM is instructed to adjust the DC's instead, so if the base DC would have been 15, the Rogue gets a DC of 10 (and possibly Advantage depending on how well the Rogue read the situation) and the Barbarian gets a DC of 20. The Circumstance Modifiers just adjust the DC instead of applying to the roll.

[Zaq]

I guess the difference is that to me the notion that the barbarian who hasn't by background studied arcana can't succeed at arcana is so obvious that it doesn't even take a thought. In my eyes it's not a rule because it's so obvious and self evident as to be an axiom, and the rule was excluded because it would just be a waste of space on the page.

If they wanted certain checks to only be possible with proficiency, they had plenty of space to say that. But they didn't. They half-included an element of "only proficient people need apply" (when crafting an item, you have to be proficient with the appropriate tool or you can't even try), so they weren't unfamiliar with the concept, but they didn't say that. There's no rule that a Barbarian can't try to roll Arcana. Nor would it necessarily be an improvement. If you're playing a Barbarian and you don't want your character to know about Arcana, don't roll for it. But if a Barbarian chooses to roll Arcana, they should be able to. After all, you don't get many skill proficiencies—if we start closing off certain rolls to be "trained only," then you end up with vanishingly few things you can actually do. (I would argue that it's so bloody hard to be good at anything in 5e that it's already the case that there's vanishingly few things you can do reliably in the first place, but that may or may not be a separate matter.)

I'm with Grod that it doesn't make sense that something would be DC 10 for the Wizard and DC 20 for the Barbarian—literally the entire point of having stats is that we don't have to determine individual DCs for individual people; you put up a DC and roll your bonus against it. If the Wizard is better at it than the Barbarian, that's represented by the Wizard having a higher bonus. And the reverse is true when it comes time to roll Athletics or whatever. That's why the Wizard has a STR score and why the Barbarian has an INT score—so we don't have to make up new DCs for every person.

That said, while I don't have a problem with a Barbarian rolling Arcana, I do agree with the greater problem that a Wizard can't end up more than like 40-60% better than the Barbarian at it. That bothers me a lot. But the solution to "the Wizard isn't that much better at Arcana than the Barbarian" is not "ban the Barb from rolling Arcana!" Honestly, that's how we got crap like 3.5 requiring a specific class feature to be able to try to find traps (no matter how good you were at Search and Disable Device, if you didn't have Trapfinding, you couldn't find anything above DC 20 by RAW). That was dumb then and it would be dumb now.

I repeat, if YOUR Barbarian doesn't want to roll Arcana, then don't roll Arcana. But if the rules say that my Barb has a 25% chance of knowing a DC 15 Arcana item, why should I be prevented from trying? The problem isn't that the Barbarian can succeed; the problem is that the Wizard can't outpace the Barbarian that much.

I mean, if we go down that road, what other restrictions are "so obvious and self evident as to be an axiom"?

[Tanarii]

The problem isn't that the Barbarian can succeed; the problem is that the Wizard can't outpace the Barbarian that much.
*Puts on programmer hat*
That's not a problem. It's a feature.

[joaber]

I think the greater point of 5e is exactly bounded accuracy. You need the dice to succeed. Why you rolled the dice in early editions if you only need a 3? In the end, you're playing dice, that's why the game was made and everybody will be in the same boat.

Now you always have a chance to fail, but if you play wise, you can reduce that chance with advantage, ispiration points, expertise, bless, guidance, bardic inspiration, precision strike, lucky feat. Limited resources that allow you to break bounded accuracy.

Easiest way to go is thinking in ways to get advantage "always".

If you still can't handle BA, the easiest way to go is don't play 5e

[TheTeaMustFlow]

This does become much less of a thing (I hesitate to say 'problem', because I'm not sure I have a problem with it) at higher levels. Our campaign just went epic, and I've often been finding that my Rogue could* get better on a 1 than the rest of the party can get on a 20. Admittedly, a lot of that's due to Reliable Talent, but even without it, I think the range of skill modifiers in the party is 17 or so, and most of us are rolling at least +13 to hit (I think I was rolling +16*). Rolls have become a lot more predictable.

*Prior to an unrelated case of sudden onset being a dragon syndrome. Protip: Don't let your familiars draw from the Deck of Many Things. And don't trust them with Wishes.

[Telok]

I think the 5e may just be schitzo over ability scores. The barbarian may be 8 Int, but an ogre is 5 and a gorilla is 6. So what does inteligence, as measured by the bonus, really mean?

The fighter is str 16 and trained in athletics, the wizard is str 8, and the cat has str 3. We're loking at +5 vs -1 vs -4 and a d20 roll. The cat has the raw strength to try to out wrestle the fighter half the time. The cat has to roll well but the rules tell us that the cat can push the fighter because it's strong enough. We're looking at something like 8th level before there's no chance of a house cat pushing an armored fighter around. Then we've got the ogre at -3 versus the wizard at +5 (first level) to +10 (eightth? level). Sure the wizard usually knows more arcana than a creature dumber than an ape, but not always.

[Zaq]

I think the 5e may just be schitzo over ability scores. The barbarian may be 8 Int, but an ogre is 5 and a gorilla is 6. So what does inteligence, as measured by the bonus, really mean?

The fighter is str 16 and trained in athletics, the wizard is str 8, and the cat has str 3. We're loking at +5 vs -1 vs -4 and a d20 roll. The cat has the raw strength to try to out wrestle the fighter half the time. The cat has to roll well but the rules tell us that the cat can push the fighter because it's strong enough. We're looking at something like 8th level before there's no chance of a house cat pushing an armored fighter around. Then we've got the ogre at -3 versus the wizard at +5 (first level) to +10 (eightth? level). Sure the wizard usually knows more arcana than a creature dumber than an ape, but not always.

I'm inclined to agree. Ability scores have never meant much at the high end and at the low end, but the whole "bounded accuracy" fetishization has exacerbated that to a great degree, since you no longer have other bonuses to augment (or compensate for) the base stats. The cat pushing around the armored Fighter is a fantastic example (and way more entertaining to me than a housecat killing a level 1 Wizard). I mean, technically, the Shove rules (PHB pg. 195) specify that you have to be within one size of your target (so the 16 STR Fighter would have to be a halfling or a gnome or something), but still, it's really just a technicality preventing the actual numbers from getting weird.

But yeah. Ability scores have always been wacky, but "bounded accuracy" makes them simultaneously super important (because they're like your only source of numbers) and super pointless (for the reasons you've mentioned).

[Gurifu]

I've only been experimenting with it for a short time, but I've had a positive impression of my change to 2d10 instead of 1d20 in the game I started DMing. Improve crit ranges by 2 to keep the probability of a crit the same despite the change to a curved dice system.



d20 % 2d10 %
20 5% 18-20 6%
19-20 10% 17-20 10%
18-20 15% 16-20 15%


Probability of most-skilled rolling higher than least-skilled at level 3 (+5 vs -1)
d20: 77.25%
2d10: 82.4%

Probability of beating dc13 most vs least-skilled at level 3 (+5 vs -1)
d20: 65% (+5) vs 35% (-1), split of 30%, 1.9 times as likely
2d10: 79% (+5) vs 28% (-1), split of 51%, 2.8 times as likely

Most changes are subtle. The only really noticeable one is at the extreme high end of DCs, which only regularly occurs in AC. AC20, often achievable by 3, makes you almost untouchable by the +5 attacks being thrown at you at that level (10.5% chance to hit). This strikes me as realistic, but not particularly balanced: it makes for a significant low-level buff to armored bruisers, who are already arguably the strongest low-level characters. (That being said, cover, often overlooked, can easily boost your AC to the point at which only a crit can hit you, even for a modest 3rd level AC of 16.)

[caden_varn]

I do agree that it seems silly that the Wizard has such a small advantage over the Barbarian, but looking at it another way, it means the party still has a chance to make arcana checks if they don't have an arcane character, and there is no guarantee that each party would have one. This is especially helpful if you are running a published module, as they won't be tuned to your party composition.

[djreynolds]

But the bonuses aren't small and choices change in creation and what level you are.

Starting at level 10, Heavy Armor master seems silly.

At level one for a human variant fighter who also grab +1 defensive style, you are the big hero every combat. Or the fighter with archery style.

You paladin may not live to smite.

We are starting tonight with the Ravenloft, or at least rolling up level 1 characters. Starting off an archery fighter, say would elf, its like having a 20 in dexterity, but you have a 16. Archery style is huge. But as the game progresses, yeah, that wizard becomes dangerous. Getting advantage at second level, or having shield master at 1st, these are game changers for players playing in the moment. And its tough to even think about multiclassing that early, fighters are getting so much by 2,3,4,6 its tough to stay the course and become the wizard or what have you.

A wizard may have a lot hit points now depending on ability selection, and look very barbarian like, like 9 or 10. A barbarian could have 16. And if you go the books way with standardized hit points, level 2 a barbarian is getting 7 a level vs a wizards 4.

[Lines]

I do agree that it seems silly that the Wizard has such a small advantage over the Barbarian, but looking at it another way, it means the party still has a chance to make arcana checks if they don't have an arcane character, and there is no guarantee that each party would have one. This is especially helpful if you are running a published module, as they won't be tuned to your party composition.

Why would you tune anything to party composition? The world is the world, I'm not going to have its contents magically change just because they brought a rogue and druid instead of a barbarian and a monk. Different classes are good at different things, it makes those differences meaningless if you're just going to adjust the world around them and adjusting the world around anything destroys the feeling that it's a living breathing place, turns it into a video game centered around the PCs.

[Lines]

But the bonuses aren't small and choices change in creation and what level you are.

Starting at level 10, Heavy Armor master seems silly.

At level one for a human variant fighter who also grab +1 defensive style, you are the big hero every combat. Or the fighter with archery style.

You paladin may not live to smite.

We are starting tonight with the Ravenloft, or at least rolling up level 1 characters. Starting off an archery fighter, say would elf, its like having a 20 in dexterity, but you have a 16. Archery style is huge. But as the game progresses, yeah, that wizard becomes dangerous. Getting advantage at second level, or having shield master at 1st, these are game changers for players playing in the moment. And its tough to even think about multiclassing that early, fighters are getting so much by 2,3,4,6 its tough to stay the course and become the wizard or what have you.

A wizard may have a lot hit points now depending on ability selection, and look very barbarian like, like 9 or 10. A barbarian could have 16. And if you go the books way with standardized hit points, level 2 a barbarian is getting 7 a level vs a wizards 4.

Which doesn't change the fact that the wizard who has spent his life studying the arcane is only slightly more likely to recall a specific arcane fact or analyze a magical pattern than the barbarian who hasn't read a book or seen a spell in his life.

It changes a bit as you level, by the end the barbarian will only sometimes be recalling arcane trivia that the wizard didn't know - observe the following situation. Zazax the Loremaster, master wizard with proficiency bonus 6 and int bonus 5, most knowledgable in magic in all the worlds vs Krug the great at hitting things, barbarian who has reached level 20 and still hasn't touched a book in seeing a rune and trying to work out from the way it glows which of the arcane universities from centuries past it is from. DC is 17, which means Krug has a 20% chance to know and Zazax has a 70% chance.

Which means given a bunch of such checks, there will be several instances where the master arcanist who has dedicated his life to magical lore is stumped and Krug the still-only-wipes-his-ass-with-pages will be all like 'oh yeah, that's definitely from Drarell's Comments on the Harnessing of Residual Transmutation, vol IX'. And that won't be a freak occurrence that baffles everyone, it'll have a one in five chance of happening every time Zazax can't think of the answer.

[ad_hoc]

Way fewer rolls is probably a given; the whole "only when failure is interesting


This is the way the game is designed. It's not 3.x and if you are playing it as though it is, you're going to have a bad time.

Characters are also not experts of skills. Even those with expertise aren't the true experts of the worlds they inhabit.

They have abilities which make them great at other things.

Leave the blacksmithing to the blacksmiths.

[djreynolds]

Which doesn't change the fact that the wizard who has spent his life studying the arcane is only slightly more likely to recall a specific arcane fact or analyze a magical pattern than the barbarian who hasn't read a book or seen a spell in his life.

It changes a bit as you level, by the end the barbarian will only sometimes be recalling arcane trivia that the wizard didn't know - observe the following situation. Zazax the Loremaster, master wizard with proficiency bonus 6 and int bonus 5, most knowledgable in magic in all the worlds vs Krug the great at hitting things, barbarian who has reached level 20 and still hasn't touched a book in seeing a rune and trying to work out from the way it glows which of the arcane universities from centuries past it is from. DC is 17, which means Krug has a 20% chance to know and Zazax has a 70% chance.

Which means given a bunch of such checks, there will be several instances where the master arcanist who has dedicated his life to magical lore is stumped and Krug the still-only-wipes-his-ass-with-pages will be all like 'oh yeah, that's definitely from Drarell's Comments on the Harnessing of Residual Transmutation, vol IX'. And that won't be a freak occurrence that baffles everyone, it'll have a one in five chance of happening every time Zazax can't think of the answer.

Excellent example.

But in defense of Krug, he has had to survive wizards trying to kill him for 20 levels. He knows the sound of a wizard casting and what he might be casting, hence his advantage on dexterity stuff he can see. Magic isn't exclusive to wizards, they are trying to master it. But a barbarian, has killed a lot of casters and Zazax has told him how to kill them (he wants their spell book).

Just like Zazax can swing a staff really well, same +6 as Krug, cause Krug showed him how to fight.

Just tweak it. Worst thing about 3.5, my rogue in the cemetery. Digging his own grave practically. NO more of that nonsense. Zombie, you're arm is falling off, let me take it off of you. Sneak Attack

[Lines]

This is the way the game is designed. It's not 3.x and if you are playing it as though it is, you're going to have a bad time.

Characters are also not experts of skills. Even those with expertise aren't the true experts of the worlds they inhabit.

They have abilities which make them great at other things.

Leave the blacksmithing to the blacksmiths.

Why not? If I have a level 20 wizard with training in arcana and 20 int and +6 proficiency why am I not a true expert in the arcane?


Excellent example.

But in defense of Krug, he has had to survive wizards trying to kill him for 20 levels. He knows the sound of a wizard casting and what he might be casting, hence his advantage on dexterity stuff he can see. Magic isn't exclusive to wizards, they are trying to master it. But a barbarian, has killed a lot of casters and Zazax has told him how to kill them (he wants their spell book).


Just tweak it. Worst thing about 3.5, my rogue in the cemetery. Digging his own grave practically. NO more of that nonsense. Zombie, you're arm is falling off, let me take it off of you. Sneak Attack

Just like Zazax can swing a staff really well, same +6 as Krug, cause Krug showed him how to fight.

But that's still the case if you got to level 20 hewing through armies of githyanki, only fighting psions and warriors and never encountering a spell other than what Zazax was doing. At level 1 krug had the same chance to work it out as he has at 20, even if level 1 krug just came from the steppes and has never heard of magic before.

Edit for edit:

Just tweak it. Worst thing about 3.5, my rogue in the cemetery. Digging his own grave practically. NO more of that nonsense. Zombie, you're arm is falling off, let me take it off of you. Sneak Attack
Sentence weird and disjointed, not sure what you mean. Are you saying that you didn't like that sneak attack didn't work against the undead in 3.5 unless you have penetrating strike, grave strike, a greater truedeath crystal or similar?

[DanyBallon]

Which doesn't change the fact that the wizard who has spent his life studying the arcane is only slightly more likely to recall a specific arcane fact or analyze a magical pattern than the barbarian who hasn't read a book or seen a spell in his life.

It changes a bit as you level, by the end the barbarian will only sometimes be recalling arcane trivia that the wizard didn't know - observe the following situation. Zazax the Loremaster, master wizard with proficiency bonus 6 and int bonus 5, most knowledgable in magic in all the worlds vs Krug the great at hitting things, barbarian who has reached level 20 and still hasn't touched a book in seeing a rune and trying to work out from the way it glows which of the arcane universities from centuries past it is from. DC is 17, which means Krug has a 20% chance to know and Zazax has a 70% chance.

Which means given a bunch of such checks, there will be several instances where the master arcanist who has dedicated his life to magical lore is stumped and Krug the still-only-wipes-his-ass-with-pages will be all like 'oh yeah, that's definitely from Drarell's Comments on the Harnessing of Residual Transmutation, vol IX'. And that won't be a freak occurrence that baffles everyone, it'll have a one in five chance of happening every time Zazax can't think of the answer.

Such a situation just won't happen in my game, as unless Krug have any reason to know such thing, he will auto-fail by DM fiat. But if, let say, during its adventuring career, Krug came upon that book, then he could have a chance to roll an arcana check and remember that this large leather cover book with shiny writing was the one he's been told not to touch in the wizard tower that one time "you know that book Damien's blablating, about resident transmuter?"
- "You mean Drarell's Comments on the Harnessing of Residual Transmutation, vol IX? Yeah you're right."

[Lines]

Such a situation just won't happen in my game, as unless Krug have any reason to know such thing, he will auto-fail by DM fiat. But if, let say, during its adventuring career, Krug came upon that book, then he could have a chance to roll an arcana check and remember that this large leather cover book with shiny writing was the one he's been told not to touch in the wizard tower that one time "you know that book Damien's blablating, about resident transmuter?"
- "You mean Drarell's Comments on the Harnessing of Residual Transmutation, vol IX? Yeah you're right."

So that first part means you are agreeing that the randomness of skill checks is bad, right? I mean if you're having to alter how the rules work, they're plainly not working that well.

This is why I liked 3.x trained skills, some skills you couldn't make checks for unless you had skill ranks in them. Anyone can roll to climb a tree, but you couldn't make a spellcraft check unless you had some reason to actually know something about it.

[DanyBallon]

Why not? If I have a level 20 wizard with training in arcana and 20 int and +6 proficiency why am I not a true expert in the arcane?

Zazax a 20th level wizard with Int 20 and proficient in Arcana have +11 to Intelligence (Arcana) checks while Krug the 20th level barbarian who isn't proficient in Arcana at all and have only Int 10, gets a +0

With such difference it's safe to say that Zazax is far more proficient than Krug in order to find obscure arcana lore. Which again, he should be able to roll only if there's a slight chance for him to know about it, otherwise it's an auto-fail. On the other hand Zazax shouldn't have to roll for more common lore that a great wizard like him should already know about.

That's the beauty of 5e, you don't have to roll for everything like it was in 3.P. You roll only if it makes sense or if it makes the game more fun.

[DanyBallon]

So that first part means you are agreeing that the randomness of skill checks is bad, right? I mean if you're having to alter how the rules work, they're plainly not working that well.

I'm not altering the rules. The rules states that you roll only if there's a chance for failure (or success), and it's up to the DM to decide.


This is why I liked 3.x trained skills, some skills you couldn't make checks for unless you had skill ranks in them. Anyone can roll to climb a tree, but you couldn't make a spellcraft check unless you had some reason to actually know something about it.

Yet with 3.X trained skills, Krug wouldn't be able to do what I just describe because he won't spend a single skill point in arcana to recall a book he has seen once in it's life.

5e skill system is much more flexible, but it requires that players rely more on the DM, instead of relying on hard coded rules for everything and more.

[Lines]

I'm not altering the rules. The rules states that you roll only if there's a chance for failure (or success), and it's up to the DM to decide.

Yeah, but the point of say a DC17 check is that the dice decide what the chance is. If there's no chance of success, you're saying that the check would be out of the players reach - say they're asking for something so ludicrous that it would be a DC35 check if it could be made at all, and since nobody is near that you don't give them a chance because there is none.

For a DC17 check, the barbarian has a chance, there's nothing to support not letting him make the check when he could potentially make it.

[Kurald Galain]

I do agree that it seems silly that the Wizard has such a small advantage over the Barbarian, but looking at it another way, it means the party still has a chance to make arcana checks if they don't have an arcane character, and there is no guarantee that each party would have one. This is especially helpful if you are running a published module, as they won't be tuned to your party composition.

It is "helpful" only if the adventure contains such bottlenecks as "ok, now your party must roll a difficult check on one particular skill only, and there's no other way to proceed".

But that's a bad way to write adventures.

[djreynolds]

A good DM and even good players know not to hog the spotlight. Is there a chance a novice can beat an expert? There has to be.
Otherwise why roll dice. Adjust rules. But a natural 20 is an automatic yes on all things, even concentration checks that may kill you.

Dice rolls and advantage and luck are much more potent in 5E than in 3.5.

Both games are fun.

[DanyBallon]

Yeah, but the point of say a DC17 check is that the dice decide what the chance is. If there's no chance of success, you're saying that the check would be out of the players reach - say they're asking for something so ludicrous that it would be a DC35 check if it could be made at all, and since nobody is near that you don't give them a chance because there is none.

For a DC17 check, the barbarian has a chance, there's nothing to support not letting him make the check when he could potentially make it.

A DC17 check is meant to be relatively hard to beat by someone, even if you are proficient, and can seems near impossible for people not trained. As long as the character as a reason to be trying this check he can try. But even with a DC 5 to read a wanted poster on a wall, I won't allow a illiterate Barbarian a check to be able to read what's written on the poster. He might be able to reconized it as a wanted poster, but nothing more.

[Lines]

A good DM and even good players know not to hog the spotlight. Is there a chance a novice can beat an expert? There has to be.
Otherwise why roll dice. Adjust rules. But a natural 20 is an automatic yes on all things, even concentration checks that may kill you.

Dice rolls and advantage and luck are much more potent in 5E than in 3.5.

Both games are fun.

A natural 20 is only an automatic hit. It isn't an auto success on a skill - if you're an average guy trying to balance on a cloud, the DC will be like 40 and no matter how many times you try you'll fail 100% of the time.

[Cazero]

Yeah, but the point of say a DC17 check is that the dice decide what the chance is. If there's no chance of success, you're saying that the check would be out of the players reach - say they're asking for something so ludicrous that it would be a DC35 check if it could be made at all, and since nobody is near that you don't give them a chance because there is none.

For a DC17 check, the barbarian has a chance, there's nothing to support not letting him make the check when he could potentially make it.

Your assumption is that the check is described by DC. It isn't. The check is described by words, and then a DC is extrapolated from those words.

A specific rune used in an obscure spell? It is impossible (DC close to 30) for an uneducated barbarian to know anything about it. But a trained wizard will simply find it challenging (DC around 17).

This allows the strength of bounded accuracy (you never need a specialist) to not break without preventing people to feel that their training doesn't matter. The obvious problem being that the skills should be enough to tell the difference. But with something as swingy as a d20, it's simply not possible to have that cake and eat it too.

[DanyBallon]

Your assumption is that the check is described by DC. It isn't. The check is described by words, and then a DC is extrapolated from those words.

A specific rune used in an obscure spell? It is impossible (DC close to 30) for an uneducated barbarian to know anything about it. But a trained wizard will simply find it challenging (DC around 17).

This allows the strength of bounded accuracy (you never need a specialist) to not break without preventing people to feel that their training doesn't matter. The obvious problem being that the skills should be enough to tell the difference. But with something as swingy as a d20, it's simply not possible to have that cake and eat it too.

While you're right, it's harder to do so with published adventure. And personally, I don't like to give different DC for a same task, i.e. crossing a chasm suspended to a rope, in my opinion it should be the same DC for eveyone, but the agile rogue might cross it with ease (proficient in acrobatic or athletics) while the frail and clumsy wizard may have a harder time doing so. The DM could as well decide that it's not a real challenge for the rogue and give him a auto-success, as still ask the wizard to roll.

[ravenkith]

...Reliable Talent...

It seems to me that the answer lies somewhere in the reliable talent class ability of the rogue.

Perhaps a series of feats?

Off the top of my head....

Feat: Journeyman (Skill)
Requirements - Level 4, Proficiency in skill to be chosen

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to a 10 if the actual number shown on the die is less than 10.

Feat: Master (Skill)
Requirements - Level 8, Proficiency in skill to be chosen and either Journeyman feat for skill to be chosen -OR- Reliable Talent class feature

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to a 15 if the actual number shown on the die is less than 15.

Theoretically, at least these feats would remove the chances of ridiculously low rolls for those characters that are truly dedicated to a single skill, thus increasing reliabilty...it wouldn't do anything to their skill maximum, however, so you'd still be shopping for bonuses. They don't really teach you how to be awesome, so much as how to suck less.

Combining these feats with Lucky may be worthwhile, or it may not.

Fortunately these feats also do not reduce the utility of reliable talent, or that of jack of all trades, as those abilities apply to multiple skills as opposed to a single one.

It might take a little monkeying with the specific numbers (maybe 8 and 12 instead of 10 and 15, or something), but I think this is probably as close as you can get to replicating that 'master of one skill' thing you are talking about, and frankly, I don't anticipate anyone at the table going "that's SOOO broken" when the opportunity cost is an ASI (2 if you want both feats), and you can't access these at first level.

I mean, when compared to shield master, which can give you evasion at level one, for instance, my first reaction is that it's not broken, but.....

What do you guys think?

EDIT: It might be necessary to change the level requirements from 4 to 8 and from 8 to 12, respectively...again, something to consider.

[Joe the Rat]

If you want a little more consistency in outcomes, I'd incline towards d20 for attacks and saves, 2d10 for ability checks.

This leaves combat swingy and random, and avoiding bad things inclined towards fortune - which I feel are the places that luck should play a big role. You also don't have to retweak crit ranges (and auto-fail ranges, if you really want the 5% auto win / lose in combat). Tending towards 11 on your ability checks makes your performance more consistent, and it's less likely (but not impossible) that a house cat will successfully grapple the barbarian, but still leaves room for potential (though reduced) greatness from anyone. This also means that passive checks and Reliable Talent are matching or beating 40% of possible rolls, instead of 50%. Same relationship to given DCs, but a little more probability space above the fixed number.

So what about (Dis)Advantage? The simple approach would be to roll an extra 2d10, but you might want to look at rolling 3d10, keeping the best / worst 2. Look at the probability curves, and find the mode that best fits how much you feel circumstances should affect performance.

[Lines]

Your assumption is that the check is described by DC. It isn't. The check is described by words, and then a DC is extrapolated from those words.

A specific rune used in an obscure spell? It is impossible (DC close to 30) for an uneducated barbarian to know anything about it. But a trained wizard will simply find it challenging (DC around 17).

This allows the strength of bounded accuracy (you never need a specialist) to not break without preventing people to feel that their training doesn't matter. The obvious problem being that the skills should be enough to tell the difference. But with something as swingy as a d20, it's simply not possible to have that cake and eat it too.

Neat! So if they're trying to climb up a difficult slope and the DC is 17 for Krug, who has his +7 strength and +6 proficiency and DC 30 for the wizard, who has no real experience at this because in normal circumstances he'd just use magic?

Because that's not how it works. The DC is the same for the both of them, Krug doesn't get an easier DC because he's a barbarian - because he's a barbarian, he has a greater modifier to his roll. You don't double dip on this, if you want the DC to be 13 easier for a class, give them +13 to that skill.

[Segev]

To begin with 5e relies a bit more on DM fiat or rulings. For example even if your barbarian made an Arcana role, assuming that he got a 20 what should he find out? It's arcane magic! Because that's probably the limit of what he knows about magic, unless maybe he's seen that magic in an earlier session specifically.


It sounds like the problem is that your DM is just allowing people to do things their characters shouldn't be able to or is failing to consider what your characters individual traits are when ruling things.

The reason the accuracy is so bounded even for skills is specifically so that the die is relevant (cause when you have a +20 it really isn't at all). That doesn't mean that a roll of a 3+4 for 7 should be a total failure you should still be told obvious basics, you just won't know the critical info the DC required.



I mean, you can fiat everything like that, but... part of the point of having a game with established rules is that you don't--or shouldn't--need to do that sort of fudging. I don't want to have to already decide whether or not my players succeed--not in the sort of dramatic moment when you're supposed to be actually rolling. I shouldn't need to say "well, it's a hard check for a barbarian, but an easy one for a wizard;" that's not why we have character sheets. When a check needs to be made, there's supposed to be an objective DC: DC 13 to avoid the wizard's spell, DC 11 to stab the goblin, DC 15 to climb the smooth wall. The game tries to present rules for "can I do this;" if it requires the DM to consistently overrule it, that's a major problem. 5e has the rules it needs for that sort of thing, but the math seems to be letting it down.


Look at it this way. In 3.x, Circumstance Bonuses and Penalties were given out to reflect different circumstances at the DM's discretion. Like, say a Rogue is trying to sweet talk a noble into getting them an invitation some important event. This Rogue happens to have known the noble previously, dresses to impress, and sweetens the deal by giving them a gift that past experience has taught that the noble likes. The DM decides this is worth a +5 to their diplomacy check.

But say it was a brutish Barbarian dressed in animal hides, scuffed and dirty from travel with the foul odor of not having bathed in weeks clinging to him. The Barbarian angrily demands an invitation. The DM decides this is a -5 to the Check.

Putting that together, there's a +10 difference between the Circumstance Modifiers between the two against the same DC. In 5e, the DM is instructed to adjust the DC's instead, so if the base DC would have been 15, the Rogue gets a DC of 10 (and possibly Advantage depending on how well the Rogue read the situation) and the Barbarian gets a DC of 20. The Circumstance Modifiers just adjust the DC instead of applying to the roll.

I have picked some comments above which I think convey a large part of the conversation, in order to contextualize my own thoughts.

I agree that arbitrarily changing the DC based on who's rolling is...not the best way to go about things. If you're going to do that, don't roll. Just give it to the guy you think should know it. Or be able to do it. In the 5e game I started playing in a few weeks ago, nearly every PC has minor illusion, because so many ways exist to pick up cantrips and it's such a useful one. The DM did, however, determine that the character who is an Illusionist and has History trained was better able to replicate an illusion of a cool ancient coin we'd seen than the character who just happened to know minor illusion from his training in ninjitsu (a shadow monk). This was similar to the notion of the barbarian who doesn't read not being able to succeed on the Arcana check even though the wizard can...except the DM wasn't pretending there was chance involved; he just assigned success values as he felt appropriate given character backgrounds and foci in this case.

It is probably worth noting, however, that I think you can make a distinction between "proficient-only" checks and ones that anybody can make. If your barbarian really hasn't studied Arcana, that's probably represented by not even being proficient. So you can set up boundaries without making it "Bob can't, but George can." "It's DC 17 if you're proficient in Arcana" works, because it reflects an actual mechanical difference. But if it's something you want to do based solely on backgrounds and fluff, you should probably just give it to the one you think should succeed. It's silly to set up different DCs specifically for each PC.

Circumstance bonuses, too, divorce it a touch from "Bob can, George can't." The +10 bonus the rogue got in the above example is due to a number of important things, some of which could have affected the rogue had he been acting differently, and some of which could have improved for the barbarian if he had taken different efforts. Cleaning up and dressing nicely in the style of the court, for instance, would have removed -2 or more of the circumstance penalties.

The only thing that stands out as "rogue's background" in particular is "the rogue knows the noble from his past." And that's where circumstance bonuses do apply. But even then, it's not an altered DC; it's a shift to the roll's value. Mathematically, this may not matter, but conceptually, it does. It colors how the judgment is made now and in the future. And it isn't arbitrary; it's +/- 2 per applicable circumstance. Not a hand-wavey "I really don't want the Barbarian succeeding, so I'll make the DC ridiculous for him, but still pretend it's the dice that decide."

In short: if the DM is going to arbitrarily decide who succeeds at something, that's fine, but the fig leaf of the roll is kind-of silly at that point. Everything has its place; rolls for success/failure should happen only when you haven't already decided which PC(s) should succeed/fail.

[Cazero]

Neat! So if they're trying to climb up a difficult slope and the DC is 17 for Krug, who has his +7 strength and +6 proficiency and DC 30 for the wizard, who has no real experience at this because in normal circumstances he'd just use magic?

Because that's not how it works. The DC is the same for the both of them, Krug doesn't get an easier DC because he's a barbarian - because he's a barbarian, he has a greater modifier to his roll. You don't double dip on this, if you want the DC to be 13 easier for a class, give them +13 to that skill.
If you really want to make a proper parralel between a mountain cliff and an obscure rune, you need to pick an illiterate barbarian and a quadraplegic wizard. Not just a wizard.
And as I already said...

The obvious problem being that the skills should be enough to tell the difference.
But it can't. Because a d20 is absurdly swingy for anything vaguely reliable.

[ravenkith]

In retrospect,


It seems to me that the answer lies somewhere in the reliable talent class ability of the rogue.
Feat: Journeyman (Skill)
Requirements - Level 4, Proficiency in skill to be chosen

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to a 10 if the actual number shown on the die is less than 10.

Feat: Master (Skill)
Requirements - Level 8, Proficiency in skill to be chosen and either Journeyman feat for skill to be chosen -OR- Reliable Talent class feature

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to a 15 if the actual number shown on the die is less than 15.



The above seem a little OP.

With changes:

Feat: Journeyman (Skill)
Requirements - Level 8, Proficiency in skill to be chosen

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to an 8 if the actual number shown on the die is less than 8.

Feat: Master (Skill)
Requirements - Level 12, Proficiency in skill to be chosen and either Journeyman feat for skill to be chosen -OR- Reliable Talent class feature

When you take this feat, choose one skill that you are proficient in. Whenever you roll a skill check for that skill, any number rolled on the d20 used for the check gets raised to a 12 if the actual number shown on the die is less than 12.

This means that, at level 8, upon selection of the first feat, a Character with an 18 in the relevant score (+4 bonus), will have an automatic floor for their chosen skill of 8 (Roll minimum) + 3 (Proficiency) + 4 (modifier), for a total of 15 - This means they would succeed at most normal tasks involving the check.

At level 12, the same character would have a floor of 12+4+4, for a total of 20.

Of course a character with expertise in the chosen skill would get double the proficiency on each of these, meaning their minimum becomes 18 as a Journeyman and 24 as a Master.

At level 20, a Journeyman character would have either 8+6+4 (18) or 8+12+4 (24), depending on expertise, as their floor, while a Master character would have 12+6+4 (22) or 12+12+4 (28) as his floor.

These numbers seem more balanced, I think.

[MaxWilson]

This particular DM made skills the most offensive, I think (he made the gnome make two separate checks to get on and off a barstool :smallfurious:), but the issue is general. 25% just isn't enough, I don't think.

I've had the same experience at some tables. One DM made my Acrobatics-proficient Dex 16 PC roll a DC 10(?) check when I declared "climb a tree" as my action in order to stay away from some wolves that were chasing an (apparently) little boy. I failed it, two rounds in a row. That's partly poor luck with dice, but it's also an implausible outcome to demonstrates that the DC (IMO) was poorly-chosen in the first place. At my table I'd either not roll at all, or assign a DC low enough to be an auto-success for anyone trained, e.g. DC 4 or 5.

The OP talks about how 5E works well for DMs who are good but doesn't give much guidance for those who are okay. That fits my assessment. 5E's skill system, what there is of it, works pretty well for any DM who understands probability intuitively and knows how to choose DCs/etc. appropriately... which isn't much of a recommendation because it basically means "5E's skill system works for any DM who is willing to invent their own skill system." There's a space there for some good OGL/DMsGuild projects like a Complete Book of Mundane Skill Checks, with DCs and circumstantial modifiers, written by someone who does understand probability, as well as diverse fields from camouflage to cooking to athletics so that he can choose the right modifiers.

Skills is one of the worst aspects of 5E design from a simulationist viewpoint. Either you house-rule it extensively, or you just accept that 5E is better at combat than exploration/social skill use and go from there (which might mean not choosing 5E for certain campaigns). Given my temperament I lean towards "accept it temporarily until you can create good rules."

[Segev]

To be fair, "Climb a tree while being chased by wolves" is actually pretty stressful and prone to small errors magnifying. I can see even a trained tree-climber panicking and thus not gripping right, or choosing too weak a limb to support himself because he's not taking time to test before moving. It may be a small chance, but since time is of the essence, mistakes that would normally be compensated for by a little extra time and not even be noticed will instead fail miserably.

I've managed to drop my keys while trying to unlock my back door twice in a row, and the only thing I had complicating it was a water bottle in my other hand. I'm not particularly clumsy, compared to the average person. And that's the kind of thing one normally wouldn't roll at all. If only because those rare moments when you drop your keys even when calm and collected don't really do much more than add another 3 seconds to the time it takes to roll your eyes at yourself, stoop down, pick them up, and actually unlock the darned door.

If I'd been being chased by a vampire or something and getting inside my threshold was safety? You can bet the roll would be appropriate! It matters how long it takes me to actually get the stupid thing unlocked!

[Regulas]

As was pointed out in the other thread. In 5e you are not supposed to be rolling for every single possible check or action, rather you should be rolling when there is a possibility of either success or failure. And depending on the character even a success or failure can have variable results. As in if a wizard fails his arcana roll he won't know exactly what spell but he might still recognise that it's necromancy.

Furthermore the barbarian who knows nothing about magic... knows nothing about magic. Even if you for some reason let him make the role it would only tell him "Yes this is magic", though realistically your DM probably shouldn't be letting him make a roll unless there is a reason he would possibly know something (such as if he saw the spell before).

[Knaight]

The downside to curved RNGs is that they make extraordinary success less likely.

So if, for example, your Fighter gets grabbed by a Giant, your prospects become really, really bleak. Or, even more likely, it gets even rockier when an enemy spellcaster is targeting one of your weak saves.

This isn't to say it's a bad idea, at all. It's just often overlooked when people are considering 2d10 or 3d6.
The trimming of extraordinary success is part of the point of curved RNGs.


Well the DM shouldn't be telling the players necessarily that you failed because your a Barbarian, however on the flip side I don't think there should be a need to tell the Barbarian every time that you don't know anything on the subject, once really should be enough. But yes it should be pretty obvious without needing more then a second thought for the DM to know that a Barbarian can't tell what spell that is no matter what he roles.

And I can understand it would take some getting used to, 3.5 edition pretty much is the "rules for everything" whereas 5th edition is explicitly aiming to simply be "general framework", with a stronger emphasis on roleplaying and decision making.


I guess the difference is that to me the notion that the barbarian who hasn't by background studied arcana can't succeed at arcana is so obvious that it doesn't even take a thought. In my eyes it's not a rule because it's so obvious and self evident as to be an axiom, and the rule was excluded because it would just be a waste of space on the page.


Or to put it another way, the designers wanted the rules to be flexible enough for the DM to decide when a check was called for and what success & failure looked like. As opposed to trying to hard-code every possible DC for all characters.
Two concepts are being conflated here. The transition to general framework isn't the bizarre skill override system people are proposing, or varying DC by class, or anything like that. It's the use of a simple Easy-Medium-Hard ladder instead of every skill having a DC list with specific examples under it.

Varying DC by class or only allowing certain characters to roll by class regardless of actual skill bonuses is just a weird system of hidden bonuses and penalties stacked on top of the ones on the character sheet. It's a clumsy patch that adds needless complexity because the base rules aren't properly calibrated, thus producing oddities.


Why would you tune anything to party composition? The world is the world, I'm not going to have its contents magically change just because they brought a rogue and druid instead of a barbarian and a monk. Different classes are good at different things, it makes those differences meaningless if you're just going to adjust the world around them and adjusting the world around anything destroys the feeling that it's a living breathing place, turns it into a video game centered around the PCs.

[Kurald Galain]

As was pointed out in the other thread. In 5e you are not supposed to be rolling for every single possible check or action,
Nor are you in any other RPG printed, ever, so that's not much of an argument.

[Regulas]

Nor are you in any other RPG printed, ever, so that's not much of an argument.

The only way that any of the problems that people are complaining about are relevant is if you ARE rolling for checks and actions you shouldn't be.

For example people are complaining their untrained people are nearly as good as there trained people, except those untrained people shouldn't be allowed to roll, or if they are there results shouldn't be the same.

[Grod_The_Giant]

The only way that any of the problems that people are complaining about are relevant is if you ARE rolling for checks and actions you shouldn't be.

For example people are complaining their untrained people are nearly as good as there trained people, except those untrained people shouldn't be allowed to roll, or if they are there results shouldn't be the same.
Except that there's nothing in the books telling you that's what you're supposed to be doing. In fact, not doing that is one of the biggest points of Bounded Accuracy-- that anyone can try. The one objective measure of "what you should be able to do"--your modifier--pretty distinctly says that untrained is nearly as good as trained.

[Kurald Galain]

For example people are complaining their untrained people are nearly as good as there trained people, except those untrained people shouldn't be allowed to roll, or if they are there results shouldn't be the same.

So what you're saying is that when the party comes by a river, only characters trained in athletics should be allowed to roll swim checks, and the others automatically fail?

[Regulas]

So what you're saying is that when the party comes by a river, only characters trained in athletics should be allowed to roll swim checks, and the others automatically fail?

Now you are just being an ass taking an obvious twist of the situation to portray what I said differently. However if I want to be pedantic I could point out that anyone who hasn't been taught how to swim actually likely wouldn't be able to in real life and would drown.

Anyway a skill check should be made when there is a reasonable chance of success/failure. People trying to climb a cliff even untrained has a reasonable chance to succeed. Someone trying to determine the precise properties of a spell who has never heard of magic before is not reasonable, there is no reasonable chance they would succeed so they shouldn't be making a roll, unless you as a DM want to give them a fake result or something else that still equates to guaranteed failure.


Except that there's nothing in the books telling you that's what you're supposed to be doing. In fact, not doing that is one of the biggest points of Bounded Accuracy-- that anyone can try. The one objective measure of "what you should be able to do"--your modifier--pretty distinctly says that untrained is nearly as good as trained.

At this point I'm wondering if you're just a munchkin type who always tries to bend any rules into your favour in defiance of common sense?

[DanyBallon]

Except that there's nothing in the books telling you that's what you're supposed to be doing. In fact, not doing that is one of the biggest points of Bounded Accuracy-- that anyone can try. The one objective measure of "what you should be able to do"--your modifier--pretty distinctly says that untrained is nearly as good as trained.

You don't need an actual rule to say that it's up to DM fiat, because 5e rely more on DM fiat and rule of fun instead of hardcoded rules. "Ask your DM" is all over the PHB, no need to write it down for every single aspect of the game, you get the feel when you read it as a whole.



So what you're saying is that when the party comes by a river, only characters trained in athletics should be allowed to roll swim checks, and the others automatically fail?

You kinda missed the last part, that if they're allowed, they might not have the smae result, so in you example, maybe the elf ranger would cross the river quite easily, but the wizard, will manage to pull it out (if he succeed its check) but may take longer and/or be push further down the stream before reaching the other side.

[Kurald Galain]

You kinda missed the last part, that if they're allowed, they might not have the smae result, so in you example, maybe the elf ranger would cross the river quite easily, but the wizard, will manage to pull it out (if he succeed its check) but may take longer and/or be push further down the stream before reaching the other side.
The problem is that now, your character's swimming prowess depends not on his athletics modifier, but on whether your DM's headcanon of elves matches your character concept. There's a big difference between a Tolkien elf and a Pini elf, after all.

The more elegant approach would be to have a rule that (e.g.) "elves get +2 on athletics checks". At least that way it's (1) consistent and (2) known when you're building your character.

[Grod_The_Giant]

You don't need an actual rule to say that it's up to DM fiat, because 5e rely more on DM fiat and rule of fun instead of hardcoded rules. "Ask your DM" is all over the PHB, no need to write it down for every single aspect of the game, you get the feel when you read it as a whole.
There doesn't need to be a rule that says "the GM can adjust things to taste," no, but if the expectation is that "only character proficient in X skill can attempt Y checks," the rules should say so. They did in 3.5 with trained skills. That was fine. I have yet to see anything of the sort in 5e. If the expectation isn't "everyone can attempt any check and your abilities on your character sheet say you have a decent shot at making that Medium check" I'd like to be told. I'd like new DMs to not have to figure that out.


At this point I'm wondering if you're just a munchkin type who always tries to bend any rules into your favour in defiance of common sense?
Don't be an ass. I'm trying to make sense of the rules, and--if they remain dysfunctional--find a solution. Preferably one that doesn't rely on the DM deciding what the result should be and then dictating the rolls to match that. Gating checks behind proficiency is, I suppose, one way of getting around the "barbarian knows more than the wizard" issue, though it doesn't do anything about the "wizard has a good chance of failing basic checks" issue.

[DanyBallon]

The problem is that now, your character's swimming prowess depends not on his athletics modifier, but on whether your DM's headcanon of elves matches your character concept. There's a big difference between a Tolkien elf and a Pini elf, after all.

The more elegant approach would be to have a rule that (e.g.) "elves get +2 on athletics checks". At least that way it's (1) consistent and (2) known when you're building your character.

D&D have always been about the relation between the players and the DM, some editions gave more hardcoded rules, some less, but every editions need you to talk to your DM before hand. If you get in a game assuming that elves are Tolkien elves, while your DM think differently and never bother to ask, then you have a serious problem, but it's a communication problem not a problem with the rules. In the case of Adventure League, there's a guide and a setting that gives both player's, and DM alike, enough background so that when they talk about elves, they picture the same type of creature.

And again, it's all about the interaction between you and your DM. When facing the same river, one can tell its DM that his character is an expert swimmer (hence the +6 Athletic), the DM could agree (the river is not wide, shallow, calm, etc.) and let the character swim across the river without any roll, or knowing that they might be consequence for failure (strong current, underwater hazard or creatures...), the DM ask for a roll. The +6 to Athletic will probably let the character cross the river without any trouble, but rolling reflects the chance of failure.

[DanyBallon]

There doesn't need to be a rule that says "the GM can adjust things to taste," no, but if the expectation is that "only character proficient in X skill can attempt Y checks," the rules should say so. They did in 3.5 with trained skills. That was fine. I have yet to see anything of the sort in 5e. If the expectation isn't "everyone can attempt any check and your abilities on your character sheet say you have a decent shot at making that Medium check" I'd like to be told. I'd like new DMs to not have to figure that out.


Never said that a character can never try an untrained skill check, what I said, is that if there is no good reason for someone to make an untrained skill check, then the DM should tell the character just can't and why. The rule are quite simple, skill checks (when to roll and DC) are up to the DM, and use "is there a chance for failure or for success?" and "does it make sense, does it push the story forward, and/or is it fun?" as guideline.

[Kurald Galain]

And again, it's all about the interaction between you and your DM. When facing the same river, one can tell its DM that his character is an expert swimmer (hence the +6 Athletic),
But wait: by the rules, your character doesn't have a +6 to athletic. For most of the character's career, it will be +2 or +3. That's the premise of this very thread here, that the difference between a skilled and an unskilled character is too small.

What if a player claims his character is an expert swimmer even though he's not trained in athletics? What if the DM claims that another PC is bad at swimming because dwarves can't swim, even though he is trained in athletics? What if the DM decides that since he himself cannot swim, clearly crossing a river is difficult, or DC 20?

A better-written skill system would answer all those questions in a heartbeat.

[Regulas]

Gating checks behind proficiency is, I suppose, one way of getting around the "barbarian knows more than the wizard" issue, though it doesn't do anything about the "wizard has a good chance of failing basic checks" issue.

The point is that it's not supposed to be a hard fixed gate. In 3.5 the game was heavily automated, but this version isn't, the DM has to think about his decisions and not just rely on the rulebook far moreso, much like how most other RPG systems work.

If it's basic enough should the wizard even need to roll for it? Furthermore the DC may prevent him from remembering precise details but he could still probably tell basics. Essentially it's up to the DM to figure out what a Wizard in his world might reasonably know by default and what he might need to make a roll for and what he would learn depending on the roll, cause even if your wizard fails his check, that doesn't mean he's going to confuse that fireball spell for a squirrel he'll just not know what exact spell you are casting.


But wait: by the rules, your character doesn't have a +6 to athletic. For most of the character's career, it will be +2 or +3. That's the premise of this very thread here, that the difference between a skilled and an unskilled character is too small.

What if a player claims his character is an expert swimmer even though he's not trained in athletics? What if the DM claims that another PC is bad at swimming because dwarves can't swim, even though he is trained in athletics? What if the DM decides that since he himself cannot swim, clearly crossing a river is difficult, or DC 20?

A better-written skill system would answer all those questions in a heartbeat.

But simply put that is explicitly the DM's job to figure out.

[Blue Lantern]

Don't be an ass. I'm trying to make sense of the rules, and--if they remain dysfunctional--find a solution. Preferably one that doesn't rely on the DM deciding what the result should be and then dictating the rolls to match that. Gating checks behind proficiency is, I suppose, one way of getting around the "barbarian knows more than the wizard" issue, though it doesn't do anything about the "wizard has a good chance of failing basic checks" issue.

Just because you don't like how a system work does not make it dysfunctional or wrong.

5e was designed in a specific way to make the roll be important at all level and remove the 3.5 like possibility of a character being able completely remove the dice as a factor and to make skill challenges possible for everyone; if you don't like that there are always other system to look for that works in the way you want.

Alternatively I think making some skill roll only available for proficient, like that arcana issue that seems to bother you so much, is an easy house rule to add.

[Kurald Galain]

But simply put that is explicitly the DM's job to figure out.
Indeed. The point is that if the system did that for him, then it would make the DM's job easier, and would give him more time to focus on other things.

Since pretty much every commercial RPG does precisely that, it strikes me as a fair criticism of 5E that it does not, and a fair request to ask the forum how to fix that.


5e was designed in a specific way to make the roll be important at all level and remove the 3.5 like possibility of a character being able completely remove the dice as a factor
What you're missing is that there's a huge gap between "dice aren't a factor" and "proficiency makes only a marginal difference".

[georgie_leech]

Just because you don't like how a system work does not make it dysfunctional or wrong.

By the same token, not liking the skill rules as presented doesn't make someone a munchkin.

[Blue Lantern]

What you're missing is that there's a huge gap between "dice aren't a factor" and "proficiency makes only a marginal difference".

I am not missing anything, I am just not bothered by it, I don't consider a 10-30% increase of success (doubled for expertise) to be so marginal as you do.


By the same token, not liking the skill rules as presented doesn't make someone a munchkin.

Care to point out where and when I have called someone a munchkin exactly?

It is fair not liking something and it is fair to ask for suggestion on how house rule it (which is different from fixing it), and I just give my opinion on that, but criticising the system and all the players that like it for how it is I think is bad manners.

[DanyBallon]

But wait: by the rules, your character doesn't have a +6 to athletic. For most of the character's career, it will be +2 or +3. That's the premise of this very thread here, that the difference between a skilled and an unskilled character is too small.

My 5th level EK would disagree with you :smalltongue: It's quite easy for a 1st level character to have +5 in a proficient skill that is related to it's main ability score.


What if a player claims his character is an expert swimmer even though he's not trained in athletics? What if the DM claims that another PC is bad at swimming because dwarves can't swim, even though he is trained in athletics? What if the DM decides that since he himself cannot swim, clearly crossing a river is difficult, or DC 20?

A better-written skill system would answer all those questions in a heartbeat.

If such a player tell me that he's an expert swimmer and don't have anything to back it up, then tough luck. If there is a good reason for him to be a good swimmer, but not to take the athletic skill, I'll give him advantage on his roll. Your dwarf example is again a problem of communication between the players and the DM. And as for your last scenario, it's still a communication problem, but at least you'll now know that in this campaign, swimming is a hard task for everyone.

[Kurald Galain]

My 5th level EK would disagree with you :smalltongue: It's quite easy for a 1st level character to have +5 in a proficient skill that is related to it's main ability score.
It's also quite easy for a 1st level character to have a +5 in a non-proficient skill. Once again, the point is that proficiency makes too little of a difference.


If there is a good reason for him to be a good swimmer, but not to take the athletic skill, I'll give him advantage on his roll.
And that's precisely the issue.

[RickAllison]

It's also quite easy for a 1st level character to have a +5 in a non-proficient skill. Once again, the point is that proficiency makes too little of a difference.


And that's precisely the issue.

Correction: It is possible to get a +5 in a non-proficient skill at level 1. That requires a character to roll an 18 (certainly no guarantee) and also have a race give +2 to that ability. Far from easy, and actually impossible if you are using point-buy. At least the +5 in a proficient skill was reasonable, this theory requires someone to roll stats and be at the peak of humanoid strength.

[DanyBallon]

It's also quite easy for a 1st level character to have a +5 in a non-proficient skill. Once again, the point is that proficiency makes too little of a difference.


And that's precisely the issue.

Seems to me that you don't like 5e way of giving more power into the hands of the DM, from there, you can either adapt 3.x skills to 5e, or revert back playing 3.x, because, I fear that 5e will never satisfy you as presented in the core rule books. And sincerly, I don't want 5e to revert to hardcoded rules for everything.

[newsman77]

I'm trying to make sense of the rules, and--if they remain dysfunctional--find a solution. Preferably one that doesn't rely on the DM deciding what the result should be and then dictating the rolls to match that.

Just a suggestion, but perhaps play 5e at certain levels (1, 5, 10, 15, 20) to see if what you're saying holds up. As you said, you only played a level 1 character. Levels 1 to 3 you're nothing more than a super-commoner who has the spark to take up adventuring. You character may not know much... but you'll get there eventually. I think you may be theory crafting in your head and it doesn't necessarily translate to paper.

Remember, you'll have your proficiency bonus that goes up with level and your stat ability that will add to your ability checks. Level 1 is a bad example to base your decision on whether or not the system is dysfunctional.

[Kurald Galain]

I don't want 5e to revert to hardcoded rules for everything.

And nobody is suggesting that.

[Regulas]

Since pretty much every commercial RPG does precisely that, it strikes me as a fair criticism of 5E that it does not, and a fair request to ask the forum how to fix that.


I would argue on the contrary most RPG's don't do that at all. 3rd in particular has always had far far far more rules then any other RPG I've ever seen trying to create rules for every single last possibility, with other systems tending between 5e level all the way to many popular ones being little more then vague guides.

[pwykersotz]

I love the fact that the skill system is open and that it encourages players who want a bit more complexity to add in a few things. The books even mention all these solutions at one point or another, allowing only the proficient to roll, granting advantage/disadvantage...I love that it's so easily hackable and tweakable.

What I don't like as a GM or as a player is when the skill system is not these things. For example, only once in about 5 years played with a GM who used the skill system as written. But the skill stuff was so hard-coded that this changed a LOT of intangibles and expectations, and created far more difficulty than a system which does not set that baseline. I love that I can talk to one of my friends and he can say "I'm going to run the skill system with laymen's physical limits, so a DC 30 would be moving something that Captain America might move" and another DM can say "30 is nearly impossible, that would be pushing open the Giant's castle doors that weigh ten tons", and another can say "DC 30 is for demigods, you're moving concepts such as grabbing ahold of the wind and dragging it in a different direction.

In a hard coded system, this causes havoc. I know (allegorical, I know) because I played in all those ways with 3.5. In 5e it's a 2 minute conversation to set the expectations between easy, hard, and impossible.

[Lines]

As was pointed out in the other thread. In 5e you are not supposed to be rolling for every single possible check or action, rather you should be rolling when there is a possibility of either success or failure. And depending on the character even a success or failure can have variable results. As in if a wizard fails his arcana roll he won't know exactly what spell but he might still recognise that it's necromancy.

Furthermore the barbarian who knows nothing about magic... knows nothing about magic. Even if you for some reason let him make the role it would only tell him "Yes this is magic", though realistically your DM probably shouldn't be letting him make a roll unless there is a reason he would possibly know something (such as if he saw the spell before).

But he can make the roll. It's within his skillset, as evidenced by the fact that the roll is within the capabilities of his total skill modifier for arcana, so he should be allowed to try.

[Kurald Galain]

But he can make the roll. It's within his skillset, as evidenced by the fact that the roll is within the capabilities of his total skill modifier for arcana, so he should be allowed to try.

It's actually interesting how differently people respond to the notion that bonuses are too small.

Some say, well obviously all characters are intended to have more-or-less the same chance of success at everything, so there's no problem.
Others say, well obviously characters that are supposed to be bad at something aren't be allowed to roll for it, so there's no problem.
And yet others claim that this part of gameplay is not DM-dependent :smallbiggrin:

I suppose the main conclusion we can draw here is that the rulebooks provide insufficient guidance.

[Knaight]

The point is that it's not supposed to be a hard fixed gate. In 3.5 the game was heavily automated, but this version isn't, the DM has to think about his decisions and not just rely on the rulebook far moreso, much like how most other RPG systems work.

It's not just a matter of the DM having to think about their decisions, it's a matter of the math of the system producing an incredibly tight band in which you either accept that outside of combat situations and a few edge cases characters aren't significantly different in ability, or you produce weird system alterations with hidden bonuses and penalties. This isn't a matter of hard coded rules; that's where the list of DCs for every skill as opposed to the ladder come in (a change that I suspect is largely supported by the people criticizing the bonus size).

Let me use an example here, from a non D&D 3.5 system. I mostly GM Fudge. It's not automated, the GM has to use their judgement, so on and so forth. However, when it comes time for people to roll a skill for an identical task, I can just have them both roll the skill. Between the curvature of the dice probabilities and the use of a sufficiently large range, characters who are bad at tasks actually coming across as bad at tasks when they roll for them happens, as does characters being good at them. Every advantage about being able to calibrate what constitutes easy vs. hard is still there, it's just that I don't have to fight the system to get character variation. When it comes to recognizing the magic, the barbarian who knows nothing about magic* would have a terrible skill, and thus for an average task have a 5/81 chance of succeeding. Meanwhile the great wizard has a 5/81 chance of failing. The dice still matter, but the skill of the characters shines through. If it is instead some sort of really complex thing, the barbarian might have no chance at all, while even the wizard only has a 31/81 chance of succeeding.

*Which isn't necessarily part of a core barbarian concept, given that tropes about barbarian shamans are absolutely a thing, and which would thus be irritating if it wasn't a player decision to have a bad magic skill but instead a secret penalty they didn't know about due to picking class.

[Regulas]

It's actually interesting how differently people respond to the notion that bonuses are too small.

Some say, well obviously all characters are intended to have more-or-less the same chance of success at everything, so there's no problem.
Others say, well obviously characters that are supposed to be bad at something aren't be allowed to roll for it, so there's no problem.
And yet others claim that this part of gameplay is not DM-dependent :smallbiggrin:

I suppose the main conclusion we can draw here is that the rulebooks provide insufficient guidance.

Mostly I just see 3.5 as having been so rules intense that a lot of it's core players have become to used to having every last thing defined by a rule. And I think part of the problem is the difficulty in finding a good DM, one of 3.5 strong suits was that, especially if you used a pre-made adventure, you could have a pretty ****ty DM and still make it work.

Another trait I've noticed is that players who like having tons of rules also tend to like applying things as absolutes even when they shouldn't be.



But he can make the roll. It's within his skillset, as evidenced by the fact that the roll is within the capabilities of his total skill modifier for arcana, so he should be allowed to try.


This is a great example, as this is basically rules lawyering. To me it's like arguing that you can swim up a waterfall, or heck climb air. There's nothing in the rules that says you can't after all! Social interactions don't consume an action, so howabout in combat I spend infinite rolls trying to convince them to join our side, eventually I'll roll enough 20's!. There's nothing in the rules that says I can't, common sense sure does but the rules?

[Knaight]

Mostly I just see 3.5 as having been so rules intense that a lot of it's core players have become to used to having every last thing defined by a rule. And I think part of the problem is the difficulty in finding a good DM, one of 3.5 strong suits was that, especially if you used a pre-made adventure, you could have a pretty ****ty DM and still make it work.

Another trait I've noticed is that players who like having tons of rules also tend to like applying things as absolutes even when they shouldn't be.
:smallsigh:
This isn't about having every last thing defined by a rule. People are perfectly happy with a general rule of roll a d20, add your skill, compare it to a DC set by the DM from this convenient table. It's the specific numbers chosen that people aren't liking. The argument about having every last thing defined by a rule is a completely separate one in completely separate threads, and just about everyone in this thread you're trying to paint as a 3.5 partisan is on the 5e side of that particular argument.



This is a great example, as this is basically rules lawyering. To me it's like arguing that you can swim up a waterfall, or heck climb air. There's nothing in the rules that says you can't after all! Social interactions don't consume an action, so howabout in combat I spend infinite rolls trying to convince them to join our side, eventually I'll roll enough 20's!. There's nothing in the rules that says I can't, common sense sure does but the rules?
It's not even remotely comparable. There's no support for the idea that this particular barbarian knows nothing about magic - odds are good they work with a mage of some sort, they're almost guaranteed to have been on the receiving end of spells, they have some background. On top of that, this is a part of the system that is defined, and they have a skill number there which interacts with the DCs in known ways. There's every indication that they have a chance at a DC 10 or 15 check (or a 20, if they have proficiency or a positive modifier in the relevant attribute). This is just being yanked out, to compensate for how 5e doesn't distinguish between characters well in the skill system.

Meanwhile, the swimming up waterfalls and climbing air situations are also nicely handled by the general principles. If a task is impossible you don't get to roll, problem solved. Or in the case of swimming up a waterfall (which might be possible in more cinematic campaigns), you handle it with the structure in place.

[Xetheral]

The only way that any of the problems that people are complaining about are relevant is if you ARE rolling for checks and actions you shouldn't be.

For example people are complaining their untrained people are nearly as good as there trained people, except those untrained people shouldn't be allowed to roll, or if they are there results shouldn't be the same.

I'm going to join the chorus stressing strenuous objection. Forbidding untrained people to roll and/or interpreting the same die results differently are both very much additions to the rules in the PHB that together are merely one potential solution to the problems being discussed here, and come with costs that many players and DMs are unwilling to pay.

That the 5e skill system as written fails to meet the needs of a variety of posters whose tables want to play 5e is a weakness of the system. That you've found a solution to this problem that works for your table doesn't mean it isn't a problem for others whose play style makes your solution unpalatable. Suggesting that everyone else "shouldn't" have a problem if only they'd adopt your solution doesn't help find a fix for those posters, nor does it in any way undermine their statement that 5e fails to meet their needs in this regard.

[Regulas]

snip

The waterfall/air example is just showing how it looks to me when someone tries to argue something that goes against common sense, it's as if you are arguing that you can climb air.


For the rules, let me put it this way, I see making a judgement call as to what someone finds out from an Arcana check based on there background and knowledge a fairly quick simple and easy decision that a DM could make in seconds on the fly in any situation. At least in my own case if I were DMing a campaign, I would almost universally give my arcana proficient characters relatively more information then I would for non proficient regardless of the DC's or rolls. This isn't something that I have to think about in depth. Both the wizard and barbarian make there rolls? Example (terrible spell choice but still):
Barb: You know that it's a fire spell, and you've mostly seen similar looking magic used to blast large areas.
Wizard: You know that it is specifically the fireball spell, and you'd judge it's probably being cast as a 4th level spell.

This doesn't even take serious thought, it's barely more then an afterthought, but it accounts for there difference in knowledge on magic.


snip.

In my eyes I feel like all I'm actually advocating is people use common sense.

[Tanarii]

That the 5e skill system as written fails to meet the needs of a variety of posters whose tables want to play 5e is a weakness of the systemThe so-called "failures" of 5e is a system built in reaction to the failure and commonly complained about problems in 3.5e/pathfinder.

All the things that are being complained about were SPECIFICALLY done to move away from the things the posters in this thread want. Because the things these posters want created too many problems for the majority of players.

Things wanted that existed in 3.5e/pathfinder that 5e intentionally moved away from, because they were considered failures in the skill system by the majority of players:
1) Large differences in success chance between "trained" and "untrained" for PCs
2) Total bonuses outweighing the d20 randomness (ie being close to or more than +20)
3) Fixed and defined DCs for specific tasks

The 5e skill system isn't a failure. It's a reaction to the failures of previous editions, for the majority of players. If it doesn't meet your needs, it's house-rule time. Which is awesome. Do it. Some good suggestions were made in this thread already. Including having differing levels of what is "Easy" or doesn't require a check for any given PC, based on the character. (ie the Barbarian/Wizard Arcana discussion.)

[georgie_leech]

The so-called "failures" of 5e is a system built in reaction to the failure and commonly complained about problems in 3.5e/pathfinder.

All the things that are being complained about were SPECIFICALLY done to move away from the things the posters in this thread want. Because the things these posters want created too many problems for the majority of players.

Things wanted that existed in 3.5e/pathfinder that 5e intentionally moved away from, because they were considered failures in the skill system:
1) Large differences in success chance between "trained" and "untrained" for PCs
2) Total bonuses outweighing the d20 randomness (ie being close to or more than +20)
3) Fixed and defined DCs for specific tasks

Quibble, moving away from something that causes problems or towards something that a vocal portion of the audience wants doesn't mean there won't be other problems. See also: the rather nasty Edition Wars after 4e came out.

[Tanarii]

a vocal portion of the audienceWhich is why we have 5e skill system instead of 3.5's. It's based on huge amounts of playtesting and surveys. If the 'vocal portion' had it's way, we'd still have 3.5 semi-simulation skill system going on.

[Xetheral]

The so-called "failures" of 5e is a system built in reaction to the failure and commonly complained about problems in 3.5e/pathfinder.

All the things that are being complained about were SPECIFICALLY done to move away from the things the posters in this thread want. Because the things these posters want created too many problems for the majority of players.

Things wanted that existed in 3.5e/pathfinder that 5e intentionally moved away from, because they were considered failures in the skill system by the majority of players:
1) Large differences in success chance between "trained" and "untrained" for PCs
2) Total bonuses outweighing the d20 randomness (ie being close to or more than +20)
3) Fixed and defined DCs for specific tasks

The 5e skill system isn't a failure. It's a reaction to the failures of previous editions, for the majority of players. If it doesn't meet your needs, it's house-rule time. Which is awesome. Do it. Some good suggestions were made in this thread already. Including having differing levels of what is "Easy" or doesn't require a check for any given PC, based on the character. (ie the Barbarian/Wizard Arcana discussion.)

I never said the 5e skill system was a failure; I said it failed to meet the needs of many of the posters in this thread, and I said that was a weakness of the system. I stand by my statement. Any element of any system can be a weakness if it fails to meet the needs of people who want to play the system. That doesn't preclude the possibility that such an element can be seen as a strength by others (although I've seen nothing to suggest that either view on this element commands a majority). And I completely agree that when a system doesn't meet your needs it's house-rule time, hence my participation in this thread devoted to discussing the weakness and how to work around it.

[pwykersotz]

I never said the 5e skill system was a failure; I said it failed to meet the needs of many of the posters in this thread, and I said that was a weakness of the system. I stand by my statement. Any element of any system can be a weakness if it fails to meet the needs of people who want to play the system. That doesn't preclude the possibility that such an element can be seen as a strength by others (although I've seen nothing to suggest that either view on this element commands a majority). And I completely agree that when a system doesn't meet your needs it's house-rule time, hence my participation in this thread devoted to discussing the weakness and how to work around it.

I love the 5e skill system, and I can agree with this statement. Well reasoned.

[newsman77]

I never said the 5e skill system was a failure; I said it failed to meet the needs of many of the posters in this thread, and I said that was a weakness of the system. I stand by my statement. Any element of any system can be a weakness if it fails to meet the needs of people who want to play the system. That doesn't preclude the possibility that such an element can be seen as a strength by others (although I've seen nothing to suggest that either view on this element commands a majority). And I completely agree that when a system doesn't meet your needs it's house-rule time, hence my participation in this thread devoted to discussing the weakness and how to work around it.

Seconded. I love the skill system in 5e. I basically like 5e in general and don't see really any part as a failure. I see it as a response to the ridiculousness that was 3.5e. When you get +20 to +30 to a stat or skill check, well, why roll dice at that point? I call 5e a success!!

[DanyBallon]

I join my voice to those who like 5e skill system! I like its simplicity, its flexibility and that it push foward discussion between the DM and the players in order to build a better story. With this in mind, I think 5e skill system being vague and relying more on DM fiat is more a strenght than a weakness.

[ad_hoc]

Seconded. I love the skill system in 5e. I basically like 5e in general and don't see really any part as a failure. I see it as a response to the ridiculousness that was 3.5e. When you get +20 to +30 to a stat or skill check, well, why roll dice at that point? I call 5e a success!!

Yeah, I also love the 5e system. If a character is supposed to be an extra special blacksmith or whatever then they will have abilities that make them better at that.

PCs are not assumed to be the best blacksmiths so they don't have those abilities.

I am perfectly content to not only not have a +30 to my use rope skill but also to not be making skill checks whenever I do something. I just do the thing. If the outcome is in doubt I make a roll.

[Segev]

I know that in 3e/PF, there are some skills and uses of skills that are "trained only." No matter your bonus from stat or magic items, if you lack any ranks in the relevant skill, you can't roll it for "trained only" things.

I take it 5e does not have any such clause. No "can only roll this if proficient in the skill" things. I think that would be a good house rule to fix some of the problems that have been brought up here. Primarily the "Barbarian with no Arcane and Int 8 rolling a natural 18 and knowing more about the magical rune than the wizard with Int 20 and proficiency who rolls a natural 5."

That it is only some things that are "proficient only" means that you also don't run into "only people trained in athletics can even try to swim across the river."

[Kurald Galain]

I take it 5e does not have any such clause. No "can only roll this if proficient in the skill" things. I think that would be a good house rule to fix some of the problems that have been brought up here. Primarily the "Barbarian with no Arcane and Int 8 rolling a natural 18 and knowing more about the magical rune than the wizard with Int 20 and proficiency who rolls a natural 5."

That it is only some things that are "proficient only" means that you also don't run into "only people trained in athletics can even try to swim across the river."

I concur that this would be a good idea.

Something else to consider, for some but not all skills, is that in an opposed skill check, a character who's trained automatically beats a character who's not. So everybody can make a roll to swim across a river, but if there's a swimming race then the trained swimmer will always swim faster than the untrained.

[Shaofoo]

I know that in 3e/PF, there are some skills and uses of skills that are "trained only." No matter your bonus from stat or magic items, if you lack any ranks in the relevant skill, you can't roll it for "trained only" things.

I take it 5e does not have any such clause. No "can only roll this if proficient in the skill" things. I think that would be a good house rule to fix some of the problems that have been brought up here. Primarily the "Barbarian with no Arcane and Int 8 rolling a natural 18 and knowing more about the magical rune than the wizard with Int 20 and proficiency who rolls a natural 5."

That it is only some things that are "proficient only" means that you also don't run into "only people trained in athletics can even try to swim across the river."

I said this in another topic but I think part of the problem is that the knee jerk reaction to automatically reach for the d20 anytime a situation would arise regardless of situation.

If the barbarian shouldn't know what the rune was then I think the player should be able to reason with himself that his barbarian wouldn't have any reason to know so there shouldn't be a reason to chuck that d20 even if the DM would be all right with you doing it.

But you can also play the success in another way. Instead of the Barbarian saying that he knows whatever say the Barbarian gives enough clues to the Wizard so that he can figure it out, basically as an extended version of the Help action.

Of course asking players to be mindful of role play is not a good answer to the mechanical problem of chucking a d20 at it.

[DanyBallon]

I know that in 3e/PF, there are some skills and uses of skills that are "trained only." No matter your bonus from stat or magic items, if you lack any ranks in the relevant skill, you can't roll it for "trained only" things.

I take it 5e does not have any such clause. No "can only roll this if proficient in the skill" things. I think that would be a good house rule to fix some of the problems that have been brought up here. Primarily the "Barbarian with no Arcane and Int 8 rolling a natural 18 and knowing more about the magical rune than the wizard with Int 20 and proficiency who rolls a natural 5."

That it is only some things that are "proficient only" means that you also don't run into "only people trained in athletics can even try to swim across the river."

There's no need for that, you only need common sense. i.e. 2 barbarians with low int and not proficient in history, one from region A and one from region B (on the other side of the world). For some reason both face a challenge where theyh must name the previous chieftain of the elk tribe in Region A. Common sense would say that the barb from region A is allowed a roll, and the one from region B fail automatically because he just have no mean to have such knowledge.

A character from region B proficient in history or that had a interest to what happened in region A, might get to roll, while a character that is related to the late chieftain would have an auto success even if not proficient.

This is the kind of flexibility offered by 5e skill system.

[Segev]

I said this in another topic but I think part of the problem is that the knee jerk reaction to automatically reach for the d20 anytime a situation would arise regardless of situation.

If the barbarian shouldn't know what the rune was then I think the player should be able to reason with himself that his barbarian wouldn't have any reason to know so there shouldn't be a reason to chuck that d20 even if the DM would be all right with you doing it.

But you can also play the success in another way. Instead of the Barbarian saying that he knows whatever say the Barbarian gives enough clues to the Wizard so that he can figure it out, basically as an extended version of the Help action.

Of course asking players to be mindful of role play is not a good answer to the mechanical problem of chucking a d20 at it.
RP is great. I love it. I often make choices along the lines of, "My character would have no experience with this, so I am not going to roll." However, that is not the right choice to make in all situations, and at all tables. In fact, at many tables, doing that has led to the problem becoming unsolvable. Because my character mechanically could, in theory, have done it, and the others who tried had (by luck) failed. My choice not to try becomes mechanically disadvantageous, and actively can diminish the fun of everybody else at the table.

Now, obviously, if my character CANNOT do it rather than merely DOES NOT, the end result isn't much different, but at least then I'm not deliberately choosing to fail, OOC. At that point, the resources that could be brought to bear have been.

And, as a GM, it is often frustrating to know that a PC has the capability, but the player (perhaps without having told you) decided that he wouldn't have any experience with it, despite having the mechanical ability to try. And then you watch what you thought was a doable challenge become insoluble because the players won't even TRY "because RP."

[Kurald Galain]

I said this in another topic but I think part of the problem is that the knee jerk reaction to automatically reach for the d20 anytime a situation would arise regardless of situation.
Not at all. The problem has never been rolling too many checks. The problem is checks not getting the probability of outcomes that people expect.

The idea that checks should only be rolled when they matter isn't unique to 5E, after all. People just like to claim it is, but it equally applies to every other version of D&D, as well as most other RPGs on the market. That was not the topic of the thread to begin with.

[DanyBallon]

RP is great. I love it. I often make choices along the lines of, "My character would have no experience with this, so I am not going to roll." However, that is not the right choice to make in all situations, and at all tables. In fact, at many tables, doing that has led to the problem becoming unsolvable. Because my character mechanically could, in theory, have done it, and the others who tried had (by luck) failed. My choice not to try becomes mechanically disadvantageous, and actively can diminish the fun of everybody else at the table.

Now, obviously, if my character CANNOT do it rather than merely DOES NOT, the end result isn't much different, but at least then I'm not deliberately choosing to fail, OOC. At that point, the resources that could be brought to bear have been.

And, as a GM, it is often frustrating to know that a PC has the capability, but the player (perhaps without having told you) decided that he wouldn't have any experience with it, despite having the mechanical ability to try. And then you watch what you thought was a doable challenge become insoluble because the players won't even TRY "because RP."

In such case, the players needs to find a new way to overcome the challege and the DM needs to adapt its quest, the same way creative thinking from players may overcome in an unexpected way a well planned encounter by the DM.

[Shaofoo]

RP is great. I love it. I often make choices along the lines of, "My character would have no experience with this, so I am not going to roll." However, that is not the right choice to make in all situations, and at all tables. In fact, at many tables, doing that has led to the problem becoming unsolvable. Because my character mechanically could, in theory, have done it, and the others who tried had (by luck) failed. My choice not to try becomes mechanically disadvantageous, and actively can diminish the fun of everybody else at the table.

Now, obviously, if my character CANNOT do it rather than merely DOES NOT, the end result isn't much different, but at least then I'm not deliberately choosing to fail, OOC. At that point, the resources that could be brought to bear have been.

And, as a GM, it is often frustrating to know that a PC has the capability, but the player (perhaps without having told you) decided that he wouldn't have any experience with it, despite having the mechanical ability to try. And then you watch what you thought was a doable challenge become insoluble because the players won't even TRY "because RP."

Well at the very least why would the characters ask the uneducated and slightly sub average intelligence to check the complex arcane rune? The RP buy in has to apply to the entire group. If you are the only one that is RPing then there will be some friction.

Maybe another thing that could help is that success gives different results. For the Wizard getting a success will give a detailed description while the Barbarian will give a couple of clues. Basically success is individually defined.

Also in tangential to the previous part maybe don't make failure as if you don't know anything. Even if you roll a 1 you still learn something, maybe not much and probably some things might be erroneous but it isn't "I spend the hour staring at the sun as I apparently forgot how to open a book during that time". It is a common rule when the DM has to use tracking to keep the story going and everyone botches it, this might be useful if the DM wants the players to succeed on a skill check but of course the expanded results will require more work for the DM.

But there is no right answer and I feel there never will be especially when differing ideologies come to the table (including players who only care about the mechanical side that will take every opportunity to get any advantage world building be damned). It will either require cooperation from the entire group and/or more work for the DM. There is no easy way to fix this.


Not at all. The problem has never been rolling too many checks. The problem is checks not getting the probability of outcomes that people expect.

The idea that checks should only be rolled when they matter isn't unique to 5E, after all. People just like to claim it is, but it equally applies to every other version of D&D, as well as most other RPGs on the market. That was not the topic of the thread to begin with.

The complaint was that people are rolling checks when they supposedly had no business to roll a check, it is too much in the sense that not everyone should be doing it. The problem is the bad habit of reaching out and taking the skill check. Even if the Barbarian and Wizard rolling Arcana was the only check in the game that is still a problem. The number of checks is not the problem (in fact I dare say that in general the number of skill checks is too low).

[Kurald Galain]

The complaint was that people are rolling checks when they supposedly had no business to roll a check,
Check the thread title please.

[Tanarii]

I never said the 5e skill system was a failure; I said it failed to meet the needs of many of the posters in this thread, and I said that was a weakness of the system. I stand by my statement. Any element of any system can be a weakness if it fails to meet the needs of people who want to play the system. That doesn't preclude the possibility that such an element can be seen as a strength by others (although I've seen nothing to suggest that either view on this element commands a majority). And I completely agree that when a system doesn't meet your needs it's house-rule time, hence my participation in this thread devoted to discussing the weakness and how to work around it.Fair enough, and well put. And I think some of the suggestions made in this thread are good ones. At the very least, it's made me consider how I adjudicate skill checks.

[Shaofoo]

Check the thread title please.

It seems that the title is about implementing the old style of adding bonuses and penalties for every event that can be quantified, a system that was thrown out in 5e. It seems that the "problem" is that there are people wanting the old system of skills. Which if you want to do it then that is fine and can be done with the new system but it isn't a problem in so much as the system not catering to your specific needs.

Also I was talking about other situations that other people presented as well so when I said the complaint I wa talking about that particular quote that I was saying.

[Xetheral]

When you get +20 to +30 to a stat or skill check, well, why roll dice at that point?

Why, to find out just how awesome the character performs right then, of course! :smallsmile: Which is important if they're trying to do something that even such an expert would find challenging.

One of the most memorable characters from one of the games I ran back in the 3.5 days maxed Profession (Cook)--it was a defining part of the character. Routinely rolling above 30 led to the aromas attracting interesting NPCs from the road when the party camped, the ability to host truly stupendous banquets that put everyone in a good mood and gave a circumstance bonus to diplomacy checks, and, when stepping up to help during a crisis, a morale boost for townsfolk surviving on limited (but now very tasty!) rations. It was awesome to see and frankly is just about the only thing I remember of that character--I couldn't even tell you what class it was.

In 5e, as written, I have two choices: I can set the DCs for such tasks low enough to give characters proficient with cooking tools a reasonable chance of success (thus giving everyone else at the table an implausibly-high slight chance of success), or I can set the DC for such tasks high enough to preclude the non-proficient characters from succeeding (thus giving the proficient character a disappointingly-low chance of success). Either way, there is no way to use the rules to model a character skilled enough to routinely accomplish tasks that no one else can.

Many of the house-rules/rulings/interpretations suggested in this thread and elsewhere don't entirely solve this problem. Setting the DCs low and forbidding non-proficient characters from rolling (or interpreting the same result differently between proficient and non-proficient characters) only helps a little: any proficient character can now accomplish these amazing tasks with nearly exactly the same likelihood of success (note that there is no way to gain expertise in cooking tools). When all proficient cooks can accomplish "amazing" things, none of them can truly stand out from the rest, meaning that even with this supposed fix, there are still no true experts.

Completely rescaling skill bonuses (in line with the double-proficiency, +3 DC idea) has more promise--the inability to create a standout expert in 5e is a solvable problem with a radical enough overhaul. But, to paraphrase a returning player from a recent 5e character creation session: "So wait... there's no way to create [that old cook character] in 5e? Remind me again why this edition is a good thing?"

My point is that, while the removal of extremely high bonuses may have solved problems that were occurring at many tables, it did so in a way that stripped from every table the ability of the base system to model true experts in mundane skills. (Of course, not every table had players wanting to play such a character, but the ability was there.) A player who wants to play such a true expert can work with their DM to adapt 5e in such a way as to add that feature back in at their table, but I find that to be a very disappointing state of affairs and one that diminishes my enjoyment of 5e. Given that threads such as this are fairly common, I feel justified in inferring that I'm not the only one who feels that way.

[newsman77]

Why, to find out just how awesome the character performs right then, of course! :smallsmile: Which is important if they're trying to do something that even such an expert would find challenging.

One of the most memorable characters from one of the games I ran back in the 3.5 days maxed Profession (Cook)--it was a defining part of the character. Routinely rolling above 30 led to the aromas attracting interesting NPCs from the road when the party camped, the ability to host truly stupendous banquets that put everyone in a good mood and gave a circumstance bonus to diplomacy checks, and, when stepping up to help during a crisis, a morale boost for townsfolk surviving on limited (but now very tasty!) rations. It was awesome to see and frankly is just about the only thing I remember of that character--I couldn't even tell you what class it was.

In 5e, as written, I have two choices: I can set the DCs for such tasks low enough to give characters proficient with cooking tools a reasonable chance of success (thus giving everyone else at the table an implausibly-high slight chance of success), or I can set the DC for such tasks high enough to preclude the non-proficient characters from succeeding (thus giving the proficient character a disappointingly-low chance of success). Either way, there is no way to use the rules to model a character skilled enough to routinely accomplish tasks that no one else can.

Many of the house-rules/rulings/interpretations suggested in this thread and elsewhere don't entirely solve this problem. Setting the DCs low and forbidding non-proficient characters from rolling (or interpreting the same result differently between proficient and non-proficient characters) only helps a little: any proficient character can now accomplish these amazing tasks with nearly exactly the same likelihood of success (note that there is no way to gain expertise in cooking tools). When all proficient cooks can accomplish "amazing" things, none of them can truly stand out from the rest, meaning that even with this supposed fix, there are still no true experts.

Completely rescaling skill bonuses (in line with the double-proficiency, +3 DC idea) has more promise--the inability to create a standout expert in 5e is a solvable problem with a radical enough overhaul. But, to paraphrase a returning player from a recent 5e character creation session: "So wait... there's no way to create [that old cook character] in 5e? Remind me again why this edition is a good thing?"

My point is that, while the removal of extremely high bonuses may have solved problems that were occurring at many tables, it did so in a way that stripped from every table the ability of the base system to model true experts in mundane skills. (Of course, not every table had players wanting to play such a character, but the ability was there.) A player who wants to play such a true expert can work with their DM to adapt 5e in such a way as to add that feature back in at their table, but I find that to be a very disappointing state of affairs and one that diminishes my enjoyment of 5e. Given that threads such as this are fairly common, I feel justified in inferring that I'm not the only one who feels that way.


Wow, that's a pretty good story. Wish I was there for that one. I bet your entire table still remembers that. 5e can still provide those memorable moments, without having to dramatically alter the rules to mirror 3.5. In fact, our table has had more memorable experiences with 5e, than any of the previous editions, and we've all been playing since the 80's. 5e is very "freeing" for players and DM's.... but if you don't know how to take advantage of that freedom, then of course you're going to dislike it and pine for the previous edition.

I guess I'm very satisfied with the skill system, because I don't think PC's are necessarily "experts" at their given skills. Not so much that they deserve an skill bonus so high they don't need to roll dice. Sure, you can craft, cook, swim, run or whatever else... but you're not an artisan or champion level, like the guy spending his entire life mastering his art. You are a fighter, cleric, etc... and you spend a lot of time wandering the world, righting wrongs (or causing them). You can always work with your DM to create a "Cook" class and take levels in that, but that's not what a majority of players want.

I agree with many posters, the 5e skill system is good. It puts the focus on story telling, flexibility and fun instead of how many mechanical bonuses can I accumulate to prove I'm awesome. Sure, there are many people who hate on anything that's not 3.5, because that's what they're used to. Heck, I run into them in my community all the time. However, I would wager with 5e's success, there's a equal or even larger percentage of people who are just fine with the skill system, that don't say a word about it. A failure to be just like a previous edition is not a failure at all.

[Grod_The_Giant]

It seems that the title is about implementing the old style of adding bonuses and penalties for every event that can be quantified, a system that was thrown out in 5e. It seems that the "problem" is that there are people wanting the old system of skills. Which if you want to do it then that is fine and can be done with the new system but it isn't a problem in so much as the system not catering to your specific needs.

Also I was talking about other situations that other people presented as well so when I said the complaint I wa talking about that particular quote that I was saying.
Since we seem to be drifting, let me reiterate the gist of my dilemma:

I like 5e characters, I like the idea of proficiency, I like the idea that there's always a chance that the wizard can succeed on his Stealth roll, that the rogue still has to worry about low-level guards, and that hordes of weak monsters are always a threat.

What I find worrysome is the lack of distinction between training and not-training. The difficulty of immersion created by the bard routinely flubbing social checks that the barbarian makes, or by the rogue and wizard being about equally good at sneaking around. For what it's worth, I did just run a 5e game last night, at level 4. I did my best to only call for rolls when they were relevant, to let people who were proficient do their thing. It... didn't always help; when the dice did come out there was an upsetting lack of consistency.

As I've said before, the issue isn't so much the bonus size as it is the bonus size relative to the RNG--to the die size and distribution. I don't care if I don't have a +30 to Stealth, I just want my character who's supposed to be trained in stealth to have more than a 50% chance of succeeding on basic checks (moderate difficulty at low levels, hard difficulty at mid levels, and very high at high levels). A trained professional, in my opinion, should have about a 70% success chance on level-appropriate, general-difficulty checks; that's enough to feel like a pro but not so much that there's not a chance of failure. 50% makes for a much more slapstick-y feel--appropriate for Paranoia, maybe, but not D&D.

Options for addressing that seem to be:

"Suck it up, muchkin, 5e is perfect and you're just a cheating optimizer who wants to always win": You're not going to persuade me that there's not a problem *for my table*, so please either make helpful suggestions of leave the thread.
Clear definitions of what difficulties mean: Worth talking about at the table, and a clear failure of the book (I spend several thousand words talking about this when I worked on a DM-driven rules light system). Doesn't help with the math, though.
Only roll when things are interesting: Obvious; works for "roll often enough and of course you'll fail a bunch" issues but doesn't address success rate
Bring back trained-only checks: Works for some cases (contests), but doesn't address the success rate issue.
Write a more in-depth permissions system, to add functionality to skills you're proficient in: Would make having proficiency feel much more useful, but (probably) doesn't fix the success rate issue. It would also take a lot of work. Might be fun anyway?
Use different dice: Can easily solve probability problems, but takes away a bit of the D&D feel and can potentially get complicated.
Increase the rate of proficiency scaling: Solves a lot of the issues, but potentially requires changes to a lot of combat numbers on monsters.

Anything else? Thoughts on the above?

[Tanarii]

In 5e, as written, I have two choices: I can set the DCs for such tasks low enough to give characters proficient with cooking tools a reasonable chance of success (thus giving everyone else at the table an implausibly-high slight chance of success), or I can set the DC for such tasks high enough to preclude the non-proficient characters from succeeding (thus giving the proficient character a disappointingly-low chance of success). Either way, there is no way to use the rules to model a character skilled enough to routinely accomplish tasks that no one else can.Third choice: Only make checks for things that have a chance of failing. IMO the default assumption of 5e is that Skills are the things that all PC adventurers should be capable of to some degree, and can make a check regardless of proficiency or not. And proficiency just means "as good as a naturally skilled PC with a high attribute", not "trained". However, yes, if you don't like that interpretation, you have to create some artificial divide as to who may make checks and who can't.

Also note that for Tools, you have to be proficient or you can't use them at all.

[DanyBallon]

Options for addressing that seem to be:

"Suck it up, muchkin, 5e is perfect and you're just a cheating optimizer who wants to always win": You're not going to persuade me that there's not a problem *for my table*, so please either make helpful suggestions of leave the thread.
Clear definitions of what difficulties mean: Worth talking about at the table, and a clear failure of the book (I spend several thousand words talking about this when I worked on a DM-driven rules light system). Doesn't help with the math, though.
Only roll when things are interesting: Obvious; works for "roll often enough and of course you'll fail a bunch" issues but doesn't address success rate
Bring back trained-only checks: Works for some cases (contests), but doesn't address the success rate issue.
Write a more in-depth permissions system, to add functionality to skills you're proficient in: Would make having proficiency feel much more useful, but (probably) doesn't fix the success rate issue. It would also take a lot of work. Might be fun anyway?
Use different dice: Can easily solve probability problems, but takes away a bit of the D&D feel and can potentially get complicated.
Increase the rate of proficiency scaling: Solves a lot of the issues, but potentially requires changes to a lot of combat numbers on monsters.

Anything else? Thoughts on the above?


Use common sense: Needs a good collaboration between players and DM, for a given task, some character might roll, other autosucceed and other autofail, given the situation and character background. Skill checks happens only when it makes sense, build some tention, push the story forward, and/or bring more fun.

[Segev]

Another imperfect solution (because it requires making it something central to the character) would be to customize a background to have the background-granted ability be something that makes you able to do something amazing with a given skill.

"Chef" as a background might come with the "gourmet" power, which specifically gives you advantage when cooking and lets you add your proficiency bonus, and also makes anything you prepare really delicious. I'm not sure the mechanical modeling for that, but by making it specifically that yours is above and beyond even what the dice say in quality would be feasible.

But that only helps with ONE thing the character can be good at, and comes at a hefty investment (can't be a Sage, or a Hermit, or a Criminal, or anything else).

[pwykersotz]

Options for addressing that seem to be:

"Suck it up, muchkin, 5e is perfect and you're just a cheating optimizer who wants to always win": You're not going to persuade me that there's not a problem *for my table*, so please either make helpful suggestions of leave the thread.
Clear definitions of what difficulties mean: Worth talking about at the table, and a clear failure of the book (I spend several thousand words talking about this when I worked on a DM-driven rules light system). Doesn't help with the math, though.
Only roll when things are interesting: Obvious; works for "roll often enough and of course you'll fail a bunch" issues but doesn't address success rate
Bring back trained-only checks: Works for some cases (contests), but doesn't address the success rate issue.
Write a more in-depth permissions system, to add functionality to skills you're proficient in: Would make having proficiency feel much more useful, but (probably) doesn't fix the success rate issue. It would also take a lot of work. Might be fun anyway?
Use different dice: Can easily solve probability problems, but takes away a bit of the D&D feel and can potentially get complicated.
Increase the rate of proficiency scaling: Solves a lot of the issues, but potentially requires changes to a lot of combat numbers on monsters.

Anything else? Thoughts on the above?

Theres always group ability checks on PHB 175 for when multiple characters are trying the same thing. I didn't think it was relevant at first, but the focus here seems to be the Wizard failing his good thing while the Barbarian succeeds, and this encourages a mindset closer to the Barbarian mentioning something that jogs the Wizard's memory and such.

[Zaq]

Also note that for Tools, you have to be proficient or you can't use them at all.

This isn't fully supported by the rules. It is true that you must be proficient with a relevant crafting tool to perform crafting during downtime (PHB pg. 187), but I can't find any rules saying that you have to be proficient in a tool to use it at all. (If you do have such a citation, please share it.)

If it were the case that you needed to be proficient with a tool to even use it, the "draw a map" activity on PHB pg. 183 is misleading (it doesn't say a damn thing about catrographer's tools).

Or, worse, it means that basically everyone in the D&D universe is the equivalent of the culinarily clueless college freshman who lives on pop-tarts, because most people aren't proficient with cook's utensils. So if we accept your argument that you must be proficient with a tool to use it, then almost no one knows how to make dinner. And I mean, I guess we can roll with that if you're going for an especially silly/comedic game where everyone is just running around going "HOW DO I FOOD" and trying to shove raw eggs in their mouths (and only the occasional Guild Artisan or Folk Hero can actually cook, though I guess that explains why the Folk Hero is a hero), but I somehow doubt that this is actually how the universe is assumed to work.

[pwykersotz]

This isn't fully supported by the rules. It is true that you must be proficient with a relevant crafting tool to perform crafting during downtime (PHB pg. 187), but I can't find any rules saying that you have to be proficient in a tool to use it at all. (If you do have such a citation, please share it.)

If it were the case that you needed to be proficient with a tool to even use it, the "draw a map" activity on PHB pg. 183 is misleading (it doesn't say a damn thing about catrographer's tools).

Or, worse, it means that basically everyone in the D&D universe is the equivalent of the culinarily clueless college freshman who lives on pop-tarts, because most people aren't proficient with cook's utensils. So if we accept your argument that you must be proficient with a tool to use it, then almost no one knows how to make dinner. And I mean, I guess we can roll with that if you're going for an especially silly/comedic game where everyone is just running around going "HOW DO I FOOD" and trying to shove raw eggs in their mouths (and only the occasional Guild Artisan or Folk Hero can actually cook, though I guess that explains why the Folk Hero is a hero), but I somehow doubt that this is actually how the universe is assumed to work.

Wait, what? Tools can be trained. There could be plenty of people with plenty of proficiencies if you were dead set on engulfing NPC's with these rules.

[mgshamster]

Another imperfect solution (because it requires making it something central to the character) would be to customize a background to have the background-granted ability be something that makes you able to do something amazing with a given skill.

"Chef" as a background might come with the "gourmet" power, which specifically gives you advantage when cooking and lets you add your proficiency bonus, and also makes anything you prepare really delicious. I'm not sure the mechanical modeling for that, but by making it specifically that yours is above and beyond even what the dice say in quality would be feasible.

But that only helps with ONE thing the character can be good at, and comes at a hefty investment (can't be a Sage, or a Hermit, or a Criminal, or anything else).

Background features shouldn't be granting advantage or a proficiency bonus - it should be something that bypasses a skill check altogether, or allows a skill check when no one else would ever be able to. Something like, "You can always prepare delicious and nutritious meals even from the lowest quality ingredients. In addition, rations last twice as long for you and up to 5 others." Maybe something else, to. But a simple "have advantage and proficiency" is way too limiting for a background feature.

[Segev]

Background features shouldn't be granting advantage or a proficiency bonus - it should be something that bypasses a skill check altogether, or allows a skill check when no one else would ever be able to. Something like, "You can always prepare delicious and nutritious meals even from the lowest quality ingredients. In addition, rations last twice as long for you and up to 5 others." Maybe something else, to. But a simple "have advantage and proficiency" is way too limiting for a background feature.

Good points.

[Grod_The_Giant]

Use common sense: Needs a good collaboration between players and DM, for a given task, some character might roll, other autosucceed and other autofail, given the situation and character background. Skill checks happens only when it makes sense, build some tention, push the story forward, and/or bring more fun.

Downside: Potentially leads to game-delaying arguments about whether or not you can/have to/shouldn't roll (and they will happen, even with good groups. They might not be bitter, they might not be prolonged, but I really don't want to have to discuss every roll); still doesn't address the issue of swingy results and poor success rates when the dice come out.

[newsman77]

I guess I don't see the big deal here. Even in real life, there's always a chance some novice is going to remember or do something that astounds the experts. Case in point, the scientists who spend hours upon hours studying astronomy, then some student picks out some weird piece of evidence everyone overlooked that turns out to be something like a new comet. It happens.

What are the DC's... like Trivial (5), Easy (10), Moderate (15), up to Hard (20). If you set them appropriately, then the PC's who are proficient in a skill will have a moderately successful time, where the PC's that aren't, will mostly fail, but get the occasional success. Just like real life, that success happens. It won't be as often as you're making it out to be.

At level 4, most PC's will most likely have +5 - +7 in a skill they're proficient with. To his a 15 - 20 sounds reasonable. It's even easier with a Cleric on board.

[Tanarii]

This isn't fully supported by the rules. It is true that you must be proficient with a relevant crafting tool to perform crafting during downtime (PHB pg. 187), but I can't find any rules saying that you have to be proficient in a tool to use it at all. (If you do have such a citation, please share it.)Chapter 5, Equipment:
"A tool helps you to do something you couldn’t otherwise do, such as craft or repair an item, forge a document, or pick a lock."
The individual tool descriptions let you know what you can do with them. Artisan's tools are very broad, and say "These special tools include the items needed to pursue a craft or trade."


Or, worse, it means that basically everyone in the D&D universe is the equivalent of the culinarily clueless college freshman who lives on pop-tarts, because most people aren't proficient with cook's utensils. So if we accept your argument that you must be proficient with a tool to use it, then almost no one knows how to make dinner. And I mean, I guess we can roll with that if you're going for an especially silly/comedic game where everyone is just running around going "HOW DO I FOOD" and trying to shove raw eggs in their mouths (and only the occasional Guild Artisan or Folk Hero can actually cook, though I guess that explains why the Folk Hero is a hero), but I somehow doubt that this is actually how the universe is assumed to work.1) You're trying to qualify any task related to the name of the artisan's tool as requiring the tool to succeed, not just artisan level tasks (or highly skilled tasks). As I quoted above, artisan tools are needed to pursue a craft or trade. You don't need cooks utensil to cook food. You need them to pursue the trade of being a chef, or professional cook. Or a professional map maker. Or carpenter.

Similarly you need proficiency in a disguise kit to apply a disguise, not put on makeup. You need to know how to play a game or instrument to play them at all. Navigators to set a ships course. Non-artisan tools tell you how they are used.

2) You're assuming doing anything requires a check. "Every action taken requires a rule/check" is simulationist thinking.

Edit: I accept it's equally possible to interpret the chapter 5 equipment wording as meaning: You need the tools to do X. If you're not procient, do X without a proficiency bonus. If you are proficient, do X with a proficiency bonus. ie you need the tool to do the task, not *proficiency* in the tool to do the task. Second point still stands though, as does not needing artisan's tools for non-artisan tasks.

[mgshamster]

Downside: Potentially leads to game-delaying arguments about whether or not you can/have to/shouldn't roll (and they will happen, even with good groups. They might not be bitter, they might not be prolonged, but I really don't want to have to discuss every roll); still doesn't address the issue of swingy results and poor success rates when the dice come out.

Has anyone even had any game delaying arguments over this, or is this just a hypothetical?

In my own games, these "arguments" take maybe 30 seconds to a minute. It's just a quick, "I don't think your character would be able to do that; thoughts?" With a quick reply from the player, "I agree" or "Here's why I think my haracters should be able to." With a final yes or no from me. The game moves on. Most of the time, no one makes a roll, as everyone is simply able to accomplish what they set out to do. Sometimes, they make a roll to see if they succeed - and then sometimes only certain characters can make the check. And sometimes, no one makes a roll simply because it's an auto failure.

Example from one of my games: the party was trying to swim up a hole and then to the surface of the water from a sudden underwater cave (collapse of a cave ceiling led to sudden waterfall filling up the chamber; they could swim up to escape). Those with proficiency in athletics and no armor (or light armor) didn't have to make a check, they auto succeeded. Those with medium/heavy armor OR no proficiency had to make a check. Those with medium/heavy armor AND no proficiency auto failed and required assistance from others to make the swim.

[Xetheral]

Wow, that's a pretty good story. Wish I was there for that one. I bet your entire table still remembers that. 5e can still provide those memorable moments, without having to dramatically alter the rules to mirror 3.5. In fact, our table has had more memorable experiences with 5e, than any of the previous editions, and we've all been playing since the 80's. 5e is very "freeing" for players and DM's.... but if you don't know how to take advantage of that freedom, then of course you're going to dislike it and pine for the previous edition.

Can you provide an example of how to have a memorably-skilled cook in 5e? How do you prefer to change the base rules to accommodate that? (Or, if you're claiming that the base rules already accommodate that, can you please explain further?)


I guess I'm very satisfied with the skill system, because I don't think PC's are necessarily "experts" at their given skills. Not so much that they deserve an skill bonus so high they don't need to roll dice. Sure, you can craft, cook, swim, run or whatever else... but you're not an artisan or champion level, like the guy spending his entire life mastering his art. You are a fighter, cleric, etc... and you spend a lot of time wandering the world, righting wrongs (or causing them). You can always work with your DM to create a "Cook" class and take levels in that, but that's not what a majority of players want.

I don't see character classes as a profession. But even if I did, what's to say the fighter didn't start life as a world-class chef before circumstances forced him to pick up a sword and defend his kitchen? Are you truly ok with simply forbidding a PC from being a master of a skill? Doesn't that artificially curtail the range of permissible characters?


However, I would wager with 5e's success, there's a equal or even larger percentage of people who are just fine with the skill system, that don't say a word about it.

By purchasing the books and actively playing the system I'm a part of 5e's success. And yet I still don't like the skill system. I don't see how you can infer from the success of the whole product that a majority of people like any specific individual element.

----------


Third choice: Only make checks for things that have a chance of failing. IMO the default assumption of 5e is that Skills are the things that all PC adventurers should be capable of to some degree, and can make a check regardless of proficiency or not. And proficiency just means "as good as a naturally skilled PC with a high attribute", not "trained". However, yes, if you don't like that interpretation, you have to create some artificial divide as to who may make checks and who can't.

Also note that for Tools, you have to be proficient or you can't use them at all.

Good point on tool proficiency. But I don't see how that helps: everyone is either (equally) competent or can't even try. It also introduces the new problem that for tasks like "make a dinner to impress a minor NPC with an unsophisticated palate" (e.g., DC 5-10), those without proficiency can't possibly succeed.

And only rolling when there is a chance of failure (which I already do) isn't a third a choice at all. It's equally compatible with both of the options for setting the DC that I listed, because I'm only going to set a DC if I ask for a roll.

The rest of what you wrote suggests that you agree with me that under the base rules of 5e a character cannot be a true expert in a mundane skill? Or am I misunderstanding?

[Shaofoo]

Options for addressing that seem to be:

"Suck it up, muchkin, 5e is perfect and you're just a cheating optimizer who wants to always win": You're not going to persuade me that there's not a problem *for my table*, so please either make helpful suggestions of leave the thread.
Clear definitions of what difficulties mean: Worth talking about at the table, and a clear failure of the book (I spend several thousand words talking about this when I worked on a DM-driven rules light system). Doesn't help with the math, though.
Only roll when things are interesting: Obvious; works for "roll often enough and of course you'll fail a bunch" issues but doesn't address success rate
Bring back trained-only checks: Works for some cases (contests), but doesn't address the success rate issue.
Write a more in-depth permissions system, to add functionality to skills you're proficient in: Would make having proficiency feel much more useful, but (probably) doesn't fix the success rate issue. It would also take a lot of work. Might be fun anyway?
Use different dice: Can easily solve probability problems, but takes away a bit of the D&D feel and can potentially get complicated.
Increase the rate of proficiency scaling: Solves a lot of the issues, but potentially requires changes to a lot of combat numbers on monsters.

Anything else? Thoughts on the above?

Well I have a couple of ideas from another topic about success.

-Make a group check but the one most proficient still gets to be the one making the discovery in RP. For example if the barbarian finds more info about a rune than the Wizard you can say the barbarian gave the Wizard enough hints to help the Wizard. This will probably not sit well with everyone.

-Make success individual per character. The Wizard making his check will find out detailed information about the rune, the Barbarian making the check will find some facts about it. This also ties in with failure not meaning you don't learn anything, failure means that what you know might either be incomplete or erroneous but you learn something, failure should never look like the world didn't change.

[DanyBallon]

Downside: Potentially leads to game-delaying arguments about whether or not you can/have to/shouldn't roll (and they will happen, even with good groups. They might not be bitter, they might not be prolonged, but I really don't want to have to discuss every roll); still doesn't address the issue of swingy results and poor success rates when the dice come out.

Swingy result are part of the game and can become part of the character.

i.e. in one campaing I'm playing, the DM still ask for roll for every challenge as it was in 3.x (he's new to 5e and we don't bother much yet.) So in the game, my character with +3 to Arcana and History (the best bonuses in the party so far) constently roll low (most of the time a 3 whatever the d20 I use) on the dice, and the Warlock with a whooping -1 for both skills keep rolling 15+ for these specific checks, so my character is now seen as knowledgable but incapable to remember anything on the fly, while the warlock has the Dark Ones whispering him the answers.


As for a reply to the downside you see, I'll refer to mgshamster post as he expressed my thoughts, much more clearly that I would.

[Tanarii]

Good point on tool proficiency. But I don't see how that helps: everyone is either (equally) competent or can't even try. It also introduces the new problem that for tasks like "make a dinner to impress a minor NPC with an unsophisticated palate" (e.g., DC 5-10), those without proficiency can't possibly succeed.

And only rolling when there is a chance of failure (which I already do) isn't a third a choice at all. It's equally compatible with both of the options for setting the DC that I listed, because I'm only going to set a DC if I ask for a roll.Addressed that above, you don't need artisan's tools just to cook dinner. You need them to practice your craft/trade. Similarly, you shouldn't need to make a check unless it matters.

But yeah, your specific example of "make a dinner to impress a minor NPC" seems like a good one to me. That sounds like a Cha (Cook's Utensil's) check. And yeah, either you can't do it at all if you aren't proficient, or you can but get no bonus. (Depending on how you interpret the Chapter 5 equipment wording.)


The rest of what you wrote suggests that you agree with me that under the base rules of 5e a character cannot be a true expert in a mundane skill? Or am I misunderstanding?Qualified yes, I agree.

I think that PCs are adventurers. They can do all the mundane adventure related stuff they need to without a check. They can also do any exceptional adventure related stuff they need to with an ability check, with success on DM's determination of difficulty, and natural skill. They can also do some specific exceptional stuff (listed under PHB Skills & Tools) with an ability check, and be just as good as if they were naturally skilled if they are proficient, by adding a proficiency bonus approximately equal to a "high" natural ability score for someone of their level.

I think they are "true experts" in adventure related skills/tools when they have both proficiency and high natural skill. And I think being a "true expert" at an adventuring skill/tool means they have 10-30% more skill than an adventurer that is merely proficient or naturally talented, and 20-60% better than one that is neither.

I think that PC adventurers are compared to other PCs, and "mundane" isn't really a relevant term for skills/tool capability when it comes to adventurers. They're adventurers! Comparing them to a full time smith, or a sage, or whatever, doesn't really make much sense. And if you want something more in line with what those guys can do, background features can provide it.

For example, no Barbarian or Wizard, regardless of their Arcana or Nature or Religion score, can compare to what a Sage can do with Research, or a Hermit with Discovery. No Ranger can outdo a Outlander Wanderer. No Assassin a Charlatan with a False Identity. Although those last two come pretty close and have a lot of overlap. ;)

[newsman77]

I don't see character classes as a profession. But even if I did, what's to say the fighter didn't start life as a world-class chef before circumstances forced him to pick up a sword and defend his kitchen? Are you truly ok with simply forbidding a PC from being a master of a skill? Doesn't that artificially curtail the range of permissible characters?

I'm not saying forbid a PC from being a master of a skill. I'm just saying it typically doesn't happen by level 4.

The 5e system already has an element for letting you start as a world-class chef who was forced to become a fighter. It's called custom background. You and the DM make it. Your class feature can be "World Class Cuisine" - You prepare a splendid dish and the aroma brings in people from all parts of the surrounding countryside (insert terrain here). Then let the DM create his own little table for how many people and who shows up. Still uses 5e system.

5e is about fun, make it that way. :)

[newsman77]

Addressed that above, you don't need artisan's tools just to cook dinner. You need them to practice your craft/trade. Similarly, you shouldn't need to make a check unless it matters.

But yeah, your specific example of "make a dinner to impress a minor NPC" seems like a good one to me. That sounds like a Cha (Cook's Utensil's) check. And yeah, either you can't do it at all if you aren't proficient, or you can but get no bonus. (Depending on how you interpret the Chapter 5 equipment wording.)

Qualified yes, I agree.

I think that PCs are adventurers. They can do all the mundane adventure related stuff they need to without a check. They can also do any exceptional adventure related stuff they need to with an ability check, with success on DM's determination of difficulty, and natural skill. They can also do some specific exceptional stuff (listed under PHB Skills & Tools) with an ability check, and be just as good as if they were naturally skilled if they are proficient, by adding a proficiency bonus approximately equal to a "high" natural ability score for someone of their level.

I think they are "true experts" in adventure related skills/tools when they have both proficiency and high natural skill. And I think being a "true expert" at an adventuring skill/tool means they have 10-30% more skill than an adventurer that is merely proficient or naturally talented, and 20-60% better than one that is neither.

I think that PC adventurers are compared to other PCs, and "mundane" isn't really a relevant term for skills/tool capability when it comes to adventurers. They're adventurers! Comparing them to a full time smith, or a sage, or whatever, doesn't really make much sense. And if you want something more in line with what those guys can do, background features can provide it.

For example, no Barbarian or Wizard, regardless of their Arcana or Nature or Religion score, can compare to what a Sage can do with Research, or a Hermit with Discovery. No Ranger can outdo a Outlander Wanderer. No Assassin a Charlatan with a False Identity. Although those last two come pretty close and have a lot of overlap. ;)

(Note that 3e & 4e has the reverse problem. You can only do stuff you are trained in and. You are bad at everything else. So bad you might as well not try a check at all for any task related to the skill.)

Agree 100%!!!

[maxweasel]

The waterfall/air example is just showing how it looks to me when someone tries to argue something that goes against common sense, it's as if you are arguing that you can climb air.


For the rules, let me put it this way, I see making a judgement call as to what someone finds out from an Arcana check based on there background and knowledge a fairly quick simple and easy decision that a DM could make in seconds on the fly in any situation. At least in my own case if I were DMing a campaign, I would almost universally give my arcana proficient characters relatively more information then I would for non proficient regardless of the DC's or rolls. This isn't something that I have to think about in depth. Both the wizard and barbarian make there rolls? Example (terrible spell choice but still):
Barb: You know that it's a fire spell, and you've mostly seen similar looking magic used to blast large areas.
Wizard: You know that it is specifically the fireball spell, and you'd judge it's probably being cast as a 4th level spell.

This doesn't even take serious thought, it's barely more then an afterthought, but it accounts for there difference in knowledge on magic.



In my eyes I feel like all I'm actually advocating is people use common sense.




Theres always group ability checks on PHB 175 for when multiple characters are trying the same thing. I didn't think it was relevant at first, but the focus here seems to be the Wizard failing his good thing while the Barbarian succeeds, and this encourages a mindset closer to the Barbarian mentioning something that jogs the Wizard's memory and such.

A combination of these two strategies is what tend to use when I GM. Someone failing a check doesn't necessarily mean that they "failed." The 5e system seems to place a lot of emphasis on chance in the form of rolling a d20 and this can be worked into the RP aspect of the game. If a rogue fails to pick a lock it doesn't have to simply be "You weren't good enough this time." Something like "While picking the lock your character suddenly sneezes and breaks a tumbler" can be more fun.

I understand this doesn't meet the criteria of the original complaint dealing with the desire for a proficient character to have more than a "50% chance of succeeding on basic checks" but this solution has made my games move smoothly without many complaints. Especially combined with stuff like pwykersotz's suggestion above.

I personally try to find RP solutions for most issues as messing with the math of the game can have negative unintended consequences. Really as long as everyone is having fun whatever solution works for you is fine.

[Tanarii]

Also be aware of the problems you can introduce by insisting on a large spread between experts and non-experts, AND having simulation level checks for all tasks, including mundane tasks: You can only do the few things that you are "good" at. Since you're very likely to fail the other tasks, you might as well not try.

That was a very common problem in 3.5e/Pathfinder (Large difference between skilled and unskilled) and 4e (scaling difficulty by level with loss of success chance as you leveled). 5e attempts to address both of those problems.

Of course, this problem, just as the one in 5e, was exacerbated by DMs calling for checks for things that should have been auto-succeed or auto-fail. And mitigated by DMs only calling for checks when failure is both meaningful and ... uh chanceful.

(Removed this from my previous post because it's kind of a separate thing that needed its own post.)

[ad_hoc]

5e isn't set up to play Cooks and Kitchens and I think that is okay. The thing is, if you want to play a cook, you can have that be a background and have your feature be that you can make great meals.

One house rule that can be used in group checks is to set 2 DCs. A lower one where x number of PCs to hit and a higher one where only 1 PC needs to hit it.

It is much more unlikely to be able to hit a 22 DC without proficiency, and you could single handedly save the day when the rest of the party fails. Alternatively the rest of the party can help out when the talented character flubs it.

[mephnick]

One house rule that can be used in group checks is to set 2 DCs. A lower one where x number of PCs to hit and a higher one where only 1 PC needs to hit it.

It is much more unlikely to be able to hit a 22 DC without proficiency, and you could single handedly save the day when the rest of the party fails. Alternatively the rest of the party can help out when the talented character flubs it.

I like this quite a bit, though the "single-handed" DC would have to be significantly higher. I'm going to try it out.

[Xetheral]

Another imperfect solution (because it requires making it something central to the character) would be to customize a background to have the background-granted ability be something that makes you able to do something amazing with a given skill.

"Chef" as a background might come with the "gourmet" power, which specifically gives you advantage when cooking and lets you add your proficiency bonus, and also makes anything you prepare really delicious. I'm not sure the mechanical modeling for that, but by making it specifically that yours is above and beyond even what the dice say in quality would be feasible.

But that only helps with ONE thing the character can be good at, and comes at a hefty investment (can't be a Sage, or a Hermit, or a Criminal, or anything else).


The 5e system already has an element for letting you start as a world-class chef who was forced to become a fighter. It's called custom background. You and the DM make it. Your class feature can be "World Class Cuisine" - You prepare a splendid dish and the aroma brings in people from all parts of the surrounding countryside (insert terrain here). Then let the DM create his own little table for how many people and who shows up. Still uses 5e system.


And if you want something more in line with what those guys can do, background features can provide it.

For example, no Barbarian or Wizard, regardless of their Arcana or Nature or Religion score, can compare to what a Sage can do with Research, or a Hermit with Discovery. No Ranger can outdo a Outlander Wanderer. No Assassin a Charlatan with a False Identity. Although those last two come pretty close and have a lot of overlap. ;)

Using backgrounds is a decent idea. Unfortunately most background features are rather limited in scope as compared to the open-endedness of a skill check. Still, there's room to play around here if you're willing to push the envelope. Giving expertise in a set of tools other than thieves tools would be a possibility, particularly when combined with the idea of doubling all proficiency bonuses for skills and using higher DCs.

----------


Similarly, you shouldn't need to make a check unless it matters.

You keep mentioning this, despite the fact that I agree with you. Are you reiterating it for emphasis or am I missing something?


But yeah, your specific example of "make a dinner to impress a minor NPC" seems like a good one to me. That sounds like a Cha (Cook's Utensil's) check. And yeah, either you can't do it at all if you aren't proficient, or you can but get no bonus. (Depending on how you interpret the Chapter 5 equipment wording.)

And that doesn't strike you as problematic?


I think they are "true experts" in adventure related skills/tools when they have both proficiency and high natural skill. And I think being a "true expert" at an adventuring skill/tool means they have 10-30% more skill than an adventurer that is merely proficient or naturally talented, and 20-60% better than one that is neither.

We have a VERY different idea of what it means to be an expert in probabilistic terms. Assuming that you're equating your expert's "30% better" with a +6 bonus, your expert will only win against someone with an equal stat but without proficiency 73.75% of the time. 3-to-1 odds sounds to me more like "somewhat skilled" than "true expert". And you only get 3-1 odds near max level--at first level a proficient character only beats an equal-stat non-proficient character 57.25% of the time, which is about 4-to-3 odds.


I think that PC adventurers are compared to other PCs, and "mundane" isn't really a relevant term for skills/tool capability when it comes to adventurers. They're adventurers! Comparing them to a full time smith, or a sage, or whatever, doesn't really make much sense.

It makes perfectly good sense to me. I see no reason we can't have an adventuring smith or an adventuring sage! And why shouldn't the mundane be relevant to adventurers?


Also be aware of the problems you can introduce by insisting on a large spread between experts and non-experts, AND having simulation level checks for all tasks, including mundane tasks: You can only do the few things that you are "good" at. Since you're very likely to fail the other tasks, you might as well not try.

That's only true if you believe skill DCs have to be set high enough to challenge the most-skilled character at the table. If instead you just use normal (i.e. low) DCs, the phenomenally-skilled characters either autosucceed or roll to see just how amazingly-well they did, and everyone else still has a chance. Unfortunately, it was quite common to see DMs set DCs based on the skill bonus of the most-skilled character. Rather than try to address this problem directly in 5e, it seems they instead simply removed any possibility for a character to be skilled enough for this style of DC-setting to create a problem for the other characters.

----------


5e isn't set up to play Cooks and Kitchens and I think that is okay. The thing is, if you want to play a cook, you can have that be a background and have your feature be that you can make great meals.

Are you saying it's ok that the 5e skill system doesn't meet the needs of my table? Or that it's ok because at your table it isn't a problem--no one wants to play such a character anyway? Either way comes across as rather dismissive of other playstyles.

Personally, I think the cook character who played in my old campaign was one of the best I've ever seen. It added player-initiated depth and vibrancy to my campaign. It sounds like you're saying I'm wrong to value such characters as a DM.

[JackPhoenix]

I know that in 3e/PF, there are some skills and uses of skills that are "trained only." No matter your bonus from stat or magic items, if you lack any ranks in the relevant skill, you can't roll it for "trained only" things.

I take it 5e does not have any such clause. No "can only roll this if proficient in the skill" things. I think that would be a good house rule to fix some of the problems that have been brought up here. Primarily the "Barbarian with no Arcane and Int 8 rolling a natural 18 and knowing more about the magical rune than the wizard with Int 20 and proficiency who rolls a natural 5."

That it is only some things that are "proficient only" means that you also don't run into "only people trained in athletics can even try to swim across the river."

In 3.x, you can roll for "trained only" skills, but only for a DC 10 or lower. You need to put at least one skill point to be able to roll for higher DCs.

[pwykersotz]

Are you saying it's ok that the 5e skill system doesn't meet the needs of my table? Or that it's ok because at your table it isn't a problem--no one wants to play such a character anyway? Either way comes across as rather dismissive of other playstyles.

Personally, I think the cook character who played in my old campaign was one of the best I've ever seen. It added player-initiated depth and vibrancy to my campaign. It sounds like you're saying I'm wrong to value such characters as a DM.

In my opinion, it's not a matter of it being okay, just a matter of recognizing and adapting. This edition is fairly modular and flexible and can take a lot more change than 3.5 could without breaking.

The problem, as was stated upthread, is that implementing this system as the norm encourages the same "only roll if you bothered to max your skill ranks" paradigm that 3.5 has. Also, 3.5 has a remarkably similar skill system to 5e with the exception of the point system and slightly flattened numbers, rolling high technically did nothing. You were still supposed to role against a DC or make an opposed check.

So since you ruled in that the roll was just a measure of extremity anyway, why not add something similar to this system? It's a little more work because without the point scaling baked into the system it doesn't happen naturally, but you can (for example) have a cook who has proficiency in cooking tools and went ahead and trained for a further 2500 days (and a further 2500 gold) and they get to add +10 to their roll (still at a max of 30). Maybe add a clause about it not affecting Thieves Tools (unless you want a Rogue to auto-pass the DC, which now that I think about it, you might!) and let the specially trained characters rock those checks. Or instead of a static increase, say that for each extra 2500 days and 2500 gp, they can't roll below a 5, then a 10, then a 15. So for the low, low price of 21 years and 7750gp, you literally cannot roll below a 15 on your cooking check. A true master in the field!

[Tanarii]

Using backgrounds is a decent idea. Unfortunately most background features are rather limited in scope as compared to the open-endedness of a skill check. Still, there's room to play around here if you're willing to push the envelope. Giving expertise in a set of tools other than thieves tools would be a possibility, particularly when combined with the idea of doubling all proficiency bonuses for skills and using higher DCs.I agree more backgrounds are needed. I've also wanted expertise in other tools would be a good house-rule, although spending a precious expertise on something like that as an adventurer is pretty crazy.


You keep mentioning this, despite the fact that I agree with you. Are you reiterating it for emphasis or am I missing something?CYA on my part. I have a tendency to include redundant information so that a paragraph from one post is less easily taken out of context, because it includes all the info in one place, as opposed to assuming someone will put together someone I've said in one post with one said 10 posts later. More importantly, someone other than the person the post was actually responding to is far more likely to overlook comments from different posts.


And that doesn't strike you as problematic?Not really. Cooking an awesome dinner for an NPC being a major plot point in an adventurer's life, to the point where it's important for them to be an expert chef compared to other adventurers, comes up how often? Also, generally speaking, if it's possible to spend precious character creation/development resources to become an "true expert" in such a thing, at the cost of other more important ones, that's considered a trap option and a flaw in system rules. Specifically talking about weird corner non-typical-for-adventurers skills here. Underwater basket-weaving.

Or were you addressing my uncertainty on how the tools rules work? Yeah, I consider that a problem. I'd like to know which way they're supposed to work. ;)


We have a VERY different idea of what it means to be an expert in probabilistic terms. Assuming that you're equating your expert's "30% better" with a +6 bonus, your expert will only win against someone with an equal stat but without proficiency 73.75% of the time. 3-to-1 odds sounds to me more like "somewhat skilled" than "true expert". And you only get 3-1 odds near max level--at first level a proficient character only beats an equal-stat non-proficient character 57.25% of the time, which is about 4-to-3 odds.I assume it's fair to say that a PC adventurer who is a true expert in a specific adventuring task compared to a different PC adventurer who is less expert, but has the basic skill level necessary to be an adventurer at that task, has a difference of no more than 60% chance of success. Or adventurer's opponents. So Yes. We have different views on what it means to be a true expert among adventurers, at adventuring type tasks. It's like comparing two NBA players ... they're both fairly amazing no matter which one is the true expert.


It makes perfectly good sense to me. I see no reason we can't have an adventuring smith or an adventuring sage! And why shouldn't the mundane be relevant to adventurers?Then you're making different assumptions from 5e, which is fair enough. But the other things to consider is : if you spend your character resources developing those skills to become an expert at those non-adventuring tasks, what are you sacrificing to gain it? Only 3e really set up a system to do it that way, and both 4e and 5e moved away from that due negative reactions to the problems it caused.


That's only true if you believe skill DCs have to be set high enough to challenge the most-skilled character at the table. If instead you just use normal (i.e. low) DCs, the phenomenally-skilled characters either autosucceed or roll to see just how amazingly-well they did, and everyone else still has a chance. Unfortunately, it was quite common to see DMs set DCs based on the skill bonus of the most-skilled character. Rather than try to address this problem directly in 5e, it seems they instead simply removed any possibility for a character to be skilled enough for this style of DC-setting to create a problem for the other characters.Sounds to me like instead of the DM judging something an automatic success or automatic failure, you only want it when the dice decide it. That's a fair point. I personally like that D&D 5e doesn't make it possible for a character to auto-succeed on any adventuring task, or conversely auto-fail on a adventuring task. Or waste character development resources trying to keep up with the treadmill for contest skills (perception vs stealth especially).

Edit: Lots of stuff that's responses here, but something that's important to consider if you're creating house rules, is the character resources thing, which is why I came back to it several times. Experienced players don't necessarily mind a weakness in their character, or 'trap' options, so long as what they're getting into. But it's worth being aware if that's the system you're creating with your house rules.

[Tanarii]

To be constructive, instead of just pushing back against needing alternatives, I'll make a couple of house rule suggestions.

Based on keeping DCs the same (DC 5 very easy to DC 30 impossible):
1) allow Rogues (and maybe Bards) to apply Expertise to 'roguish' Tools. Thieves Tools, Disguise Kit, Forgery Kit, Poisoners Kit.
2) same as above but to any Tool.
3) allow any character to use the Training rules for Tools (250 days and 250gp) to get double proficiency bonus with a Tool they are already proficient in. Suggest excluding 'roguish' tools.
4) same as above but also allow combining the same tool proficiency from 2 different sources (Race/Class/Background) to get double proficiency starting at level 1. Still suggest excluding roguish tools.

Change the difficulty level slightly, let's say by adding a new category at the top, DC 35 (miraculous):
5) double proficiency bonus (+4 to +12) for skills. (Note you may want to exclude some rules that use skills, like grappling rules.) Expertise add 1/2 new bonus, and Jack-of-All-Trades & Champion version of it add 1/4 to non-proficient.
6) allow combining the same skill from two different sources (race, class, background) to get double proficiency bonus. Expertise and other things that add double bonus instead add triple (total) if used with an already doubled skill.

7) Change DCs & AC to be double the current range above 10, and double all bonuses and penalties from all sources. So Leather plus Dex 16 is AC 18, Full Plate and shield is AC 30, a first level typically has +8-10 to hit, a max level one +22 to hit. Skill DCs range from 5 (very easy) to 10 (easy) to 30 (hard) to 50 (impossible). Skill modifiers range from -2 (ability score 8, no proficiency) to +22 (Max prof & ability score) to +34 (expertise + max).

[georgie_leech]

7 has a few issues, as that breaks the expectation that sizable numbers of low level creatures remain a factor even at higher levels. Whatever bugs lowish proficiency caused with the skill system, let's not throw the baby out with the bath water, yeah?

[Tanarii]

7 has a few issues, as that breaks the expectation that sizable numbers of low level creatures remain a factor even at higher levels. Whatever bugs lowish proficiency caused with the skill system, let's not throw the baby out with the bath water, yeah?8) same as 7, but only applies to ability/skill checks, not attack rolls or saves. May require tweaking for certain combat related use of skills such as grappling.



Note: Once I start to throw away the idea of skill system balance on the same underlying concept as attack rolls, to whit that low level creatures remain a factor in skill checks even at higher levels, at least one baby is already thrown out as far as I am concerned. So I ran an entire gamut of options, from tweaking unimportant-to-adventuring Tools to complete baby-tossing. :) the main thing I was thinking for #7 is it is really easy to calculate what's going on vs the current system, since it's a global rule for all resolution to do anything. Otoh #8 is too, just for one subsystem of the resolution system.

[georgie_leech]

8) same as 7, but only applies to ability/skill checks, not attack rolls or saves. May require tweaking for certain combat related use of skills such as grappling.



Note: Once I start to throw away the idea of skill system balance on the same underlying concept as attack rolls, to whit that low level creatures remain a factor in skill checks even at higher levels, at least one baby is already thrown out as far as I am concerned. So I ran an entire gamut of options, from tweaking unimportant-to-adventuring Tools to complete baby-tossing. :) the main thing I was thinking for #7 is it is really easy to calculate what's going on vs the current system, since it's a global rule for all resolution to do anything. Otoh #8 is too, just for one subsystem of the resolution system.

Fair enough. That still leads to just automatically being able to, say, elude all low level observers or out shove an infinite number of orcs at a high enough level, but they at least remain a relative combat threat. That much is something so baked into the design that I'm not sure you could still call it 5e after you got rid of it. It'd be an overly simple 3.x with the variety of customization options removed; instead of putting skill points into skills every level, arranging as you see fit, each skill you put a point into got 2 every level instead.

[Tanarii]

Fair enough. That still leads to just automatically being able to, say, elude all low level observers or out shove an infinite number of orcs at a high enough level, but they at least remain a relative combat threat. That much is something so baked into the design that I'm not sure you could still call it 5e after you got rid of it. It'd be an overly simple 3.x with the variety of customization options removed; instead of putting skill points into skills every level, arranging as you see fit, each skill you put a point into got 2 every level instead.Agreed. Personally my preferred tweaks would be all of #2-#4 & #6. Because they don't just stretch out the range of bonuses relative to the range of the die roll, which is what #7/8 are intended to do.

One thing I didn't address is setting DCs for things. That's because it's effectively the same thing as DM-fiat, just known by the players in advance. Any individual DM would need to develop a list of example Very Easy to Impossible tasks for each skill and ability score, then give it to the players to reference. That could be done by taking the PHB list of example things that can be done and assigning DCs, along with looking up lists of 3e and 4e target DCs and adjusting them appropriately to the desired 5e range.

If I get really bored maybe I'll tackle that project so other people can use it as a basis for their own version. ;) Edit: quick search found me this for non-epic DCs from 3.5, which would be useful in creating such a 'target DCs' list for 5e in advance. http://chet.kindredcircle.org/pdf/DnD3.5Index-Skills_Actions.pdf

[Gurifu]

Another method, not mentioned here, is to roll knowledge and awareness related checks behind a screen, come up with a true result and a false result, and give the true result to people who pass and the false result to people who fail by 5 or more. If, four times out of five, the Barbarian's "brilliant" observation about Aracana is total nonsense that his brain spurts out when he strains it too hard, the one time out of five that he properly identifies the Death-curse of Thalzamor of Thay, Dread Mage of the Fifth Epoch, is going to be suspect unless the generally-reliable Wizard agrees with him.

---


I guess I don't see the big deal here. Even in real life, there's always a chance some novice is going to remember or do something that astounds the experts. Case in point, the scientists who spend hours upon hours studying astronomy, then some student picks out some weird piece of evidence everyone overlooked that turns out to be something like a new comet. It happens.

That's a Perception or Investigation check, though, not a Nature check. It's the experts who have to identify the comet (Nature), even if the student is the one who sees it in the telescope (Perception) or picks out an unexpected spike on a chart (Investigation). Your point here actually underscores an issue: certain tasks (like spotting something anomalous) can be done by laypeople, but other tasks (like identifying what the anomaly means) are the exclusive domain of the well-trained, if not necessarily the expert.

The only way to apply this sensibly is to assume that this means that every adventurer has a strong core education in every skill. Occasionally this will still throw unreal results (like the 8-strength halfling rolling natural 20 and dead-lifting a 1000-pound portcullis), but for the most part it makes some sense.

This runs into problems with the many people who want to play uneducated, unfit, or unobservant people. Perhaps a good houserule would be to allow characters to buy "antiproficiency" in a skill, which restricts them to their Passive Skill, in exchange for a minor benefit elsewhere, like a tool proficiency, language, or hit point.

[Serket]

This has been an interesting read, so thank you everyone.

I can't decide which of these two is my favourite quote:

The 5e skill system isn't a failure. It's a reaction to the failures of previous editions, for the majority of players. If it doesn't meet your needs, it's house-rule time. Which is awesome. Do it.

Varying DC by class or only allowing certain characters to roll by class regardless of actual skill bonuses is just a weird system of hidden bonuses and penalties stacked on top of the ones on the character sheet. It's a clumsy patch that adds needless complexity because the base rules aren't properly calibrated, thus producing oddities.

I like the first because it is just so right. This system doesn't make any sense for a simulationist, but it isn't written for simulationists (note, I am primarily a simulationist). D&D has never particularly held with notions of realism nor adherence to any fiction outside of itself, but 5e takes that further than 3e did (I don't know about 4e).

Actually on topic and going on from the second quote, there was an idea floated of DCs adjusted like a circumstance modifier, which was persuasive to me on a personal level in that it's how I like games to run. But it seems to go against the flow of 5e design. If by design there's a 60% difference between untrained and expert (note, I'm just taking that on faith, I did not actually check this), then giving bonus-equivalents to experts is a contradiction of the design. The system design seems to be that your bonuses are relatively small, but exceptional circumstances might grant advantage.

If you don't like the way 5e skills work so much that it is a dealbreaker for you, my advice is that you play a different system, or hack the skills part. There are many many game systems, and many of them do not work like this. Personally, I do not like this, but for me it is not a dealbreaker.
If you are a GM and looking for the simplest possible fix, my recommendation is that you hand out advantage a lot - the mean numerical bonus isn't big, but the rolls become a lot more predictable.


My point is that, while the removal of extremely high bonuses may have solved problems that were occurring at many tables, it did so in a way that stripped from every table the ability of the base system to model true experts in mundane skills. ... Given that threads such as this are fairly common, I feel justified in inferring that I'm not the only one who feels that way.

You're not wrong, your feelings are legit, and to a major degree I feel the same. But ultimately our options are hack it, ditch it, or play anyway.
And the OP, sharing the issue:


A trained professional, in my opinion, should have about a 70% success chance on level-appropriate, general-difficulty checks; that's enough to feel like a pro but not so much that there's not a chance of failure. 50% makes for a much more slapstick-y feel--appropriate for Paranoia, maybe, but not D&D.
(list snipped)
Anything else? Thoughts on the above?

You listed changing the proficiency scaling. That's a tricky thing to alter, because it also alters attack rolls and save DCs. Even if you decouple the skill scaling from the attack roll scaling, grappling is skill based. So I just wouldn't.
As for hacks:

making it easier to gain advantage will even out the probability a bit and help your modifiers matter more.
Adding "passive" checks or the "take 10" option (which are basically the same thing) will also do that.
You can combine high skill mods with skill checks that determine things other than success. Like, this minor carpentry job is not complicated for someone with your passive score, and you will definitely succeed, but the success of your rolls will determine how long it takes (and there are consequences to taking longer).
Add more magic items. Not the game intent, but lots of magic items give you bonuses to a single thing and are also attunement. EG, the rogue has a cloak of elvenkind, so she's got significant bonuses to sneaking. It doesn't make her any better at the other things where she's proficient, but it makes her the expert at the thing she is supposed to be best at.


There's this idea that the barbarian can never succeed at arcana, and I find it kind of silly. I mean, the barbarian has probably met an arcane caster at some point. And, they've probably encountered some weird creatures at some point. And they've probably been in a tavern talking to people who have encountered weird stuff at some point. So they might have picked some stuff up. RAI, I think, is that they get to roll. After all, the designers had plenty of space to put "these skills cannot be attempted without proficiency" somewhere. They didn't.

So if you are a GM and are planning to make some skill checks impossible without proficiency, I strongly urge you to mention that to your players before they pick characters.

This brings up the additional point about how the GM conveys the information gained. Perhaps the wizard passes the check and gets told information as though they read it in a book, while the barbarian gets conveyed information as though they met something similar once, or as though they got told a story in a tavern. "Yeah, that big purplish thing looks a lot like those smaller reddish things you fought once, and normal weapons just bounced off them like they were rock!"

Weird details from this thread I'd like to address:


What if a player claims his character is an expert swimmer even though he's not trained in athletics?

Zazax the Loremaster, master wizard with proficiency bonus 6 and int bonus 5, most knowledgable in magic in all the worlds

The basic problem in these two examples is that player expectations do not align with system mechanics, leading to players creating fiction that does not fit with mechanics. This means somebody needs to explain the system to the players in question. Player 1 needs to know that swimming is a skill (and that there might be options to give them a swim speed), and player 2 needs to be advised that +11 bonus does not make them the best in the world, because there are bards and rogues out there with +17 bonuses (expertise), some of whom also have Guidance and reliable talent. Stacking all that gives a minimum roll of 28, so +11 isn't being "best in the world".


What if the DM claims that another PC is bad at swimming because dwarves can't swim, even though he is trained in athletics?

The problem is that now, your character's swimming prowess depends not on his athletics modifier, but on whether your DM's headcanon of elves matches your character concept. There's a big difference between a Tolkien elf and a Pini elf, after all.


Yes, this is a problem. Usually, games need a self-consistent game fiction. Which means communication, or a solid simulationist system. 5e does not have a solid simulationist system, so I think we have to talk more.
This ties into:


Mostly I just see 3.5 as having been so rules intense that a lot of it's core players have become to used to having every last thing defined by a rule. And I think part of the problem is the difficulty in finding a good DM, one of 3.5 strong suits was that, especially if you used a pre-made adventure, you could have a pretty ****ty DM and still make it work.

Emphasis mine. This is an astute observation. Basically, the more you mechanise stuff, the more stuff works even when the GM doesn't know how or why, because they can just follow a process. 5e relies on GM judgement a lot, and that means good judgement is required.

And... I tend to think of that as poor design. I mean, games should teach you how to play them. I have the "basic" DM book, and it isn't very supportive or explanatory. Hopefully the finished product is better.

And finally:


So if we accept your argument that you must be proficient with a tool to use it, then almost no one knows how to make dinner. And I mean, I guess we can roll with that if you're going for an especially silly/comedic game where everyone is just running around going "HOW DO I FOOD" and trying to shove raw eggs in their mouths

Nah, prestidigitation exists. We warm the egg up and make it taste of something extra first. :smallsmile:

[Segev]

While this probably doesn't help in 5e, I have long thought about where the d20 came from, and why it and the 3d6 stat range were in early D&D. I can't help but wonder if a lot of the out-of-combat resolution came from a simple assumption that your stats were all you had for "skill" type checks, and the goal was simply to roll a d20 and get under your stat to succeed.

This has led me to contemplate a system wherein rolling high is good, but rolling over your stat means you fail. "1" would be the worst possible success you could have, but (absent complicating factors, such as competition), it would be a success.

One would probably either have "skill" or "proficiency" add to the cap. Or maybe BOTH the cap AND the roll (thus making it always raise the result, not merely raise the potential for a good result).

The GM could set minimum success thresholds for particularly difficult tasks. Or multiple success thresholds for tasks.

Perhaps there's a basic piece of knowledge at DC 1 for anybody who rolls under their Intelligence on the Arcana check. Anybody who gets 5 or more is told a bit more information. 10 or more yields still more useful and precise details, up to 20. If the proficiency bonus adds to the die roll and the cap, it could go up to DC 25, potentially. And rolling over your stat means you know nothing of it.


Another approach would be not to set minimum DCs at all, but instead set difficulties by die type. If you have dice ranging from d20 to d40 to d60 to d80 to d100, the more difficult the task, the larger the die the GM requires the players to roll. Does remove a lot of "gradation" potential in the difficulties, though. That could be a good or a bad thing.

[PoeticDwarf]

For ne this is never a problem. I've seen rogues with +24 and even with just +3 instead of +0 the difference can be pretty big

[mgshamster]

Question: Do "level appropriate" challenges happen in 5e?

In 3.X, you'd change the DC to match the PC's skill level. The higher level you are, the higher the DC is to accomplish something extraordinary. (This isn't my choice; I've just seen this happen a thousand times or more in all the years I played 3.X, across many different tables and many different published adventures). For some strange reason, almost every challenge you face just happens to be appropriate for your level. The DC to spot a hidden door at level 1 is around DC 15, but at level 20 it's around DC 40. The creatures that randomly encounter you always happen to be within a few CR of your group.

However, in 5e, DCs are set by the world; whether you pass them or not is dependent on your skill level and a bit of luck, but they don't change just because you're higher level. At least, that's what I've seen from the published adventures so far.

I think one of the issues here is an expectation of "level appropriate" challenges, and that shouldn't really exist anymore.

[Segev]

Question: Do "level appropriate" challenges happen in 5e?

In 3.X, you'd change the DC to match the PC's skill level. The higher level you are, the higher the DC is to accomplish something extraordinary. (This isn't my choice; I've just seen this happen a thousand times or more in all the years I played 3.X, across many different tables and many different published adventures). For some strange reason, almost every challenge you face just happens to be appropriate for your level. The DC to spot a hidden door at level 1 is around DC 15, but at level 20 it's around DC 40. The creatures that randomly encounter you always happen to be within a few CR of your group. This is either a result of bad DMing, or of characters seeking out more challenging things (which is appropriate). The trouble is when the dungeons aren't really any more well-designed and owned/built by things of the appropriate CR; then it becomes unbelievable that the secret door and trap are all such high DC.

In 3e and 5e both, the DCs are supposed to be set by the world. The difference is that the skill and challenge of the things the players are facing should, independent of the PCs' level, be stronger. The PCs' level lets them face those challenges.

A 3e PC going back to town and finding some 1st-level kobold has trapped his door should have no trouble with that search DC; the kobold's trap has a very low hide check result compared to the PC's high-level bonus. A 5e PC will never encounter a trap harder than a certain DC, however, because of bounded accuracy. At least, in theory. So his lower bonuses only make him more likely to succeed, not ABLE to succeed.

[Tanarii]

That's a Perception or Investigation check, though, not a Nature check. It's the experts who have to identify the comet (Nature), even if the student is the one who sees it in the telescope (Perception) or picks out an unexpected spike on a chart (Investigation). Your point here actually underscores an issue: certain tasks (like spotting something anomalous) can be done by laypeople, but other tasks (like identifying what the anomaly means) are the exclusive domain of the well-trained, if not necessarily the expert.

The only way to apply this sensibly is to assume that this means that every adventurer has a strong core education in every skill. Occasionally this will still throw unreal results (like the 8-strength halfling rolling natural 20 and dead-lifting a 1000-pound portcullis), but for the most part it makes some sense. Or you can assume that every adventurer has a bare minimum education in every skill (~50% chance of making 'Easy' checks), and adventurers being what they are, never gain 'Sage' or 'Physicist' or 'Astronomer' level education in such things. Adventurers are high experience, low book study characters. Even Wizards. They learned to do the things they can do hands on. As such they get all those juicy class features, but their ability to remember esoteric book learning off the top of their head is low, or smith the perfect sword, or compose the perfect balad. Even then, the best of the best can accomplish the impossible 10% of the time. And the ones that were trained Sages before adventuring get the Research background feature on top of that. (All extrapolation obviously. Take as you will.)

Edit: Or you can NOT assume that, and make some house-rule changes. ;)

Edit2: I actually assume (as I said early) exactly what you do in the post I quoted. That adventurers are all skilled at what they do, which includes Arcana, History, Religion and Nature lore as appropriate to adventuring. That's because I consider 50% chance of doing something you're not particularly naturally talented or proficient at, but carries risk (ie requires a check at all), pretty damn good.

[mgshamster]

This is either a result of bad DMing, or of characters seeking out more challenging things (which is appropriate). The trouble is when the dungeons aren't really any more well-designed and owned/built by things of the appropriate CR; then it becomes unbelievable that the secret door and trap are all such high DC.

In 3e and 5e both, the DCs are supposed to be set by the world. The difference is that the skill and challenge of the things the players are facing should, independent of the PCs' level, be stronger. The PCs' level lets them face those challenges.

A 3e PC going back to town and finding some 1st-level kobold has trapped his door should have no trouble with that search DC; the kobold's trap has a very low hide check result compared to the PC's high-level bonus. A 5e PC will never encounter a trap harder than a certain DC, however, because of bounded accuracy. At least, in theory. So his lower bonuses only make him more likely to succeed, not ABLE to succeed.

I can see that, and I'm inclined to agree. However, "level appropriate" chellenged were so systemic in 3.X that entire published adventures were based around it and people got pissed if you would dare throw things that were too high of a CR, because they couldn't make a judgment for whether to fight, run, or something else. I've been in many an argument trying to convince people to make more realistic worlds/campaigns by having any number of CR creatures appear regardless of the PC's - based on the natural conditions of the world.

Here's an older article talking about this exact issue: http://thealexandrian.net/wordpress/2050/roleplaying-games/revisiting-encounter-design

[Tanarii]

I can see that, and I'm inclined to agree. However, "level appropriate" chellenged were so systemic in 3.X that entire published adventures were based around it and people got pissed if you would dare throw things that were too high of a CR, because they couldn't make a judgment for whether to fight, run, or something else.In 4e, it was baked right in to the system. Everything was a level appropriate challenge.

[Doug Lampert]

In 4e, it was baked right in to the system. Everything was a level appropriate challenge.

Nope, every challenge had a level. There was nothing in the rules saying you only met challenges of your level, in fact, the rules allowed for quite a wide range and even advised that exceptionally tough challenges might count double toward recovery of action points while easy ones might not count at all.

There were even fixed DCs for various environments in the DMG, a wooden door didn't get any tougher because you were level 30 than it was at level 1.

Some people got this wrong, but then that's true of 3.x and fifth too.

[Serket]

I can see that, and I'm inclined to agree. However, "level appropriate" chellenged were so systemic in 3.X that entire published adventures were based around it and people got pissed if you would dare throw things that were too high of a CR, because they couldn't make a judgment for whether to fight, run, or something else.

The "judgement" part is important. If there's stuff that will murder you if you provoke it, then you have to be reasonably able to determine that it's a threat. If it looks like a rabbit, and moves like a rabbit, and eats like a rabbit, but then when you try and make rabbit stew it polymorphs itself into a dragon and eats you... well then, why are you playing that game? Hard is challenges, punishing is dying without any rationale, and the difference between the two is important.

Personally I love having a fleshed out world in which some things are just too tough to tangle with. :smallsmile: It suits my playstyle, and encourages everyone not to resort to combat unless necessary. I'm really bored with the notion that everything is there for us to kill.

[newsman77]

That's a Perception or Investigation check, though, not a Nature check. It's the experts who have to identify the comet (Nature), even if the student is the one who sees it in the telescope (Perception) or picks out an unexpected spike on a chart (Investigation). Your point here actually underscores an issue: certain tasks (like spotting something anomalous) can be done by laypeople, but other tasks (like identifying what the anomaly means) are the exclusive domain of the well-trained, if not necessarily the expert.

The only way to apply this sensibly is to assume that this means that every adventurer has a strong core education in every skill. Occasionally this will still throw unreal results (like the 8-strength halfling rolling natural 20 and dead-lifting a 1000-pound portcullis), but for the most part it makes some sense.

This runs into problems with the many people who want to play uneducated, unfit, or unobservant people. Perhaps a good houserule would be to allow characters to buy "antiproficiency" in a skill, which restricts them to their Passive Skill, in exchange for a minor benefit elsewhere, like a tool proficiency, language, or hit point.

Incorrect sir. Even the experts get things wrong. I reject your notion that only an expert or someone well trained can know a thing or identify a anomaly with a skill check. That's just not logical.

Your example of a halfing dead lifting 1000 pound portcullis does not hold water because that won't, or shouldn't happen, even on a roll of 20. It's not logical. Sure, go ahead and roll the dice, but it's clearly out of your strength range so your dice roll means nothing.

Every adventurer has a basic to moderate training in each skill. They get that from adventuring, as represented by the increasing proficiency bonus 5e gives you to each skill & bonus you get from your stats. It makes sense. The designers went through a lot of play testing with this. I believe they've turned out a solid system that takes a lot of focus off the mechanical and puts it back on having fun and storytelling.

Edit: I agree with Tanarri: "That adventurers are all skilled at what they do, which includes Arcana, History, Religion and Nature lore as appropriate to adventuring. That's because I consider 50% chance of doing something you're not particularly naturally talented or proficient at, but carries risk (ie requires a check at all), pretty damn good. "

[Spacehamster]

If I changed one thing with skills I would let the player pick one of their skills as their specialty and for that skill they get a +2 in addition to their ability and proficiency, it would be doubled by expertise as to not step on rogues and bards tippy-toes. :)

[Tanarii]

If I changed one thing with skills I would let the player pick one of their skills as their specialty and for that skill they get a +2 in addition to their ability and proficiency, it would be doubled by expertise as to not step on rogues and bards tippy-toes. :)So, with Proficiency, Expertise, and Specialty, you could have from +8 (1st level) to +20 (17th level) bonus? With an additional -1 to +5 from ability score?

[Spacehamster]

So, with Proficiency, Expertise, and Specialty, you could have from +8 (1st level) to +20 (17th level) bonus? With an additional -1 to +5 from ability score?

Yeah all though I as a player and rogue would prob choose to put my +2 in another skill as to get one more skill I'm really good at. :)

[joaber]

Please don't translate the ability of characters in probability to do something, it is instinctive, but in the end is stupid, because is all about game balance and this is a game, not reallity. If you do that, you start to question things like why an amoeba ha 10% of the intelligence of an human? If the max you can lift is 30lbs for each point in Str, Benedikt Magnusson (raw deadlift world record) has 34 of strenght? More than Big T. An Obese old human can run as fast as Usain Bolt since they are from same race...

How you correct that? well, now a normal human will have 10 of Int, an amoeba will have 0.00000000001, the strongest man have 34 of strenght, Big T has 293850... will be an amazingly complex and boring game, congrats.

is just about the game balance.
If in your campaing you think someone need to know something, give advantage for him, if you think is to hard to the other know, give desadvantage for that. Now will be preatty hard to the barbarian know more about arcana in this check than the wizard, but still is possible.

If you think someone need to know something, don't as for a check. If it is impossible for the other, don't let he make the check.

Just roleplay instead of worry about the math, otherwise, is better looking for play something different than D&D 5e.

[Talakeal]

This thread may be the strongest argument for the existence of the Oberoni fallacy that I have ever seen.

Also, I don't think not letting players roll skill tests at all is the proper solution. I imagine the barbarian's player would be rather peeved if the DM wouldnt allow him to even attempt knowledge checks just so the wizard could feel like a special snowflake.

Being completely unable to attempt things without the requisite feat, class ability, or skill level was one of 3Es biggest weaknesses, bringing it into 5E does not make the game better.

[Tanarii]

This thread may be the strongest argument for the existence of the Oberoni fallacy that I have ever seen.Nah. Non-weapon proficiency beat 5e skills (or 4e skill challenges) hands down for excessive DM-fiat in the minds of people desiring lots of rules complexity. Which is what the Oberoni Fallacy is *really* about. Or at least, how it is most commonly mis-applied. Hell, before NWP it was just DM decides with no rules or guidelines at all beyond "my character has a background as a fisherman".

[Knaight]

The so-called "failures" of 5e is a system built in reaction to the failure and commonly complained about problems in 3.5e/pathfinder.

All the things that are being complained about were SPECIFICALLY done to move away from the things the posters in this thread want. Because the things these posters want created too many problems for the majority of players.
...
The 5e skill system isn't a failure. It's a reaction to the failures of previous editions, for the majority of players. If it doesn't meet your needs, it's house-rule time. Which is awesome. Do it. Some good suggestions were made in this thread already. Including having differing levels of what is "Easy" or doesn't require a check for any given PC, based on the character. (ie the Barbarian/Wizard Arcana discussion.)
I'd argue that it's an overreaction to the failures of previous editions. 3.x could get absolutely ridiculous amounts of skill variance, and when there's a 60 point gap between an expert and a novice on a d20, there's a very real question as to why you bother rolling. That doesn't mean dropping it down to less than 10 points for most of the game is necessarily a good idea. There's room in between.


Which is why we have 5e skill system instead of 3.5's. It's based on huge amounts of playtesting and surveys. If the 'vocal portion' had it's way, we'd still have 3.5 semi-simulation skill system going on.
It's not like there's only two systems, one of which must be chosen. WotC had every indication that people disliked the 3.5 skill system and moved away from it; I'm fine with that. I disliked the 3.5 skill system, from the way points per level worked (particularly with intelligence bonuses and making higher level starting characters/NPCs), to cross class skills (good riddance to that) to the weird ranges, to feat hunting for skill bonuses, to the particular skills chosen (collapsing Jump, Climb, and Swim into Athletics is something I'm completely on board with), to every skill having a DC table of its own, and with a second shot fired at cross class skills. Seeing that purged was beautiful.

That doesn't mean that 5e can't have problems with the skill system, and it doesn't even mean that it can't have some of the same problems. It just means that some of the problems are long, long gone.

[mgshamster]

This thread may be the strongest argument for the existence of the Oberoni fallacy that I have ever seen.



Wait. So when using the most customizable edition of D&D published to date, where the design philosophy is for people to customize the rules to their taste, it's a fallacy to say that in order to get what you want, you just have to use one of the customization options or customize your own option?

That just seems off.

[Tanarii]

That doesn't mean that 5e can't have problems with the skill system, and it doesn't even mean that it can't have some of the same problems. It just means that some of the problems are long, long gone.First of all, the 5e skill system has many similarities with 4e. So they had an entire edition's worth of response to it. And they actually reduced bonuses from 4e. Not only that, my understanding is 5e is the most heavily play tested and surveyed edition prior to release. Maybe I'm wrong on that, but it's certainly the impression I got in the lead-up to release. AND they were SPECIFICALLY trying to draw 3e/Pathfinder players back in to the fold.

All of that together, the changes made and not made from 4e --> 5e skills, in combination with the heavy surveying/playtesting and attempting to draw in 3e players, indicates to me that those thinking the 5e skill system's variance in bonuses and flexible DCs are a "problem" are in the minority of players ...

... at least at the time of release. It's possible a large number of players have changed their opinions over time, of course. ;) But I find it far more likely the "vocal minority" are those complaining about the 5e system, because something razor something explanation.

OTOH I did try to actually contribute to such folks desire for something different with a slew of possible house-rules a page (few pages) back in this thread. Because I don't just want to be a edition defender.

[mephnick]

indicates to me that those thinking the 5e skill system's variance in bonuses and flexible DCs are a "problem" are in the minority of players ...

Honestly, those that actually think about these things instead of just hanging out with their friends and playing whatever game is on the table is in the minority of players. I think 8/9 people I play games with would have no idea what any of this conversation even means.

[UberMagus]

People repeatedly cite how "play-tested" 5e was, then others cite, over and over, that with a GOOD DM, the system works, but that not all DMs are good.

Obviously, the play-tests had GOOD DMs, and thus this wasn't a problem. The issue now, is that not all DMs are good, and even the ones that WILL be, don't start that way.

To quote Syndrome "And when everyone's Super, no one will be!"
That's the problem we're hitting in 5e: The lack of the exceptional, the heroic, in favor of the average. The "everyone gets a shot at winning" mentality, which just means that player that built their character to be awesome at that one thing(see the "Cook" above) is no longer really exceptional, simply slight above average.

You won't have the crazy awesome lutist, when the average joe with a high Cha can do as well; you won't have the History nut, when jimmy street-rat can roll just as high.

I like the idea of bringing back "trained and untrained". Let everyone attempt trained checks, but say they can't do better than "medium"(DC 15). You might get lucky on that History roll, but you'll never have the chance to remember the advisor to the third emperor of the second dynasty of Ching. :p

[Grod_The_Giant]

Wait. So when using the most customizable edition of D&D published to date, where the design philosophy is for people to customize the rules to their taste, it's a fallacy to say that in order to get what you want, you just have to use one of the customization options or customize your own option?

That just seems off.
If they had an variant for larger bonuses, that would be fine. If they had a variant for different die types, or minimum rolls, or interesting auto-success rules, that would be fine. But they don't, and "make it up yourself" never counts as a game being customizable. I can write new rules for 3.5 too. I can write new rules for any game; 5e isn't special in that regard.


Honestly, those that actually think about these things instead of just hanging out with their friends and playing whatever game is on the table is in the minority of players. I think 8/9 people I play games with would have no idea what any of this conversation even means.
Counterargument: my girlfriend, who's just starting to get into RPGs, still gets somewhat confused by heavy rulesets and couldn't stand Pathfinder, voiced pretty much the exact same criticisms. But it's also irrelevant; I think about these things, and I want to provide my friends with the best possible gaming experience I can.

[mephnick]

I like the idea of bringing back "trained and untrained". Let everyone attempt trained checks, but say they can't do better than "medium"(DC 15). You might get lucky on that History roll, but you'll never have the chance to remember the advisor to the third emperor of the second dynasty of Ching. :p

That's pretty much what I do, but just for knowledges. I allow everyone to roll, but non-proficient are limited to most common knowledge. I still use monster knowledge for abilities and weaknesses etc though.


But it's also irrelevant

It's relevant in that people seem to be indicating that it's a failure of the system that is either driving people away or should be a main concern for WotC to fix. Which it is neither. It's driving away a minuscule percentage of potential clients who likely would have never made the full switch anyway (3.5 fanatics) and embraced or outright missed by many more.

[Tanarii]

I'm fairly sure this thread as a whole isn't ever going to change back to the topic of actual suggestions for house-rules on how to deal with small bonuses, but I'm interested in hearing suggestions, or hearing critique of my suggestions up-thread. Especially feedback from the OP.

To post again:
Based on keeping DCs the same (DC 5 very easy to DC 30 impossible):
1) allow Rogues (and maybe Bards) to apply Expertise to 'roguish' Tools. Thieves Tools, Disguise Kit, Forgery Kit, Poisoners Kit.
2) same as above but to any Tool.
3) allow any character to use the Training rules for Tools (250 days and 250gp) to get double proficiency bonus with a Tool they are already proficient in. Suggest excluding 'roguish' tools.
4) same as above but also allow combining the same tool proficiency from 2 different sources (Race/Class/Background) to get double proficiency starting at level 1. Still suggest excluding roguish tools.

Change the difficulty level slightly, let's say by adding a new category at the top, DC 35 (miraculous):
5) double proficiency bonus (+4 to +12) for skills. (Note you may want to exclude some rules that use skills, like grappling rules.) Expertise add 1/2 new bonus, and Jack-of-All-Trades & Champion version of it add 1/4 to non-proficient.
6) allow combining the same skill from two different sources (race, class, background) to get double proficiency bonus. Expertise and other things that add double bonus instead add triple (total) if used with an already doubled skill.

7) Change DCs & AC to be double the current range above 10, and double all bonuses and penalties from all sources. So Leather plus Dex 16 is AC 18, Full Plate and shield is AC 30, a first level typically has +8-10 to hit, a max level one +22 to hit. Skill DCs range from 5 (very easy) to 10 (easy) to 30 (hard) to 50 (impossible). Skill modifiers range from -2 (ability score 8, no proficiency) to +22 (Max prof & ability score) to +34 (expertise + max).
8) same as 7, but only applies to ability/skill checks, not attack rolls or saves. May require tweaking for certain combat related use of skills such as grappling.

[mgshamster]

I'm fairly sure this thread as a whole isn't ever going to change back to the topic of actual suggestions for house-rules on how to deal with small bonuses, but I'm interested in hearing suggestions, or hearing critique of my suggestions up-thread. Especially feedback from the OP.

In the DMG, there's an option for proficiency dice:

+2 becomes 1d4
+3 becomes 1d6
+4 becomes 1d8
+5 becomes 1d10
+6 becomes 1d12

The DMG says this makes proficiency more random and less reliable. What if we took it one step further and just maximized those proficiency dice?

Now there's a split of 8 points between the low level and high level character. Keep the general rule for DCs the same, and you have high level characters making near impossible checks fairly regularly.

... And I just realized that this is the same as your "double the proficiency bonus" suggestion.

[Xetheral]

Not really. Cooking an awesome dinner for an NPC being a major plot point in an adventurer's life, to the point where it's important for them to be an expert chef compared to other adventurers, comes up how often?

Characters use the tools they have available. If a character is skilled in cooking, they're going to try to find ways to use that skill to solve the problems at hand. So the answer to your question is that cooking will likely come up in an important fashion every time someone makes a character with a cooking specialty.


Also, generally speaking, if it's possible to spend precious character creation/development resources to become an "true expert" in such a thing, at the cost of other more important ones, that's considered a trap option and a flaw in system rules. Specifically talking about weird corner non-typical-for-adventurers skills here. Underwater basket-weaving.

I draw an important distinction between an option that is primarily for non-combat use and a "trap" option. The latter is something that looks mechanically useful but really isn't. The former is just a way to add depth and breadth to a character.


I assume it's fair to say that a PC adventurer who is a true expert in a specific adventuring task compared to a different PC adventurer who is less expert, but has the basic skill level necessary to be an adventurer at that task, has a difference of no more than 60% chance of success. Or adventurer's opponents. So Yes. We have different views on what it means to be a true expert among adventurers, at adventuring type tasks. It's like comparing two NBA players ... they're both fairly amazing no matter which one is the true expert.

You keep emphasizing that the PCs are adventurers as if that somehow necessarily makes them qualitatively different from the other inhabitants of the game world in a way that should affect the interpretation of skills. That's a common playstyle, but it's not a universal one: plenty of tables eschew the "professional adventurer" concept in favor of ordinary yet exceptionally-skilled people who, when necessity calls, happen to have the skills that make them the right person in the right place at the right time. I've run campaigns in both styles (and many others), so I can't subscribe to an interpretation of the skill system that only works for one of them.


Then you're making different assumptions from 5e, which is fair enough.

Yeah, I'm going to have to disagree again. I don't think your playstyle is more privileged than any other when it comes to the default assumptions of 5e, particularly for an edition meant to have broad appeal.


But the other things to consider is : if you spend your character resources developing those skills to become an expert at those non-adventuring tasks, what are you sacrificing to gain it? Only 3e really set up a system to do it that way, and both 4e and 5e moved away from that due negative reactions to the problems it caused.

Ideally, I would prefer it if out-of-combat flavor didn't use up character building resources that could be applied to combat. Simply curtailing mechanical support for out-of-combat flavor, as 5e had done, certainly solves that problem, but for my purposes the cure is worse than the disease.


Sounds to me like instead of the DM judging something an automatic success or automatic failure, you only want it when the dice decide it. That's a fair point. I personally like that D&D 5e doesn't make it possible for a character to auto-succeed on any adventuring task, or conversely auto-fail on a adventuring task. Or waste character development resources trying to keep up with the treadmill for contest skills (perception vs stealth especially).

Again, I don't see the need for a divide between "adventuring" and "non-adventuring" tasks. Isn't everything the adventurers do, by definition, part of the adventure? And no, I'm perfectly fine with ruling something to be an automatic success or failure even if the bonuses aren't high or low enough for that to be mandated by the dice. I'm also perfectly fine if the bonuses and penalties are so high or low that under the that mechanics the outcome is predetermined.


Or you can assume that every adventurer has a bare minimum education in every skill (~50% chance of making 'Easy' checks), and adventurers being what they are, never gain 'Sage' or 'Physicist' or 'Astronomer' level education in such things. Adventurers are high experience, low book study characters. Even Wizards. They learned to do the things they can do hands on. As such they get all those juicy class features, but their ability to remember esoteric book learning off the top of their head is low, or smith the perfect sword, or compose the perfect balad. Even then, the best of the best can accomplish the impossible 10% of the time. And the ones that were trained Sages before adventuring get the Research background feature on top of that. (All extrapolation obviously. Take as you will.)

Edit: Or you can NOT assume that, and make some house-rule changes. ;)

(Emphasis added.) Yeah, I'm not going to assume that. It removes a huge amount of flexibility in character design.

----------


Question: Do "level appropriate" challenges happen in 5e?

In 3.X, you'd change the DC to match the PC's skill level. The higher level you are, the higher the DC is to accomplish something extraordinary. (This isn't my choice; I've just seen this happen a thousand times or more in all the years I played 3.X, across many different tables and many different published adventures).

I've never played 3.5 this way, and would argue that it isn't part of the system. Sure, lots of people did play that way, but it wasn't because the game told them to. That Alexandrian article does a good job of explaining how the rules for 3.5 instead suggest you shouldn't have all challenges be level-appropriate, at least in terms of encounter design.

----------


In my opinion, it's not a matter of it being okay, just a matter of recognizing and adapting. This edition is fairly modular and flexible and can take a lot more change than 3.5 could without breaking.

The problem, as was stated upthread, is that implementing this system as the norm encourages the same "only roll if you bothered to max your skill ranks" paradigm that 3.5 has. Also, 3.5 has a remarkably similar skill system to 5e with the exception of the point system and slightly flattened numbers, rolling high technically did nothing. You were still supposed to role against a DC or make an opposed check.

So since you ruled in that the roll was just a measure of extremity anyway, why not add something similar to this system? It's a little more work because without the point scaling baked into the system it doesn't happen naturally, but you can (for example) have a cook who has proficiency in cooking tools and went ahead and trained for a further 2500 days (and a further 2500 gold) and they get to add +10 to their roll (still at a max of 30). Maybe add a clause about it not affecting Thieves Tools (unless you want a Rogue to auto-pass the DC, which now that I think about it, you might!) and let the specially trained characters rock those checks. Or instead of a static increase, say that for each extra 2500 days and 2500 gp, they can't roll below a 5, then a 10, then a 15. So for the low, low price of 21 years and 7750gp, you literally cannot roll below a 15 on your cooking check. A true master in the field!

I personally never ran into that problem, even in 3.5, because I don't set DCs so high that low bonuses are worthless. Accordingly, I don't believe that problem is a fundamental result of the system permitting high skill bonuses.

Furthermore, I disagree with your interpretation of 3.5 skill mechanics. Some individual skills even had variable results based on your die roll (jump, diplomacy, etc.) right in the PHB, and there were plenty of other mechanics with differing effects depending on how much you beat a skill DC by.

It hadn't occurred to me that 5e had taken out all of the "success level depends on the degree by which the DC was passed" examples, but I see nothing in the bare-bones skill rules indicating that such an interpretation is precluded by default. So yes, I already run 5e in exactly the same way as far as degree of success (or failure) goes, in those circumstances where the outcome isn't inherently binary. My games tend not to have much in the way of downtime, so your specific suggestion probably won't see much use, but I'm certainly considering adopting some of the radical ideas that rescale the skill bonuses.

[Tanarii]

... And I just realized that this is the same as your "double the proficiency bonus" suggestion.Yeah. The other option (to make the impossible not quite so common) is just make that DC 35 necessary for impossible. It's going to be DM judgement anyway as to what's what DC.

That's something that can't be eliminated from the system, as I said in a different comment. At best, we could come up with a communal list of suggested DCs for suggested tasks, but that'd still be OUR communal DM-fiat. So to speak. But at least players could have it in advance if that's what makes them more comfortable about the 'fog-of-war' in regards to their own capabilities. Or DM consistency internally at a given table.

[mgshamster]

That Alexandrian article does a good job of explaining how the rules for 3.5 instead suggest you shouldn't have all challenges be level-appropriate, at least in terms of encounter design.

True. But you'll also notice that while he says it shouldn't be played that way, it also currently is played that way. It's so systemic in the community that published adventures do it. That should be evidence enough that it's become part of the adventure design philosophy.

I've been in many arguments over the years trying to convince people to stop making everything level appropriate - many times failing.

I'm actually really happy that people here are arguing against me, because that suggests a change in the community's thinking for this. And I've noticed a distinct difference in 5e's published adventures. More evidence that we're moving away from it.

I mean, this entire train of thought was become someone said this in this very thread:


A trained professional, in my opinion, should have about a 70% success chance on level-appropriate, general-difficulty checks; that's enough to feel like a pro but not so much that there's not a chance of failure.

That right there shows that there is still the idea that people want the challenge to change based on the level; there's an idea that a general-difficulty is different for a high level character vs a low level character, instead of a general-difficulty being based on the world. The difficultly is "level appropriate" so-to-speak. It got me thinking about the Alexandrian article and all the debates I've been in over making challenges consistent to the world rather than consistent to the PC level.

If it were static to the world, then a high level PC would not be going for the general-difficulty, they'd be going for the hard-difficulty or even the near impossible difficulty.

[Grod_The_Giant]

It's relevant in that people seem to be indicating that it's a failure of the system that is either driving people away or should be a main concern for WotC to fix. Which it is neither. It's driving away a minuscule percentage of potential clients who likely would have never made the full switch anyway (3.5 fanatics) and embraced or outright missed by many more.
And? If you look back at the original purpose of this thread, it wasn't to convince people that 5e is badwrongfun; I was hoping for help with some houserules to make the game more fun for my friends and I. I said "this is a problem for me;" you said "most people don't have this problem." As a general point it's not without merit, but for the purposes of this thread?


That right there shows that there is still the idea that people want the challenge to change based on the level; there's an idea that a general-difficulty is different for a high level character vs a low level character, instead of a general-difficulty being based on the world. The difficultly is "level appropriate" so-to-speak. It got me thinking about the Alexandrian article and all the debates I've been in over making challenges consistent to the world rather than consistent to the PC level.
That got interpreted wrong. I didn't mean "level-appropriate" in a mathematical sense, not like 4e's implicit suggestion that "if your PCs are level X, their bonus is probably Y so DCs should by 10+Y." That's goofy. "Level-appropriate challenge," to me, refers to the broader idea that you're not doing the same thing at level 20 as you are at level 1. Take Climb, for instance:

At level 1, a skilled climber should be able to scale a rough wall with lots of projections and ledges. (DC 15)
At level 5, a skilled climber should be able to scale a blank brick wall. (DC 20)
At level 10, a skilled climber should be able to scale a smooth stone castle wall (The kind of thing where you can barely fit a blade between the blocks). (DC 25)
At level 15, a skilled climber should be able to scale a glass wall. (DC 30)
At level 20, a skilled climber should be able to scale a wall made of flowing quicksilver (DC 35)

So I mean, yes, the DC for level-appropriate tasks should increase with level, but less in a "you should always have an x% chance of success" sense than in a "you have the skills to attempt progressively more difficult challenges." Those DCs correspond to the challenge, but they're static. That level 10 climber can run up the rough wall like it's a sidewalk, climb the brick wall without really thinking about it, and potentially--with plenty of luck and maybe some outside help--make it up the glass wall. Adventures should have level-appropriate challenges, both so they can actually challenge players and so players have a chance to show off their mad skills, but there should be parts that are (almost certainly) too hard, and parts where a skilled character can easily bypass things too-- just because you can hit those DC 30 climb checks doesn't mean you stop running into brick and stone walls.

Does that make sense? In combat you can defeat progressively scarier and more earth-shaking monsters; noncombat challenges should scale in the same way.


I'm fairly sure this thread as a whole isn't ever going to change back to the topic of actual suggestions for house-rules on how to deal with small bonuses, but I'm interested in hearing suggestions, or hearing critique of my suggestions up-thread. Especially feedback from the OP.
Agreed. We've gotten waaay off topic. As is inevitable, but still. We're arguing in circles; I don't think anyone who's not already convinced is going to change their mind. There's another thread to take those arguments to. As the OP, I request that we stick to discussing useable houserules. If you don't think there's a problem, saying so isn't going to help me.

Anyway...


1) allow Rogues (and maybe Bards) to apply Expertise to 'roguish' Tools. Thieves Tools, Disguise Kit, Forgery Kit, Poisoners Kit.
2) same as above but to any Tool.
Aren't these the default rules?

3) allow any character to use the Training rules for Tools (250 days and 250gp) to get double proficiency bonus with a Tool they are already proficient in. Suggest excluding 'roguish' tools.
That's not a terrible idea, especially if you expand the list to a more general NWP-type deal, more like 3e's catch-all Crafting and Profession categories. (Proficiency in cooking? Research? Farming? Sure!)


4) same as above but also allow combining the same tool proficiency from 2 different sources (Race/Class/Background) to get double proficiency starting at level 1. Still suggest excluding roguish tools.
...
6) allow combining the same skill from two different sources (race, class, background) to get double proficiency bonus. Expertise and other things that add double bonus instead add triple (total) if used with an already doubled skill.
An interesting idea, but I don't think the best one. I do prefer broader characters, and 5e already feels kind of stretched for most classes. This just encourages narrower focuses, to say nothing of more stereotypical backgrounds.


Change the difficulty level slightly, let's say by adding a new category at the top, DC 35 (miraculous):
5) double proficiency bonus (+4 to +12) for skills. (Note you may want to exclude some rules that use skills, like grappling rules.) Expertise add 1/2 new bonus, and Jack-of-All-Trades & Champion version of it add 1/4 to non-proficient.
Yeah, I think this may be the best bet overall. Or something along those lines; instead of just doubling what's written, how about a new rule of Proficiency Bonus=1/2 level (round down)+3? You'd start at +3 and get another +1 at every even level, up to +13 at 20th-- a similar range, but with a much smoother curve throughout. Expertise could upgrade that to level+3.

It works out pretty well for non-specialists, I think. +5 vs +18 isn't beyond the bounds of probability, and even a +0 has a chance-- which is about all I'd ever ask. An Expertise-wielding character with a +28 would wind up beyond the bounds of the nonproficient, but not the proficient, which again I'm quite okay with, and honestly seems appropriate.

The problem, as you note, is that Proficiency is also involved in combat skills-- while I'm okay with specialists' having a significant advantage on skills, doing the same with combat is worrysome. Attack/AC balance is probably fine; everyone will be proficient with their attack(s), so doing something like adding half your Proficiency bonus to AC should that right back up. (And also makes sense; I never really like attacks that scale with levels and AC that only scales with items).

...And come to think of it, I bet you could do the same thing with other defenses. Have a universal "Defense Bonus" equal to half your Proficiency Bonus and add it to AC, all saves, and to resisting opposed skill checks like Shoves, and just note that it gets overwritten by normal Proficiency if it applies. Possibly a bit clunky, and I admit that it'll make lower-level foes a bit less threatening, but it should be generally decent--especially if you use Pack Tactics liberally, as would make sense for low-level hordes. It would be worth checking the math at a few points, but it seems like it might be workable...


7) Change DCs & AC to be double the current range above 10, and double all bonuses and penalties from all sources. So Leather plus Dex 16 is AC 18, Full Plate and shield is AC 30, a first level typically has +8-10 to hit, a max level one +22 to hit. Skill DCs range from 5 (very easy) to 10 (easy) to 30 (hard) to 50 (impossible). Skill modifiers range from -2 (ability score 8, no proficiency) to +22 (Max prof & ability score) to +34 (expertise + max).
8) same as 7, but only applies to ability/skill checks, not attack rolls or saves. May require tweaking for certain combat related use of skills such as grappling.
Hmm... seems simple on the surface, but I feel like there's something I'm missing probability-wise.

[Tanarii]

Agreed. We've gotten waaay off topic. As is inevitable, but still. We're arguing in circles; I don't think anyone who's not already convinced is going to change their mind.You didn't change my mind, but you convinced me to stop arguing in circles and look at house-rules. At least, in this thread, since that's what it was specifically for. (The other thread on the same topic seems to ask a more open ended "what do you think" question. ;) At least two threads on the same thing can be for different purposes lol

Including which # you were responding to for clarity.

(Your response to #1 & #2)

Aren't these the default rules?No definitely not. Unless I missed an Errata. Of the Tools, Expertise can only be applied to Thieves' Tools, and only by Rogues. Honestly, allowing Rogues and Bards to apply it to any Tool is probably a worthwhile general house rule, outside of this particular discussion.

(Your response to #3)

That's not a terrible idea, especially if you expand the list to a more general NWP-type deal, more like 3e's catch-all Crafting and Profession categories. (Proficiency in cooking? Research? Farming? Sure!)Yeah. I do like those kinds of characters personally. One thing about basing things on the training rules to keep in mind is 250 days is highly campaign dependent in terms of opportunity cost. Some campaigns it might be giving up 10 levels! Effectively meaning your character has retired from adventuring, at least with that party. Others it could be all PCs taking a break at the same time, pick up a new Tool/Language, back to the game. So you'd know how possible it would be for characters in your games to be able to actually do that.

(Your response to #4 & #6)

An interesting idea, but I don't think the best one. I do prefer broader characters, and 5e already feels kind of stretched for most classes. This just encourages narrower focuses, to say nothing of more stereotypical backgrounds.Allowing double-stacking of proficiency has the advantage of requiring a cost to get a higher bonus. In other words, characters have to invest something in it. They have to be focused to become experts, and be lacking elsewhere. I agree that broader characters are more interesting, but you need to consider what goes into being an "expert" in your mind. If Expert to you just means "proficiency at level one" + "gain levels", then other options work better. If it means "focus on something to become more skilled, at the cost of other things" then options requiring doubling up on proficiency at the cost of less proficiency are better.

(Your response to #6)

Yeah, I think this may be the best bet overall. Or something along those lines; instead of just doubling what's written, how about a new rule of Proficiency Bonus=1/2 level (round down)+3? You'd start at +3 and get another +1 at every even level, up to +13 at 20th-- a similar range, but with a much smoother curve throughout. Expertise could upgrade that to level+3.

It works out pretty well for non-specialists, I think. +5 vs +18 isn't beyond the bounds of probability, and even a +0 has a chance-- which is about all I'd ever ask. An Expertise-wielding character with a +28 would wind up beyond the bounds of the nonproficient, but not the proficient, which again I'm quite okay with, and honestly seems appropriate.

The problem, as you note, is that Proficiency is also involved in combat skills-- while I'm okay with specialists' having a significant advantage on skills, doing the same with combat is worrysome. Attack/AC balance is probably fine; everyone will be proficient with their attack(s), so doing something like adding half your Proficiency bonus to AC should that right back up. (And also makes sense; I never really like attacks that scale with levels and AC that only scales with items).

...And come to think of it, I bet you could do the same thing with other defenses. Have a universal "Defense Bonus" equal to half your Proficiency Bonus and add it to AC, all saves, and to resisting opposed skill checks like Shoves, and just note that it gets overwritten by normal Proficiency if it applies. Possibly a bit clunky, and I admit that it'll make lower-level foes a bit less threatening, but it should be generally decent--especially if you use Pack Tactics liberally, as would make sense for low-level hordes. It would be worth checking the math at a few points, but it seems like it might be workable...I like smoothing out the numbers. I'm with you on avoiding affecting combat numbers. As to jacking up defenses with level with level, yes, that's a fantastic solution to countering increasing attack bonuses with level. It's one D&D 5e didn't use on purpose for a specific reason: To make sure lower level mooks were still a threat. Your increasing HPs are your defense against a bonus increasing at a slower rate the way things are. But if you increase the attack bonus much, you probably need to compensate somehow on the defensive side. Just be aware you're changing something that was intentionally done for a specific reason, and the result will be making lower level creatures less dangerous for CR than they currently are to higher level characters. (And vice versa, since I assume you'll increase creature defenses similarly.)

(Side note: 4e enabled keeping mooks dangerous a different way ... minions. Level 2 Orcs (elite) could be level 5 Orcs (Standard) could be level 10 Orcs (minions). Defenses scaled, hit points damage etc dropped. Not everyone was satisfied with the verisimilitude of that solution .)

(Your response to #7 & #8)

Hmm... seems simple on the surface, but I feel like there's something I'm missing probability-wise.
Lol yeah that's exactly what I was thinking as I typed it up. It definitely will suffer from the low-level can't touch high-level problem. As well as auto-success and can't succeed scenarios. But I think that other than that, it's actually approximately the same chance of success for any given bonus and AC/DC. It just looks wrong at first glance because target numbers are doubled from 10, not straight doubled, to account for the d20 roll. :)

[mgshamster]

That got interpreted wrong. I didn't mean "level-appropriate" in a mathematical sense, not like 4e's implicit suggestion that "if your PCs are level X, their bonus is probably Y so DCs should by 10+Y." That's goofy. "Level-appropriate challenge," to me, refers to the broader idea that you're not doing the same thing at level 20 as you are at level 1. Take Climb, for instance:

At level 1, a skilled climber should be able to scale a rough wall with lots of projections and ledges. (DC 15)
At level 5, a skilled climber should be able to scale a blank brick wall. (DC 20)
At level 10, a skilled climber should be able to scale a smooth stone castle wall (The kind of thing where you can barely fit a blade between the blocks). (DC 25)
At level 15, a skilled climber should be able to scale a glass wall. (DC 30)
At level 20, a skilled climber should be able to scale a wall made of flowing quicksilver (DC 35)

So I mean, yes, the DC for level-appropriate tasks should increase with level, but less in a "you should always have an x% chance of success" sense than in a "you have the skills to attempt progressively more difficult challenges." Those DCs correspond to the challenge, but they're static. That level 10 climber can run up the rough wall like it's a sidewalk, climb the brick wall without really thinking about it, and potentially--with plenty of luck and maybe some outside help--make it up the glass wall. Adventures should have level-appropriate challenges, both so they can actually challenge players and so players have a chance to show off their mad skills, but there should be parts that are (almost certainly) too hard, and parts where a skilled character can easily bypass things too-- just because you can hit those DC 30 climb checks doesn't mean you stop running into brick and stone walls.

Does that make sense? In combat you can defeat progressively scarier and more earth-shaking monsters; noncombat challenges should scale in the same way.

Yes, it does make sense, and it's actually what I envisioned for how I'd run the game.

Thanks for clearing that up.

So what you want is characters to have more than a 50/50 chance to accomplish those tasks at an appropriate level.

From there, we need to differentiate; do we want any given character to be able to accomplish the task, or do we want specialized characters? If so, how much soecialization (proficient vs expertise vs something else)? How much natural talent needs to be in the mix (ie how much influence should the ability score modifier have)?

The answer to those questions will lead us towards what the DC should be, what kind of proficiency bonus is needed, and if we need something between prof and expertise or after expertise.

I'd do the math, but I'm a little drunk right now. And I have game tomorrow. Maybe on Sunday.

[Regulas]

Since you basically just want to exaggerate skill so the DM isn't relevant, isn't the simplest house rule just to double proficiency for skill checks (but not attacks etc)? This alone should suffice, and if you do want anything more extreme maybe tipple the bonus for the skills you chose for your class (but not for background/race or otherwise).

You should start off with around a +7 for your proficient skills, and by level 9 probably have around +11/12 depending on asi's, compared to the +0-5 for anyone not proficient. If it's trippled then starting at +9 going up to +15 around 9.

[Kurald Galain]

That's not a terrible idea, especially if you expand the list to a more general NWP-type deal, more like 3e's catch-all Crafting and Profession categories. (Proficiency in cooking? Research? Farming? Sure!)
The main problem with tool proficiencies are that (1) they don't stack with skills, and (2) spending 250 days to learn something is not an option in most campaigns.

Actually, simply giving (trained) characters a +2 bonus for having good-quality (and likely expensive) equipment available would be a good thing. Note that in some cases (e.g. cooking, research) this may well be more than an adventuring party could reasonably carry, so it'd only work in towns.

[Xetheral]

What about adding skill perks that accrue automatically in skills you are proficient with, based on your proficiency bonus? The idea is to make proficient characters qualitatively better than non-proficient characters without needing to mess with the proficiency bonuses themselves.

The perks all follow the same structure: autosuccess at escalating DCs at proficiency bonus +2, +4, and +6, and thematically-appropriate benefits at +3 and +5 that relate to the reliability of the roll, helping others, or reducing the time required to make a check.

Note that autosuccess provided by Skill Perks does not apply to contested rolls (e.g. grappling).


Insight, Investigation, Perception

Proficiency

Perk Name
Description



+2

Basic Mastery
Autosucceed at DC 5 and below.



+3

Got your Back
So long as you are able to communicate with your nearby allies, they each receive +2 to their passive score in this skill. This bonus is not cumulative. When you take the Help action to assist someone with this skill, they gain an additional +2 bonus on the roll.



+4

Int. Mastery
Autosucceed at DC 10 and below.



+5

Focused
If you roll an unmodified 4 or less on a d20 roll, you may treat that roll as if you'd rolled a 5.



+6

Adv. Mastery
Autosucceed at DC 15 and below.




Animal Handling, Medicine, Survival

Proficiency

Perk Name
Description



+2

Basic Mastery
Autosucceed at DC 5 and below.



+3

Efficient Worker
Any task that requires more than one round to complete may be performed in half the time it would ordinarily require.



+4

Int. Mastery
Autosucceed at DC 10 and below.



+5

Veteran Campaigner
If you roll an unmodified 4 or less on a d20 roll, you may treat that roll as if you'd rolled a 5.



+6

Adv. Mastery
Autosucceed at DC 15 and below.




Arcana, History, Nature, Religion

Proficiency

Perk Name
Description



+2

Basic Mastery
Autosucceed at DC 5 and below.



+3

Efficient Research
Any task that requires more than one round to complete may be performed in half the time it would ordinarily require.



+4

Int. Mastery
Autosucceed at DC 10 and below.



+5

Reliable Recall
If you fail a check to recall knowledge, you may spend a minute doing nothing but thinking and then roll again. This roll is made against a DC 5 higher than the original roll.



+6

Adv. Mastery
Autosucceed at DC 15 and below.




Acrobatics, Athletics, Sleight of Hand, Stealth

Proficiency

Perk Name
Description



+2

Basic Mastery
Autosucceed at DC 5 and below.



+3

Instructor
When you take the Help action to assist someone with this skill, they gain an additional +2 bonus on the roll.



+4

Int. Mastery
Autosucceed at DC 10 and below.



+5

Muscle Memory
If you roll an unmodified 4 or less on a d20 roll, you may treat that roll as if you'd rolled a 5.



+6

Adv. Mastery
Autosucceed at DC 15 and below.




Deception, Intimidation, Performance, Persuasion

Proficiency

Perk Name
Description



+2

Basic Mastery
Autosucceed at DC 5 and below.



+3

Ensemble
When you take the Help action to assist someone with this skill, they gain an additional +2 bonus on the roll.



+4

Int. Mastery
Autosucceed at DC 10 and below.



+5

Cover the Gaffe
If you fail a check, you may immediately reroll. This reroll is made against a DC 5 higher than the original roll, and you must use the second result, even if it's worse.



+6

Adv. Mastery
Autosucceed at DC 15 and below.

[pwykersotz]

What about adding skill perks that accrue automatically in skills you are proficient with, based on your proficiency bonus? The idea is to make proficient characters qualitatively better than non-proficient characters without needing to mess with the proficiency bonuses themselves.

The perks all follow the same structure: autosuccess at escalating DCs at proficiency bonus +2, +4, and +6, and thematically-appropriate benefits at +3 and +5 that relate to the reliability of the roll, helping others, or reducing the time required to make a check.

Note that autosuccess provided by Skill Perks does not apply to contested rolls (e.g. grappling).

Not bad, this is a neat idea.

[Tanarii]

Are those +3/+5 bennies supposed to be specific to the skill? If so, I don't see how the 1-4 = 5 is supposed to be significantly better than auto-succeed on DC 10. i mean, it's totally dependent on having an attribute bonus of +3 or better, or you get no benefit.

The faster fieldcraft / research thing is kind of weird, but I like it.

[Grod_The_Giant]

Rank-based boosts is a cool solution, if not necessarily the easiest or, unfortunately, the most likely to deal with die randomness. Although...

Are those +3/+5 bennies supposed to be specific to the skill? If so, I don't see how the 1-4 = 5 is supposed to be significantly better than auto-succeed on DC 10. i mean, it's totally dependent on having an attribute bonus of +3 or better, or you get no benefit.
I'd actually like to see more of that sort of thing-- it subtly gives more reliable dice rolls. What if you have a universal houserule where your minimum die roll on a skill check is equal to twice your proficiency bonus? At first level rolls of 1-3 are treated as a 4, slowly scaling up. Eliminating the chance of low rolls would go a long way towards making things feel less random, I think.

[Xetheral]

Are those +3/+5 bennies supposed to be specific to the skill? If so, I don't see how the 1-4 = 5 is supposed to be significantly better than auto-succeed on DC 10. i mean, it's totally dependent on having an attribute bonus of +3 or better, or you get no benefit.

The faster fieldcraft / research thing is kind of weird, but I like it.

Yes, the benefits are specific to the skill in question. The 1-4 becomes 5 in some cases will indeed operate similarly to the auto success vs DC 10 in some situations. The ones where it won't include contests (where autosuccess doesn't apply), expertise, guidance, help (with the relevant perks) and, as you mention, situations where the attribute score is above 16.

[JoeJ]

That got interpreted wrong. I didn't mean "level-appropriate" in a mathematical sense, not like 4e's implicit suggestion that "if your PCs are level X, their bonus is probably Y so DCs should by 10+Y." That's goofy. "Level-appropriate challenge," to me, refers to the broader idea that you're not doing the same thing at level 20 as you are at level 1. Take Climb, for instance:

At level 1, a skilled climber should be able to scale a rough wall with lots of projections and ledges. (DC 15)
At level 5, a skilled climber should be able to scale a blank brick wall. (DC 20)
At level 10, a skilled climber should be able to scale a smooth stone castle wall (The kind of thing where you can barely fit a blade between the blocks). (DC 25)
At level 15, a skilled climber should be able to scale a glass wall. (DC 30)
At level 20, a skilled climber should be able to scale a wall made of flowing quicksilver (DC 35)



You set the DCs too high then. If you're standard is that a skilled climber at 1st level should be able to scale a rough wall with lots of projections and ledges about 70% of the time, then then DC should be around 11 (Or 13 if by "skilled" you mean with expertise, not just proficiency). A 20th level character with expertise can hit a DC of 23 about 70% of the time, so if you think they should be that successful at climbing a wall of quicksilver, that's where you set the DC.

The problem you'll run into, though, is that a proficient 1st level character can hit that same DC 15% of the time. That's because 5e was not designed for characters to become demigods or superheroes. It's not expected that even a 20th level character could defeat 1,000 orcs in melee, or persuade Demogorgon to become good, or climb walls that even Spiderman can't.

[Kurald Galain]

I'd actually like to see more of that sort of thing-- it subtly gives more reliable dice rolls. What if you have a universal houserule where your minimum die roll on a skill check is equal to twice your proficiency bonus? At first level rolls of 1-3 are treated as a 4, slowly scaling up. Eliminating the chance of low rolls would go a long way towards making things feel less random, I think.

That's pretty good. One of the big problems in the skill system is trained characters randomly failing at easy tasks, and this does a decent job of mitigating that.

[Grod_The_Giant]

That's pretty good. One of the big problems in the skill system is trained characters randomly failing at easy tasks, and this does a decent job of mitigating that.
Does anyone know how to calculate probabilities using a system like that?

[Kurald Galain]

Does anyone know how to calculate probabilities using a system like that?

Sure. Suppose you're level five, and have a 14 in the relevant ability. Why a 14? That's because most skills simply aren't based on your primary ability score. I'm not going to run examples at level 20 for the simple fact that most campaigns never get to level 20.

By the original system, you have +5 to your skill checks, so you have a 80% chance at an easy check (you fail one time out of five at something you're supposed to be good at), and a 55% chance at an average check, 30% at a hard check (meaning you can't do any of that reliably). Also important to know is that you'll lose a contest to an untrained person 31% of the time (roughly one time out of three).

But now let's have 1-5 count as a 6. This lets you make an easy check automatically, and you still can't reliably make an average or a hard check, as those odds haven't changed. You'll lose against an untrained person 27% of the time (about one time out of four), so that hasn't changed much.

Alternatively, suppose you double your proficiency bonus? Then you'd get a 95% chance at an easy check, 70% at average, 45% at hard; I must say I like this one better. You still lose against an untrained person 20% of the time (about one time out of five).

HTH!

[Grod_The_Giant]

Sure.
Thanks. Sounds like the increased Proficiency bonus (with 1/2 prof to defensive stuff to keep a balance) is the way to go. I can deal with losing a bit of the "low-level creatures are always a threat." Though we can mostly compensate for that by changing the Pack Tactics ability. A flat +5 instead of Advantage ought to do the trick--now those Kobolds are swinging for +10 (+2 Dex, +3 prof, +5 Pack Tactics), enough to still threaten even the most heavily armored knight.


Alternatively, suppose you double your proficiency bonus? Then you'd get a 95% chance at an easy check, 70% at average, 45% at hard; I must say I like this one better. You still lose against an untrained person 20% of the time (about one time out of five).
Agreed.

[Telok]

Thanks. Sounds like the increased Proficiency bonus (with 1/2 prof to defensive stuff to keep a balance) is the way to go. I can deal with losing a bit of the "low-level creatures are always a threat." Though we can mostly compensate for that by changing the Pack Tactics ability. A flat +5 instead of Advantage ought to do the trick--now those Kobolds are swinging for +10 (+2 Dex, +3 prof, +5 Pack Tactics), enough to still threaten even the most heavily armored knight.

Agreed.

Thought: what would happen if you double the prof bonus as applied to everything EXCEPT save DCs AND applied the base prof bonus to AC? Expertise does not double the base bonus, it adds the base bonus to the doubled bonus for a total of x3 bonus not x4.

Attacks versus AC don't change much, both increased by the proficency bonus so lower level stuff is less dangerous and higher level stuff is more dangerous but not by too much.

Skills benefit because the bonus is from 4 to 12 instead of 2 to 6. Expertise goes from 6 to 18 now.

Save DCs can't go up because of how bad nonproficient saves are already, but the good save will become almost automatic successes. Alternately the double prof bonus could not apply to saves, but saves are so bad overall and swingy already that I don't (personally) feel bad about this one.

[Knaight]

It's relevant in that people seem to be indicating that it's a failure of the system that is either driving people away or should be a main concern for WotC to fix. Which it is neither. It's driving away a minuscule percentage of potential clients who likely would have never made the full switch anyway (3.5 fanatics) and embraced or outright missed by many more.

I don't know why everyone keeps insisting that everyone who criticizes 5e is a 3.5 fanatic; it just isn't true. There is exactly one person in me extended gaming group* who I know likes 5e, two more who might possible like it, and about 20 who don't like it. The group that doesn't like it splits between just not particularly liking it and actively disliking it. However, it also contains 0 3.5 fans. The known 5e fan is a 3.5 fan, the two unknown cases are 3.5 fans, and the group that doesn't like 5e also dislikes 3.5. It breaks down differently too - there's a lot less in the way of not particularly liking it, and a lot more in the way of open contempt. There are games other than D&D out there, and it's in comparison to those that a lot of the criticisms of 5e are cropping up. In comparison to 3.5, I'm seeing a great deal more in the way of thinking that it's at least progress.

*As in, the pool of people with whom I might end up in an actual gaming group at any given point

[pwykersotz]

I don't know why everyone keeps insisting that everyone who criticizes 5e is a 3.5 fanatic; it just isn't true. There is exactly one person in me extended gaming group* who I know likes 5e, two more who might possible like it, and about 20 who don't like it. The group that doesn't like it splits between just not particularly liking it and actively disliking it. However, it also contains 0 3.5 fans. The known 5e fan is a 3.5 fan, the two unknown cases are 3.5 fans, and the group that doesn't like 5e also dislikes 3.5. It breaks down differently too - there's a lot less in the way of not particularly liking it, and a lot more in the way of open contempt. There are games other than D&D out there, and it's in comparison to those that a lot of the criticisms of 5e are cropping up. In comparison to 3.5, I'm seeing a great deal more in the way of thinking that it's at least progress.

*As in, the pool of people with whom I might end up in an actual gaming group at any given point

Yeah, technically 5e was made for 3.5 fanatics to love as well (like myself). It's just that it's the system that the most people on this forum have knowledge of. You can talk about World of Darkness or GURPS or Microscope or whatever, but I won't understand you. 3.5 on the other hand, I get.

[Arnakalar]

I've just read this whole thread and it's given me a lot to reflect on how I adjudicate skill checks/challenges, so thanks all, for that! That said, while I'll try to stick to the focus of house rule fixes, I might digress a bit - sorry.

Illiterate vs archmage, or novice vs. expert problem
I am firmly on the side that an equivalent task requires an equivalent DC - just saying 'barbarians need not apply' is insufficient to me. That said, as I read this thread one thing that came up to me is that even this does require some adjudication. Imagine a scenario where the PCs are trying to get in the good graces of the local duke: one character is a rough commoner with a criminal background - the other is a noble and happens to be the duke's nephew (this goes back to the barbarian/rogue example) - it isn't just that the second character is getting a circumstance bonus, it's that you aren't performing the same check anymore. Another example would be trying to talk your way into the xenophobic Elf-lands (charisma (persuasion)) if the charismatic bard and the rough barbarian both have to check against the same hostile (DC 25) elves, but the Elf in the party only needs a 15, this shouldn't upset anyone. From that logic, I could see setting a different DC, or more likely granting advantage/disadvantage on a check to parse a mystical rune based on circumstances. Just my 2cp.

Trained only, or hard otherwise gating
I've messed around with this some, but found it somewhat dissatisfying, both because maybe someone with a plus 5 int should have a comparable chance to the bard in recalling some history, even if they aren't trained, but also just because it's too binary (and it is in other systems like 3.5) - sure, in real life there are a lot of things you aren't going to do without significant training, but these also aren't binary situations. I personally am pretty good at math, proficient but not expert - I'm trained in calculus, environmental modelling, and system dynamics - someone who doesn't have that training has basically zero % chance of solving a problem I can solve with those skills. There's also 1000 other math skills, and more advanced math I'm worthless at. "Proficient in math" doesn't cut it if you're trying to be simulationist, so I think it's better to avoid the binary.

Doubling Skill checks
Expertise exists, and I think this is a good starting point for looking at the relationship between skill/ability checks and bounded accuracy. I think there's obviously space in the system for this sort of modification, but I'm rather leery about flat doubling of skill bonuses. I strongly dislike the idea of increasing defense incrementally to correct for bloat in skill checks

Minimum results & Reliable
The DMG suggests an alternate rule that works (something like): you can't get lower than your ability SCORE on a check you're proficient in (similar to the 18th level barbarian power) - as discussed, this removes the risk of the Str 20 Goliath slamming into a rickety door (rolls a 1) and the str 8 halfling knocking it down with ease. However in my experience, using similar rules for taking 10 and the rogue's reliable has only ever made the game less fun. IIRC the DMG talks about players becoming disengaged in the action because the PC with the highest bonus either addresses the issue at hand, or doesn't in many situations.

The suggestion of using double prof. bonus as a minimum I actually think sounds pretty interesting, as it is a much more modest effect, but does remove the risk of the expert rolling a 1-5. That said though... I still kind of dislike this, because if you're rolling at all, failure should be possible because you've already said that failure might mean something interesting - if it is a routine check they should just take 10, and if it isn't, real failure should be possible (just improbable).

Expanding Expertise Somehow
In my games, I am VERY liberal about giving NPCs/monsters Expertise and expertise-like abilities. e.g. a master blacksmith: mastery suggests ~tier three at what he does, so a prof of like 4-5, expertise, and maybe a 18 str => Call it a 12-15 at his craft, ~2-3x of a tier 1 or 2 trained adventurer. I haven't yet had that PC in 5e who wants to be really spectacular at one thing (though I have in other games, for example a character who's ambition in life was to build and run the best bar in the world, period (and adventured as a hobby/PR stunt).

Maybe a feat like
Expertise
Choose one ability score. You gain the following benefits:
- Increase the chosen ability score by 1, to a maximum of 20.
- Choose two skills or tools [typically associated with that ability?]- your proficiency bonus on checks with these skills or tools are doubled.

And/or what if, instead of automatic success if your ability/proficiency bonus is high enough, you add something like:

If the DC on an ability check you are proficient in is equal to or less than 10 plus your ability modifier, you have advantage on that check.

This would greatly normalize the results of your 'routine' checks without obviating the risk of failure or causing a huge increase in the absolute range of DCs/bonuses

[Grod_The_Giant]

Illiterate vs archmage, or novice vs. expert problem
I like Advantage/Disadvantage a lot better. At the very least a numerical modifier seems more... appropriate, I guess? I know it's the same as setting a different DC, but it feels a lot better, somehow.


Trained only, or hard otherwise gating
Agreed.


Doubling Skill checks
Expertise exists, and I think this is a good starting point for looking at the relationship between skill/ability checks and bounded accuracy. I think there's obviously space in the system for this sort of modification, but I'm rather leery about flat doubling of skill bonuses. I strongly dislike the idea of increasing defense incrementally to correct for bloat in skill checks
Might I ask why? The goal is less "big numbers for the sake of big numbers" than it is to have numbers that aren't overshadowed by the huge flat curve of the d20 result. But that's kind of a fixture of D&D, making easier solutions (decrease all DCs by 5 and roll a d10, for instance) feel odd.


Minimum results & Reliable
That's a good point.


Maybe a feat like
Expertise
Choose one ability score. You gain the following benefits:
- Increase the chosen ability score by 1, to a maximum of 20.
- Choose two skills or tools [typically associated with that ability?]- your proficiency bonus on checks with these skills or tools are doubled.
Hmm, I can see it. It certainly can't hurt to have a feat like that around; you could always dip Rogue and get a similar benefit.


If the DC on an ability check you are proficient in is equal to or less than 10 plus your ability modifier, you have advantage on that check.
That's not a bad rule.

[Tanarii]

If your goal is to reduce variation on skill rolls, rolling multiple dice to cause a bell curve result would probably be best. It'd certainly make difference in bonuses or penalties much more valuable. 2d10 or 3d6 would make fairly drastic difference in the odds of succeeding. You'd have to check the chances of success for various difficulties at various levels to see what DCs you felt were appropriate though. Because it'd be that much harder to achieve difficult tasks relative to your skill, even as it became less likely to fail easy tasks relative to your skill.

[Arnakalar]

Might I ask why? The goal is less "big numbers for the sake of big numbers" than it is to have numbers that aren't overshadowed by the huge flat curve of the d20 result. But that's kind of a fixture of D&D, making easier solutions (decrease all DCs by 5 and roll a d10, for instance) feel odd.

Ultimately it boils down to the 'curve' - or lack thereof. The problem is swinginess, not really DCs: That a 15th level Wizard with a +10 Arcana fails a 'medium' check 25% of the time, and that a 15th level barbarian w/o int or arcana passes the same check 25% of the time.

If you double the wizard's prof. bonus now the Wizard will never fail the check (which sort of has it's own problems, but whatever), but this creates two new problems: since all we've done is moved the goalposts, now the wizard still fails a DC 25 check (which we can no longer imagine to be 'extremely difficult' - and a DC 30 certainly isn't 'impossible') 50% of the time, and created a gulf between 'the people who can do x' and 'people who can't', and removing this gulf *is* a deliberate design choice for 5e that I agree with.

It also creates a problem in situations with 'moderate' difficulty, where it's entirely reasonable to imagine that different people *should* be able to accomplish the task. Say swimming across a river w/ DC 15, or even 20 - this is a joke to the character with a +10/15, but almost impossible to a character with no bonus.

This is aside to the issue of adding prof. bonus to AC, which I hate, and I see as a signal here that we're talking about bloating the whole system to adjust for the upscaling of skill bonuses. I don't think low CR characters/monsters should be irrelevant to even mid-range foes - it's a major strength of the system that you can throw a pack of CR 2 whatevers at some 8th level PCs and not have it be a waste of time - additionally, what are you going to do about summon spells? Wild shape, polymorph? The only way I can see doubling the skills but leaving combat the same would be to completely separate 'combat' (or whatever) uses from 'skill' use. 'Strenth (athletics) to x' would need to become CMB again, or something like it, or we need to record your 'base ability check bonus (unmodified)' and your 'skill check' bonus, or something stupid like that.

I know this isn't news, and as Tanarii, you, and others have said the solution to swinginess is changing the dice or curve. But I think you can't take the d20 out of d&d without a revolt - there will always be groups where this works (especially if they're used to something like Traveler that only uses 2d6) but they are probably outliers.

Part of the problem is that D&D has always used 'average result' as a baseline for difficulties in ways that strain suspension of disbelief with total disregard for probable outcomes - Another problem is 'take 10', where we're prepared to say that a 'skilled' character can accomplish a 'moderate' task 50% of the time... or 100% of the time on passive or routine checks!? A system in which a character with a passive perception of 14 fails 100% of the time, while a character with +1 better succeeds 100% of the time boggles me.


If your goal is to reduce variation on skill rolls, rolling multiple dice to cause a bell curve result would probably be best. It'd certainly make difference in bonuses or penalties much more valuable. 2d10 or 3d6 would make fairly drastic difference in the odds of succeeding. You'd have to check the chances of success for various difficulties at various levels to see what DCs you felt were appropriate though. Because it'd be that much harder to achieve difficult tasks relative to your skill, even as it became less likely to fail easy tasks relative to your skill.

My suggestion for a reliable application of Advantage is an attempt to reconcile sticking with a d20, and avoiding the bizarre results of take 10s and passive checks, but I can still think of two major problems:

- This is still *very* spiky, because almost all checks in almost all games are multiples of 5 - thus a +5 is actually (something like) +6 greater than a +4. Maybe this is ok, though, or maybe this needs to be something like a variant to 'take 10' or 'routine' checks, that isn't used in tense or stressful situations? But maybe not. I could see being pretty irritated being a 5th level wizard with a +8 Arcana who's just totally outclassed by the lore bard with a +10.
- Advantage interacts with a LOT of existing systems - namely other things that give advantage/disadvantage. If you use my rule, a character can never benefit from a circumstantial bonus to such a check; a barbarian doesn't benefit from the rage bonus to strength checks, nobody benefits from 'enhance ability' except to shore up weaknesses. What happens to passive perception & similar?

One pseudo-solution would be to adjust the break point for advantage on routine checks up: Maybe if bonus + 5 = DC you gain advantage, so the 'expert' (+10) will reliably beat 'easy' checks with no trouble, but still has to roll as normal on 'moderate' checks.

[Segev]

My suggestion for a reliable application of Advantage is an attempt to reconcile sticking with a d20, and avoiding the bizarre results of take 10s and passive checks....

I may have missed part of the discussion in here, but to what "bizarre results" of (in particular) "take 10s" are you referring?

[mephnick]

I may have missed part of the discussion in here, but to what "bizarre results" of (in particular) "take 10s" are you referring?

Likely this:



Part of the problem is that D&D has always used 'average result' as a baseline for difficulties in ways that strain suspension of disbelief with total disregard for probable outcomes - Another problem is 'take 10', where we're prepared to say that a 'skilled' character can accomplish a 'moderate' task 50% of the time... or 100% of the time on passive or routine checks!? A system in which a character with a passive perception of 14 fails 100% of the time, while a character with +1 better succeeds 100% of the time boggles me. .

[Segev]

Ah. I don't think that's nearly as bizarre as they make it out to be. A character who rolls a 9 succeeds where another who rolled a 9 failed because they had +1 difference in their bonuses. The "take 10" business sets a breakpoint for whether or not you need to take a risk, try a little "something" harder to succeed, or if you can play it safe and pull it off.

[Kurald Galain]

now the wizard still fails a DC 25 check (which we can no longer imagine to be 'extremely difficult'

Why not? Something that even a trained expert has trouble with sounds like a fair definition of "extremely difficult" to me.

[Tanarii]

The "take 10" business sets a breakpoint for whether or not you need to take a risk, try a little "something" harder to succeed, or if you can play it safe and pull it off.Actually, it sets the average roll for a task done repeatedly, or a single secret check by the DM.

For the former, it either represents the character doing average for a task, or a variety of rolls being made over and over again, from 1 to 20, without the peskyness of having to check multiple times. In the latter, it represents a way for the DM to avoid having the players know a check was being made, when the mere act of rolling dice might give that away.

In the case of the average of a task being done repeatedly, you can assume one of two things: The PC performs on average over the entire period of time in a less than stressful CHECK RIGHT NOW situation; the PC performs in wildly varying capabilities from round to round and you're using passive to represent the average at any given point of comparison. If you assume the former, it's not a valid complaint. If you assume the latter, it absolutely is a valid complaint.

In the case of the latter, there are plenty of times when it's unfair or unreasonable or un- uh, simulationist? - to assume a player rolled the average, just because it's a secret check by the DM.

[Segev]

Actually, it sets the average roll for a task done repeatedly, or a single secret check by the DM.

For the former, it either represents the character doing average for a task, or a variety of rolls being made over and over again, from 1 to 20, without the peskyness of having to check multiple times. In the latter, it represents a way for the DM to avoid having the players know a check was being made, when the mere act of rolling dice might give that away.

In the case of the average of a task being done repeatedly, you can assume one of two things: The PC performs on average over the entire period of time in a less than stressful CHECK RIGHT NOW situation; the PC performs in wildly varying capabilities from round to round and you're using passive to represent the average at any given point of comparison. If you assume the former, it's not a valid complaint. If you assume the latter, it absolutely is a valid complaint.

In the case of the latter, there are plenty of times when it's unfair or unreasonable or un- uh, simulationist? - to assume a player rolled the average, just because it's a secret check by the DM.
If you want to be pedantic about the numeric representation, "take 10" is actually .5 below average.

What it represents, in simulation, is how well somebody does when they can do something that is, for them, routine based on their training. There are no surprises in it for which they're unprepared, no panic-moments to get in the way, and no difficulties they need to rely on luck or being "in the zone" to overcome.

Which is why "OMG must do this under pressure" requires a roll (to represent that you could screw up due to panic or time-critical screw-ups magnifying into effort-ruining errors) AND you need to roll if "take 10" doesn't make the DC (because you're now having to take chances or try things outside your comfort zone, hoping you get at least a little bit lucky to complement your skill and training; your routine effort isn't going to cut it, and that means you need to chance it and hope for more-than-10 on the die).

The "passive check" bit is actually just because, on average, one of them being fixed to the average should result in the same overall success rate as both being rolled. This isn't 100% true. In fact, when you just need one success out of multiple possible tries, the advantage goes to the side that gets to roll, because they'll have much higher chance of getting a lucky one. If one failure out of a group is all you need, though, the advantage goes to the side which can take the passive score.

[Tanarii]

What it represents, in simulation, is how well somebody does when they can do something that is, for them, routine based on their training. There are no surprises in it for which they're unprepared, no panic-moments to get in the way, and no difficulties they need to rely on luck or being "in the zone" to overcome.
there is no special reason to assume that's what it represents though. That's just one possible interpretation of two.



Which is why "OMG must do this under pressure" requires a rollexcept it doesn't. A character doing OMg I must do this under pressure, round after round, can still use a passive check. For example, scouting down a potentially trap laden hallway, you are under extreme pressure every single round not to mess up. And yet passive checks still cover that situation. its still possible to interpret it as doing an average job every round, even under pressure, but it's not required to roll for a pressure situation.

(I really should have left "OMG pressure" part out of the 'doing an average job' in my last post, because it's not a requirement that you assume there is no pressure in the first interpretation of what the avg for a repeated task means. Just that you assume at, for whatever reason, the character does an average job each round. Edit: ie I just moved the goalposts ;) )

[Segev]

except it doesn't. A character doing OMg I must do this under pressure, round after round, can still use a passive check. For example, scouting down a potentially trap laden hallway, you are under extreme pressure every single round not to mess up. And yet passive checks still cover that situation. its still possible to interpret it as doing an average job every round, even under pressure, but it's not required to roll for a pressure situation.You're incorrectly conflating "passive check" and "take 10."

The former is what you describe above; the latter is "I am not under any particular pressure." The former is an official rule in 5e. The latter is not. But probably is a good house rule to employ.

They do represent different things.

[Tanarii]

You're incorrectly conflating "passive check" and "take 10."You're right. I assumed what was being discussed was passive checks. Because there is no take 10 in 5e. /facepalm

(And yeah, I was conflating them in my first post, which is why I got into the no pressure thing at all.)

[Arnakalar]

Ah. I don't think that's nearly as bizarre as they make it out to be. A character who rolls a 9 succeeds where another who rolled a 9 failed because they had +1 difference in their bonuses. The "take 10" business sets a breakpoint for whether or not you need to take a risk, try a little "something" harder to succeed, or if you can play it safe and pull it off.

Well, 5e lacks any other kind of 'gating' or skill-based tier system. You can *say* "+0 to +4 is beginner, +5 to +9 is proficient" etc. But this kind of breakpoint stills seems weird to me, given the system we're talking about.

Beyond that, where the dissonance happens hits for me is the jump between 'I can do this 45% of the time with effort' to 'I can do this 100% of the time" back to "unless it's a stressful/hazardous situation... in which case I can still only do it 50% of the time." If you're so comfortable with it in 90% of cases, why are you suddenly fumbling 50% of the time? I *get* that a stressful situation/whatever is going to make it harder, but honestly I don't buy that this is even really meant to represent a serious OMG IF I FAIL I'M DEAD stress (which, if D&D was interested in representing at all (which it isn't) would be better served by disadvantage).

'Take 10' is a synthesis of two things: conflation of an 'average result' (i.e. success) with a 'likely result', and expedience. Passive checks/take 10 (and I do get those are different things, but they are mechanically and historically related) are used because it's faster, and it cuts down on uninteresting or problematic die rolls.

The more I think about it, the more I would say my 'advantage on routine checks' thing serves the stated purpose of 'take 10' - to represent mastery in non-hectic situations without totally changing the probabilities or dynamics of a given result. Otoh it flubbs the other purpose because you're still making a roll.


Why not? Something that even a trained expert has trouble with sounds like a fair definition of "extremely difficult" to me.

Sorry, this wasn't well put I think. This was meant to be taken in context of my larger point in general, and the following clause about 'impossible checks' in particular. My main point is doubling the bonuses does nothing to alleviate the swinginess, it just moves the goalposts - The DMG says "A DC 25 task is very hard for low-level characters to accomplish, but becomes more reasonable after 10th level or so. A DC 30 check is nearly impossible for most low-level characters. A 20th-level character with proficiency and - relevant ability score of 20 still needs a 19 or 20 on the die roll to succeed at a task of this difficulty." So now at 11th level you should expect someone with a marginal investment in a class/stat appropriate skill to have a bonus of +13 - they are completing 'nearly impossible' tasks with a greater frequency than a 20th level character was prior. A 15th level character with expertise is looking at a +20, and is completing 'nearly impossible' tasks more often than they're winning a game of craps (...unless they too expertise: Gaming Set (Dice) =O )

And you still haven't increased the granularity at all, which means that a character who spent one of their ~6 starting skills in 'Religion' for fluff reasons is equal to (or possibly superior) to a 15th level Cleric who has devoted their entire like to religious pursuits. Conversely, the expansion of 'Expertise' to more characters or classes allows us, at least, three levels of differentiation.


You're right. I assumed what was being discussed was passive checks. Because there is no take 10 in 5e. /facepalm

Er. Uh. I totally remembered/knew that, and was talking about optional rules/the d20 system/history in general... yeah.

[Kurald Galain]

My main point is doubling the bonuses does nothing to alleviate the swinginess, it just moves the goalposts
But it does.

Aside from the fact that multiple dice are less swingy than one die, the swinginess is caused by the fact that the spread on the die roll is much higher than the spread on attributes, or the spread on training. The result is that luck is more important than skill, and it's worth noting that almost every other RPG does it the other way around.


The DMG says "A DC 25 task is very hard for low-level characters to accomplish, but becomes more reasonable after 10th level or so.
Well it's nice that they say so, but is it actually true? Simple math shows that it's not. A level 10 character has about +2 from attribute and +4 from training, meaning he'll succeed 10% of the time. Even at level 20 the odds are only 20%, except rogues. It's not at all reasonable to expect a character of any level to make a DC 25 check.

[pwykersotz]

Well it's nice that they say so, but is it actually true? Simple math shows that it's not. A level 10 character has about +2 from attribute and +4 from training, meaning he'll succeed 10% of the time. Even at level 20 the odds are only 20%, except rogues. It's not at all reasonable to expect a character of any level to make a DC 25 check.

Are you defining reasonable as 80% or above (or similar)? Because my players roll at that level fairly frequently at level 10. Especially with Bless or other magic active.

[Tanarii]

Especially with Bless or other magic active.Guidance. Bardic Inspiration.

[Arnakalar]

But it does.

Aside from the fact that multiple dice are less swingy than one die, the swinginess is caused by the fact that the spread on the die roll is much higher than the spread on attributes, or the spread on training. The result is that luck is more important than skill, and it's worth noting that almost every other RPG does it the other way around.

I would define swinginess as the relative likelihood of an exceptional outcome versus a common one, and thus *mostly* an outcome of the dice curve, except in situations where the *difference* between DC and bonus is large. A +10 vs AC 30 is exactly as swingy as a +0 vs AC 20, and a +50 vs DC 60 is exactly as swingy as +5 vs DC 15. So I guess I'd say large bonuses vs. die roll solves one type of swinginess (implausible failure on easy) tasks, but there's no improvement at (relative) 'moderate' tasks, and it still introduces new difficulties by increasing the gulf between 'skilled' and 'unskilled'. You've partially solved the reliability problem via brute force, but at a higher cost than adjusting the curve would have.


]Well it's nice that they say so, but is it actually true? Simple math shows that it's not. A level 10 character has about +2 from attribute and +4 from training, meaning he'll succeed 10% of the time. Even at level 20 the odds are only 20%, except rogues. It's not at all reasonable to expect a character of any level to make a DC 25 check.

I don't think we're exactly disagreeing here, in that I also don't think this is entirely reality-based, though for different reasons (i.e. reliability & spikiness). I think we are talking somewhat across each other though: I would interpret the book as referring to a 'primary' ability such as a trained skill & key stat (where I would expect something like a +8-10). I think that's also where pwykersotz is seeing different results than you are.

[UberMagus]

It almost seems as if the problem is in Expertise. If something is supposed to be "impossible", Rogues and Bards running around doing it in multiple things seems to take away from that...

I understand Rogues and Bards as "skill monkeys", but, to me, that always meant they were good at lots of things, rather than ridiculously good at a few things(and, depending on what the pick, really making things weird).

Perhaps the elimination of Expertise, over all, might be more balancing...

[Tanarii]

I think the best way to represent "realistic" skill differences is bell curve for the randomness (ie spikiness), with a bonus to place the center differently for the "average" difficulty task (base reliability). You just need to scale the width of the curve appropriately to the expected variation a person might within the skill. Then scale the bonus to the expected shift between people of various skill levels (ie one base reliability to another).

That's exactly what the current skill system does, for a value of 'variation within individual skill' = even distribution across d20, base reliability at 50% (-1 penalty vs DC10 "Easy"), and 'variation between least skilled and most skilled' = 30% @ level 1 (ie -1 to +5) or 60% at level 20 (-1 to +11). You just need to figure out what values to vary those factors by.

[Arnakalar]

Something else I was thinking about was where I have seen systems with large bonus relative to swingy dice that succeeded, and in the two examples I can come up, both have a deeply embedded mechanic of scaling success: Mutants and Masterminds & a homebrew a friend of mine has been running for ~20 years.

M&M uses a d20 and characters regularly have bonuses in the 10-20 range, however basically all effects in the game have degrees of success that scale by 5s, so a DC 15 is still somewhat meaningful to a character with a +15 because while they can be confident of success the result of the die still matters. Meanwhile, a character with a +5 still has a reasonable chance of success, but will never achieve the absolute results the +15 can.

The homebrew system I'm thinking of uses 2d6, which will not quite as swingy as 1d20 still have a pretty wild relative range (+/- 70% of the mean occurring 15% of the time, not that different than 1d20). Typical bonuses range from ~1-10/12, however exceptional bonuses often exceed that by 50 or 100% and in extreme/notable cases is totally off the charts, such as one demigod-like NPC who famously "laughs at +40s". This works mostly because a) all results are incremental by 3s, and b) in contests, the lower number is subtracted from both characters' bonuses so a fight between a 9 & a 12 begins at +0/+3.

I mention these mostly because it might be interesting to people in this thread, but also because I think it has implications for using similar systems in D&D, where at least to me incremental or tiered results don't have strong mechanical support.

[Kurald Galain]

It almost seems as if the problem is in Expertise. If something is supposed to be "impossible", Rogues and Bards running around doing it in multiple things seems to take away from that...

I understand Rogues and Bards as "skill monkeys", but, to me, that always meant they were good at lots of things, rather than ridiculously good at a few things(and, depending on what the pick, really making things weird).

Perhaps the elimination of Expertise, over all, might be more balancing...

Fair point.

Besides, if every class gets skills, then "being good at skills" is not something that defines a class. Both rogue and bard have a number of other defining features (e.g. sneak attack, songs, magic), so they don't need to have a better skill modifier than other people. Of course, I'd be in favor of giving it to all classes instead of removing it from the rogue and bard.

[Talakeal]

A question about taking a 10:

It is my understanding that most DM's keep the difficulty of a check secret from the PCs.

How do people know whether or not taking ten will succeed? Does the DM tell them? Or do they just have to take a leap of faith and hope that it is a task which they can succeed on a ten or less and if not they auto fail?

[Arnakalar]

I think the best way to represent "realistic" skill differences is bell curve for the randomness (ie spikiness), with a bonus to place the center differently for the "average" difficulty task (base reliability). You just need to scale the width of the curve appropriately to the expected variation a person might within the skill. Then scale the bonus to the expected shift between people of various skill levels (ie one base reliability to another).

That's exactly what the current skill system does, for a value of 'variation within individual skill' = even distribution across d20, base reliability at 50% (-1 penalty vs DC10 "Easy"), and 'variation between least skilled and most skilled' = 30% @ level 1 (ie -1 to +5) or 60% at level 20 (-1 to +11). You just need to figure out what values to vary those factors by.

I think this is inarguable except for intangibles like 'it isn't D&D if I'm not rolling a d20' and 'I was attacked by a vampire as a child and haven't been able to stand the sight of d10s since'.

With that in mind, let's look at some math! Your chances of success with several dice variants suggested and a relative bonus vs a given dc:
http://i.imgur.com/oyHXo5A.png

I leave learning something from the above as an exercise for the reader.

[Xetheral]

My main point is doubling the bonuses does nothing to alleviate the swinginess, it just moves the goalposts

That's only true if you're only considering a single character in a vacumn. Doubling the bonuses doubles the difference between characters who have those bonuses and those who don't. Normally, characters with proficiency succeed at a given task between 10 and 30 percentage points more often than a non-proficient character. If you double the bonuses, that difference increases to between 20 and 60 percentage points.

Even when you do consider only a single character, while it's simply "moving the goalposts" in a strict mathematical sense, at the table DCs are not evenly distributed. Accordingly, a (e.g.) 50% chance of reaching DC 15 and a 25% chance of reaching DC 20 is, in play, qualitatively different from a 50% chance of reaching DC 20 and a 25% chance of reaching DC 25.

[Arnakalar]

That's only true if you're only considering a single character in a vacumn. Doubling the bonuses doubles the difference between characters who have those bonuses and those who don't. Normally, characters with proficiency succeed at a given task between 10 and 30 percentage points more often than a non-proficient character. If you double the bonuses, that difference increases to between 20 and 60 percentage points.

Right, but I'm not sure I like that. I think this is a bug, not a feature:

Ultimately it boils down to the 'curve' - or lack thereof. The problem is swinginess, not really DCs: That a 15th level Wizard with a +10 Arcana fails a 'medium' check 25% of the time, and that a 15th level barbarian w/o int or arcana passes the same check 25% of the time.

If you double the wizard's prof. bonus now the Wizard will never fail the check (which sort of has it's own problems, but whatever), but this creates two new problems: since all we've done is moved the goalposts, now the wizard still fails a DC 25 check (which we can no longer imagine to be 'extremely difficult' - and a DC 30 certainly isn't 'impossible') 50% of the time, and created a gulf between 'the people who can do x' and 'people who can't', and removing this gulf *is* a deliberate design choice for 5e that I agree with.

It also creates a problem in situations with 'moderate' difficulty, where it's entirely reasonable to imagine that different people *should* be able to accomplish the task. Say swimming across a river w/ DC 15, or even 20 - this is a joke to the character with a +10/15, but almost impossible to a character with no bonus.



Even when you do consider only a single character, while it's simply "moving the goalposts" in a strict mathematical sense, at the table DCs are not evenly distributed. Accordingly, a (e.g.) 50% chance of reaching DC 15 and a 25% chance of reaching DC 20 is, in play, qualitatively different from a 50% chance of reaching DC 20 and a 25% chance of reaching DC 25.

I sort of agree? Except I think if you are dramatically altering the scale and range of ability checks, you are also altering the scale of DCs because now it isn't that difficult for a 5th level character to have a +10-12, or a 15th level character to have a +20. Adding one or more DCs on top of 30 was already mentioned, and if you go with this solution it seems like the *only* thing you've done is increased the gap between unskilled/skilled/expert but not done anything for reliability. Or I guess you could keep the standard DCs if you wanted to play a game where the PCs were gonzo hero-types, that would be a reasonable choice for some tables, but not my choice.

To be clear, I'm definitely not totally against increasing the skill bonuses, but I *think* adjusting the curve to increase reliability (either through changing dice to 2d10 or other, or through advantage), and then expanding access to expertise so that a character that wants to feel/be special at 1-2 skills can really shine.

[Xetheral]

Right, but I'm not sure I like that. I think this is a bug, not a feature:

It addresses one of the original poster's complaints:


My problem is with probability though-- not class features. I don't need a level 1 wizard to be a great mage, but I'd expect him to have better Arcana rolls than the Barbarian more than a quarter of the time.


I sort of agree? Except I think if you are dramatically altering the scale and range of ability checks, you are also altering the scale of DCs because now it isn't that difficult for a 5th level character to have a +10-12, or a 15th level character to have a +20. Adding one or more DCs on top of 30 was already mentioned, and if you go with this solution it seems like the *only* thing you've done is increased the gap between unskilled/skilled/expert but not done anything for reliability. Or I guess you could keep the standard DCs if you wanted to play a game where the PCs were gonzo hero-types, that would be a reasonable choice for some tables, but not my choice.

To be clear, I'm definitely not totally against increasing the skill bonuses, but I *think* adjusting the curve to increase reliability (either through changing dice to 2d10 or other, or through advantage), and then expanding access to expertise so that a character that wants to feel/be special at 1-2 skills can really shine.

Different people have different ideas of which probabilities of success translate into which english descriptors. If I were to double the proficiency modifier for skills in my game, I'd keep the DCs more-or-less the same and view it as a correction to a more reasonable success rate rather than the PCs approaching "gonzo hero-types". In the end though, as DM I can adjust the assignment of DCs to fit any success rate I want: all that I need the proficiency bonus for is to set the difference between proficient and nonproficient characters.

[JoeJ]

Even though there's no official rule for taking 20, any time there's no consequence for failure you can still retry over and over until you do. Which means that it's quite reasonable for the DM to just assume a 20 in that circumstance and not make the player roll at all.

That means that a 1st level character with proficiency and an ability score of 16 always succeeds at Very Hard tasks when they're not under any pressure. At 5th level with expertise and an ability score of 18 they can routinely succeed at Nearly Impossible ones.

[Solusek]

The problem with the skill system in 5E hit me immediately when I first started playing it. My first character was a sorcerer, and we play in a fairly large group (7 players). Pretty early on we came across an item where the DM asked for an arcane skill check to find a bit more information out about it.

"Great", I thought. My sorcerer is trained in Arcana! Here is my chance to shine a bit with my choice of skills. No one else in the party was trained in arcana, this is something that I can make myself useful doing. So lets see, I have a +2 proficiency bonus. Intelligence of 10 so no bonus there. Looks like I get d20+2.

Everyone at the table rolls. I am not the highest roll by far. Other characters know more about arcana than my sorcerer does.

Okay, okay, this isn't a big deal we are only low level character afterall. At high levels my magic using character would surely outpace everyone else who didn't bother to spend a skill point on arcana right? But no, not really.

At higher levels my arcana roll might be +3 or +4 or +5, verses everyone else having a +0 or a +1, but when we're talking about all 7 player characters rolling their skill check it will still more often than not be someone else at the table who makes the high roll and knows that bit of arcane trivia rather than my sorcerer knowing it. My sorcerer who is the only arcane magic character at the table and who is the only one that bothered to choose arcana as one of their limited skills.

The 5E skill system feels very unsatisfying from the perspective of wanting a character who is actually good at something and can fill that niche in the party. I guess their design philosophy is only rogues and bards can be any good at skill related things. Everyone else is just at the mercy of the d20 all the time. This is really not good for immersion and character niche development, imo. I don't like it.

[pwykersotz]

The problem with the skill system in 5E hit me immediately when I first started playing it. My first character was a sorcerer, and we play in a fairly large group (7 players). Pretty early on we came across an item where the DM asked for an arcane skill check to find a bit more information out about it.

"Great", I thought. My sorcerer is trained in Arcana! Here is my chance to shine a bit with my choice of skills. No one else in the party was trained in arcana, this is something that I can make myself useful doing. So lets see, I have a +2 proficiency bonus. Intelligence of 10 so no bonus there. Looks like I get d20+2.

Everyone at the table rolls. I am not the highest roll by far. Other characters know more about arcana than my sorcerer does.

Okay, okay, this isn't a big deal we are only low level character afterall. At high levels my magic using character would surely outpace everyone else who didn't bother to spend a skill point on arcana right? But no, not really.

At higher levels my arcana roll might be +3 or +4 or +5, verses everyone else having a +0 or a +1, but when we're talking about all 7 player characters rolling their skill check it will still more often than not be someone else at the table who makes the high roll and knows that bit of arcane trivia rather than my sorcerer knowing it. My sorcerer who is the only arcane magic character at the table and who is the only one that bothered to choose arcana as one of their limited skills.

The 5E skill system feels very unsatisfying from the perspective of wanting a character who is actually good at something and can fill that niche in the party. I guess their design philosophy is only rogues and bards can be any good at skill related things. Everyone else is just at the mercy of the d20 all the time. This is really not good for immersion and character niche development, imo. I don't like it.

I'm incredibly sympathetic to this argument. It's actually how I felt about 3.5 when I first started playing. I remember asking what the heck was the point of a feat that increased a skill by 2. That was such a piddly number, why not increase it by 10 if it's going to have any use? After all, reliably making my checks is the only way to make sure that my character can DO things!

Then I suffered a massive amount of backlash when I got to high levels and realized that I would never, ever be able to make a History check. Or Tumble. Or many others. And missing out on such a large portion of the game was bothersome to me. So now I love 5e's compact skill system, because everyone can participate and random numbers are king over static modifiers, which can be the enemy of simulation but which does a great job keeping this a game.

But still, I really feel ya on that point.

[Tanarii]

"Great", I thought. My sorcerer is trained in Arcana! Here is my chance to shine a bit with my choice of skills. No one else in the party was trained in arcana, this is something that I can make myself useful doing. So lets see, I have a +2 proficiency bonus. Intelligence of 10 so no bonus there. Looks like I get d20+2.

Everyone at the table rolls. I am not the highest roll by far. Other characters know more about arcana than my sorcerer does.
If the DM is allowing 7 skill checks vs an Easy or even a medium check, he's setting the DC as impossible to fail.

That's the kind of thing group checks are for. A group check represents differences of opinion as to the right answer, with some people thinking they know the answers and others being not so sure that's right, etc etc.

That's an aside to your main point of course, which is that you expect proficiency to give a significant advantage over non-proficient characters.

Edit: also with only 13 skills, and each character having 4 proficiencies, the overlap of skills within a party can get pretty high fairly fast as soon as you get more than 4 players in a group. And once you start adding in native skill (ie from ability scores only) vs characters proficiency in low ability score skills, overlap is even higher. I'm honestly surprised that a 7 person group you were the only person with Arcana proficiency, and that no one had a +2 Int bonus among those that didn't.

[Terazul]

The problem with the skill system in 5E hit me immediately when I first started playing it. My first character was a sorcerer, and we play in a fairly large group (7 players). Pretty early on we came across an item where the DM asked for an arcane skill check to find a bit more information out about it.

"Great", I thought. My sorcerer is trained in Arcana! Here is my chance to shine a bit with my choice of skills. No one else in the party was trained in arcana, this is something that I can make myself useful doing. So lets see, I have a +2 proficiency bonus. Intelligence of 10 so no bonus there. Looks like I get d20+2.

Everyone at the table rolls. I am not the highest roll by far. Other characters know more about arcana than my sorcerer does.

Okay, okay, this isn't a big deal we are only low level character afterall. At high levels my magic using character would surely outpace everyone else who didn't bother to spend a skill point on arcana right? But no, not really.

At higher levels my arcana roll might be +3 or +4 or +5, verses everyone else having a +0 or a +1, but when we're talking about all 7 player characters rolling their skill check it will still more often than not be someone else at the table who makes the high roll and knows that bit of arcane trivia rather than my sorcerer knowing it. My sorcerer who is the only arcane magic character at the table and who is the only one that bothered to choose arcana as one of their limited skills.

The 5E skill system feels very unsatisfying from the perspective of wanting a character who is actually good at something and can fill that niche in the party. I guess their design philosophy is only rogues and bards can be any good at skill related things. Everyone else is just at the mercy of the d20 all the time. This is really not good for immersion and character niche development, imo. I don't like it.

This is the problem we've been running into in our group. Proficiency just doesn't feel that significant in terms of skill differentiation unless you're one of those two classes. I had grabbed Survival as a proficiency along with the Outlander background, wanting my character to be--well, a survivalist. Getting the bonus hasn't really helped my character be any more of a trailblazer than any of the other characters, in fact due to the swingy nature of the die they've made their checks far more often than I have so far. If it wasn't for the Background abilities (can always find food, eidetic memory of layouts) that alleviated the need for checks altogether, you wouldn't even be able to differentiate my guy from the person who put their proficiency in Persuasion or something if we were basing it solely of results. Ultimately my character is a survivalist, but it's entirely due to a fiat ability, not his skill choice.

Skill check results are in a tight band, but bonuses are small enough that unless you have the appropriate stat boosted and proficiency (or are a rogue or bard), it doesn't really make much of a difference compared to everyone else.



If the DM is allowing 7 skill checks vs an Easy or even a medium check, he's setting the DC as impossible to fail.
The thing is, if it was a specific Arcana check and he wanted someone in the group to know it, the options were essentially either make the DC low enough that most individuals could roll it in theory (which seems to be implied here), or just flat out tell them. The latter is totally a solution, and he could've easily just given the answer to the only person with proficiency in Arcana! ...But if that's how skills are going to work, why bother with proficiency numbers?

The DM in question doesn't have the option of setting the DC high enough that the Arcana-person could reasonably get to it but others can't, but also since the numbers band is so low, even if he only let the Arcana-proficient person roll they might not get it either, or it's once again low enough of "why couldn't everyone?". Scaling DCs don't jive well with me. As one of my DMs put it, "the DC is the world, your character's roll is interacting with that". Things don't get easier just because the expert looks at it, it's just easier for them because their numbers are (supposedly) higher.

On the whole you get this weird mish-mash system of where if you want certain characters who have written themselves as "good at x" (and dumped a proficiency or two) to excel at those things, you need to ignore rolling altogether because (unless rogue or bard, again) the tight band/small bonuses don't make that much of a difference compared to the uninvested, but we have numbers telling us that we should be rolling anyway. It really feels like they should've just went with a full on FATE-style skill/Aspect system instead of what is given to us.



Edit: also with only 13 skills, and each character having 4 proficiencies, the overlap of skills within a party can get pretty high fairly fast as soon as you get more than 4 players in a group. And once you start adding in native skill (ie from ability scores only) vs characters proficiency in low ability score skills, overlap is even higher. I'm honestly surprised that a 7 person group you were the only person with Arcana proficiency, and that no one had a +2 Int bonus among those that didn't.

Honestly that doesn't surprise me. Unless someone is trying to buff their Int for one of the three Int-saving throws, or they're playing a Wiz/EK/AT, there is no reason to pick up Int outside of RP reasons. ASIs are way too precious to really be spending on it typically, when there's no extra benefits as there were in previous editions.

[Tanarii]

The thing is, if it was a specific Arcana check and he wanted someone in the group to know it, the options were essentially either make the DC low enough that most individuals could roll it in theory (which seems to be implied here), or just flat out tell them. The latter is totally a solution, and he could've easily just given the answer to the only person with proficiency in Arcana! ...But if that's how skills are going to work, why bother with proficiency numbers?Flat out telling them is what the DMG suggests, including telling someone with proficiency. DMG Page 237 flat out says that. The don't make a check questions are:
"Is the task so easy and so free of conflict and stress that there should be no chance of failure?"
"Is the task so inappropriate or impossible [...] that it can't work?"

If a check is something that the every single person in the party can try to do independently, with one success passing and failures having no negative affect, and as such someone is going to succeed, there IS no reason to make a check. If the odds are significant that every single PC will fail, then you might make a check.

If it's something where the group needs to pass as a whole that's what Group checks are for. It could be discussing remembered information which may or may not be accurate, or helping each other climb a wall, or what have you.

If it's something that isn't going to have wildly variable results, and the average result for doing the task for some time is enough, that's what passive checks are for. (Thinking about something to try and recall information can certainly fall into this category. Personally I prefer to save that for researching in a library or the like though.)

[Kurald Galain]

The thing is, if it was a specific Arcana check and he wanted someone in the group to know it, the options were essentially either make the DC low enough that most individuals could roll it in theory (which seems to be implied here), or just flat out tell them. The latter is totally a solution, and he could've easily just given the answer to the only person with proficiency in Arcana! ...But if that's how skills are going to work, why bother with proficiency numbers?
Yes. In my view, there are three problems with having the DM just give info to the "most skilled" character.

First, it's no longer relevant to be the second-best at anything; you can't be the "jack of all trades" character who knows almost as much as the wizard.
Second, it's more work for the DM. While this may sound like very little, it does apply to every skill and a few non-skill things and it adds up. Simply put, one of the jobs of a rules system is to make things easier for the DM.
And third, if the DM forgets for a minute or remembers it wrong, your character's ability disappears, which is just silly. I've been in the situation a couple times where the DM just assumes that some character was the best at some skill, where it was a reasonable assumption but an incorrect one. Again, keeping that from happening is one of the jobs of a rules system.

And while I realize this is YMMV, there is very little reason for a system not to do this. A system doing this if you don't care about it either way is much less annoying than a system not doing this when you do care about it.


Edit: also with only 13 skills, and each character having 4 proficiencies, the overlap of skills within a party can get pretty high fairly fast as soon as you get more than 4 players in a group. And once you start adding in native skill (ie from ability scores only) vs characters proficiency in low ability score skills, overlap is even higher. I'm honestly surprised that a 7 person group you were the only person with Arcana proficiency, and that no one had a +2 Int bonus among those that didn't.
People have a strong incentive to only take skills related to their highest two ability scores. And since almost no class except the wizard has a use for intelligence, I find it entirely plausible that very few non-wizards are trained in Arcana.

[Tanarii]

People have a strong incentive to only take skills related to their highest two ability scores. And since almost no class except the wizard has a use for intelligence, I find it entirely plausible that very few non-wizards are trained in Arcana.Huh. Clearly you play with very different players from the kind of players I play with. I only ever see that kind of thinking in game-store AL games.

[JoeJ]

Remembering information when you're not under any pressure is exactly the kind of thing I would not roll for. If you're trying to remember some obscure fact while in combat, or in the middle of negotiating with a dragon, then roll. Otherwise, I'll assume you stood there and thought about it long enough to have gotten a 20 on the check.

That said, information that is necessary for the adventure to proceed at all ("Sarah," the name the murder victim wrote in his own blood, is the first name of the prime minster's wife) is gained automatically. Information that would derail the adventure at this point (the killer is the old druid on the hill, who is also named Sarah) is automatically not gained.

[Tanarii]

Remembering information when you're not under any pressure is exactly the kind of thing I would not roll for. If you're trying to remember some obscure fact while in combat, or in the middle of negotiating with a dragon, then roll. Otherwise, I'll assume you stood there and thought about it long enough to have gotten a 20 on the check.Certainly that's a fair judgement to make. Not only that, going back (again) to the DMG pg 237, in reference to having no time limit and no consequence for failure and trying again and again, it says "to speed things up, assume that a character spending ten times the normal amount of time need to complete a task automatically succeeds on that task." (The next sentence makes it clear if it's impossible to succeed this shouldn't apply.)

[JoeJ]

Certainly that's a fair judgement to make. Not only that, going back (again) to the DMG pg 237, in reference to having no time limit and no consequence for failure and trying again and again, it says "to speed things up, assume that a character spending ten times the normal amount of time need to complete a task automatically succeeds on that task." (The next sentence makes it clear if it's impossible to succeed this shouldn't apply.)

Thanks for the page reference. I had forgotten about that.

[Knaight]

But it does.

Aside from the fact that multiple dice are less swingy than one die, the swinginess is caused by the fact that the spread on the die roll is much higher than the spread on attributes, or the spread on training. The result is that luck is more important than skill, and it's worth noting that almost every other RPG does it the other way around.
It's more the spread and the uniform die distribution. GURPS has almost as big a spread as D&D, +6 in GURPS is a huge deal. Or to use a d20 example, if the skill mechanic was to roll Nd20, take the median, and add to that, where N is some odd number (which would be a horribly clunky mechanic that should never see use, particularly if N is anything other than 3), the spreads look like this (http://anydice.com/program/7d9d). You can see how a small bonus can be very significant to the probabilities for certain parts of certain curves.


I think this is inarguable except for intangibles like 'it isn't D&D if I'm not rolling a d20' and 'I was attacked by a vampire as a child and haven't been able to stand the sight of d10s since'.
...
I leave learning something from the above as an exercise for the reader.
Absolutely. However, there's three big issues with using something other than d20.
1) It breaks symmetry. The game uses 1d20+N for every roll except damage, and that core mechanic makes it a lot easier to use.
2) It makes probability calculations harder. D&D is deliberately designed to have fairly easy to calculate probabilities; whether this is a good thing or not is up for debate (I personally don't care either way), but it's part of the design.
3) It really wouldn't feel like D&D. It's absolutely an intangible, but D&D does have a distinct feel, and it's clear that 5e was very carefully designed to feel like D&D, particularly given the backlash directed towards 4e focusing pretty heavily on it not feeling like D&D.*

*Whether 4e actually does feel like D&D is completely besides the point here, so let's not take the thread in that direction. The point is entirely that enough people felt that it didn't to make the 5e designers a bit skittish about mechanics that might make something not feel like D&D.

[Telok]

Huh. Clearly you play with very different players from the kind of players I play with. I only ever see that kind of thinking in game-store AL games.

It's also prevalent among players who want a character to be good at something and can quickly estimate numbers in their head. The difference in proficency at level 15 is five points.

A cr 0 npc commoner with no proficencies rolls a d20 on religion checks, an Int 10 cleric of 15th level rolls d20+5 on those same checks. On any dc task those 15 levels of proficency only matter if the cleric rolls one of the five numbers on the d20 that are less than the target number. So proficency only makes a difference 1/4th of the times that the d20 is rolled at 15th level.

The only other effect of proficency is as a soft gateway to the dc 21+ checks. It's a soft gate because of circumstance and stat bonuses. Short of taking rogue or bard levels nobody gets better than a 60% success rate on dc 20 checks.

Actually that gives me an idea for an objective baseline. The untrained 0 level commoner's success rate. Thus a medium difficulty knowledge check is something that half of all commoners know. So something like a dc 15 History check is anything that 25% of the population (not including people with training or study in history) immedately knows off the top of their head.

So you can equate something like math to arcane knowledge. Basic algebra, "solve <linear equation> for X", can be called a dc 5 check. Any calculus will require training so all calculus equations are dc 21+. In America doing you income taxes is what? Dc 10 or 15?

[Tanarii]

It's also prevalent among players who want a character to be good at something and can quickly estimate numbers in their head. The difference in proficency at level 15 is five points.I was referring to players only taking proficiency in things they already have a high ability score in. I'm not sure what your point is, exactly. You crunched numbers, and it's an excellent analysis. But it leaves me saying "and, so this all means ... ?"

[Kurald Galain]

It's more the spread and the uniform die distribution. GURPS has almost as big a spread as D&D, +6 in GURPS is a huge deal. Or to use a d20 example, if the skill mechanic was to roll Nd20, take the median, and add to that, where N is some odd number (which would be a horribly clunky mechanic that should never see use, particularly if N is anything other than 3), the spreads look like this (http://anydice.com/program/7d9d). You can see how a small bonus can be very significant to the probabilities for certain parts of certain curves.
Indeed. To be more precise, I should have said to compare the standard deviation of the attributes and skills to the stddev of the die roll. And of course the stddev of 1d20 is higher than that of 3d6.


Actually that gives me an idea for an objective baseline. The untrained 0 level commoner's success rate. Thus a medium difficulty knowledge check is something that half of all commoners know. So something like a dc 15 History check is anything that 25% of the population (not including people with training or study in history) immedately knows off the top of their head.
Yep. I usually go by this: "easy" means an average guy can do it about half of the time; "difficult" means an average guy basically cannot do it (note that this matches the DCs 10 and 20 in both 3E and 5E). A "trained" person can do easy tasks reliably, an "expert" can do difficult tasks reliably (note that this matches what people mean by "trained" and "expert" in real life; you hire experts precisely to perform difficult tasks). In real life, most professionals and college graduates are at the very least trained and likely experts, and can indeed reliably do things that an average guy cannot. Oh yeah, and "reliable" means at least 95%, and in real life it's commonly higher than that.


Huh. Clearly you play with very different players from the kind of players I play with. I only ever see that kind of thinking in game-store AL games.
I'd expect that Timmy and Vorthos don't particularly care about this, but Johnny and Spike do. Of course, Timmy is unlikely to be interested in knowledge skills in the first place. And yes, as Telok says, some players have an intuitive grasp of estimating numbers almost instantly.

[Tanarii]

Yep. I usually go by this: "easy" means an average guy can do it about half of the time; "difficult" means an average guy basically cannot do it (note that this matches the DCs 10 and 20 in both 3E and 5E). A "trained" person can do easy tasks reliably, an "expert" can do difficult tasks reliably (note that this matches what people mean by "trained" and "expert" in real life; you hire experts precisely to perform difficult tasks). In real life, most professionals and college graduates are at the very least trained and likely experts, and can indeed reliably do things that an average guy cannot. Oh yeah, and "reliable" means at least 95%, and in real life it's commonly higher than that.So using your definitions, D&D 5e assumes that all PCs are trained across the board without proficiency, as they can all do stuff basic stuff reliably without making a check at all. They can also do difficult things that actually require a check (DC 10 "Easy") about 50% of the time, or better if they're somewhat talented or have a bit of extra training (ability score 14-15 or low level proficiency).

Meanwhile high level PC experts (+5 prof bonus) or the very naturally talented (ability mod +5) can make those checks about 75% of the time, or if they are both Experts and Naturally talented, reliably 100% of the time. Those same PC experts who are also naturally talented can make truly challenging things that require check (DC 20 "Hard") half the time.

Edit: Or you can insist on your current way of thinking, that just because it says "Easy" for DC 10 that means it should happen more than 50% of the time when a check is made for anyone to be considered "trained", and adjust the bonuses or target numbers until you're happy.

[Kurald Galain]

They can also do difficult things that actually require a check (DC 10 "Easy")
Wait a second and read what you just said. You mention "difficult things" and equate that with an "easy" DC instead of a "difficult" DC. That's a contradiction: difficult is the opposite of easy.

(basically, you're assuming that the numbers in the DMG are automatically correct and changing the words to match that even if it contradicts the meaning of the word; whereas I'm starting from what the words mean and trying to find numbers that match)

[Tanarii]

Wait a second and read what you just said. You mention "difficult things" and equate that with an "easy" DC instead of a "difficult" DC. That's a contradiction: difficult is the opposite of easy.

(basically, you're assuming that the numbers in the DMG are automatically correct and changing the words to match that even if it contradicts the meaning of the word; whereas I'm starting from what the words mean and trying to find numbers that match)You're starting with what you think the words mean.

The DMG explicitly TELLS YOU that "Easy" means you succeed on the check 50% of the time for someone who is non-proficient. Hell, the math tells you the same thing. That defines what Easy means. I'm just aligning it with your language and definitions you've decided on.

You've defined trained as someone who accomplishes a task reliably 95% of the time. That means a 1st level adventurer is trained on everything. Because the DMG tells you that you don't make checks for things that can be done without a chance of failure. Therefore anything they need to make a check on and have a higher chance of failure on is a more difficult task for that trained person.

Edit: Alternatively, you can require checks for everything, and draw the line of "trained" as proficient. In that case, DC4 is the line for reliable completion by a trained person ("Very Easy").

[Telok]

I was referring to players only taking proficiency in things they already have a high ability score in. I'm not sure what your point is, exactly. You crunched numbers, and it's an excellent analysis. But it leaves me saying "and, so this all means ... ?"
The die is four times bigger than proficency at 15th level, so only one in four rolls is affected by whether or not you are proficent. To put it another way whether you roll high or low on the die is the determining factor of success three times more often than proficency at 15th level.

Hmm. Anything that requires training to get right on the first try must be dc 26+. Because automatic success requires a lack of threats, stresses, time/attempt limits, and failure consequences. I'm on a mobile device and scrolling is carp, so this is first draft thinking out loud. So in endevors where natural ability is insufficient the dc must be high enough for training to make the difference. Ergo all forms of free climbing, long distance swimming, higher math, baking from scratch (without a recipie book), carpentry, masonry... advanced knot tying. Any time someone must be fully trained in order to succeed on the first try without using external knowledge bases is dc 26 or higher.

So taming wild animals or training a horse for riding is dc 26 at least. As is pretty much any sort of metalworking. Fancy knots, architecture and building beyond crude housing and shacks, all forms of complicated mechanical devices (maybe not basic clocks but definitely pocket watches), non-trivial engineering, art forms, navigation, woodworking beyond rough forms, tanning hides. Man that becomes a long list rather quickly.

Perhaps the high end where training is required to make a difference isn't a good metric to use in trying to adjucate and set target numbers. On the low end... skeletons and apes are Int 6, they max out at an 18 if they roll a 20. Thus anything intelligence based that a normal human can do but that an ape or skeleton cannot must be a dc 19 or 20. Ergo reading is dc 19 or 20. That might no be a good idea.

DCs for most things in adventures and monster effects seem to be almost entirely from 10 to 20 for the majority of the game. So zero level commoners are able to do pretty much anything even high level adventurers can do (short of spellcasting) , just not as easily or as often. Random commoners make 25% of medium difficulty arcana checks, 20th level wizards fail 15% of those same arcana checks. Same with fighters and climbing, or clerics and religion, or sorcerers and intimidation.

Maybe 'proficency' isn't really equal to being skilled at something. Maybe it's more like being a talented amature.

[Tehnar]

So using your definitions, D&D 5e assumes that all PCs are trained across the board without proficiency, as they can all do stuff basic stuff reliably without making a check at all. They can also do difficult things that actually require a check (DC 10 "Easy") about 50% of the time, or better if they're somewhat talented or have a bit of extra training (ability score 14-15 or low level proficiency).


Can you explain to me then, if the skill system applies only to PCs then why do NPCs have similar bonuses to skills, and indeed use those same skills in opposed contests with PCs.

Going by your logic that means that all NPCs are also trained across the board, and that just means that no one is trained.

Instead of mindcaulking the system to work, just accept 5e's skill system doesn't work and move on.Designers did a ****ty job, and that is it. No bigger mystery behind it.

[pwykersotz]

Can you explain to me then, if the skill system applies only to PCs then why do NPCs have similar bonuses to skills, and indeed use those same skills in opposed contests with PCs.

Going by your logic that means that all NPCs are also trained across the board, and that just means that no one is trained.

Instead of mindcaulking the system to work, just accept 5e's skill system doesn't work and move on.Designers did a ****ty job, and that is it. No bigger mystery behind it.

There's a point to the discussion. You're not helping it.


As the OP, I request that we stick to discussing useable houserules. If you don't think there's a problem, saying so isn't going to help me.

[Tehnar]

There's a point to the discussion. You're not helping it.

I was refuting a proposed solution by a poster, as illogical. Also, my opinion is that system is so bad adding houserules won't really help.
Also if you really want to be technical the poster asked people who are fine with 5e's skills not to post, and said nothing about those who are not.

But since I have time here goes:

If you want to have a resolution system where you use a single d20 + modifiers vs fixed DC, to have a mechanical representation of what most people expect skilled means you have to use one of two things:

A) large modifiers to the d20 roll. The difference of modifiers should be at least half the range to gain meaningful distinction, in this case a minimum difference of 10. This leads to the 3.x way of doing things.

B) A mechanical way to allow or deny characters rolls on checks. A example of this is ranking task difficulties from 0 to 7. You always roll against DC 10, and add only your ability modifier to the roll.



Task difficulty - proficiency
action


-2
succeed without rolling


-1
roll with advantage


0
roll normally


+1
roll with disadvantage


+2
cannot succeed



This means for example that a character with a proficiency 4, automatically succeeds at tasks difficulty 0-2, rolls with advantage against task difficulty 3, rolls normally against task difficulty 4, rolls with disadvantage against task difficulty 5 and cannot succeed against tasks with difficulty 6 and 7.

Both approaches require a clear description of tasks/DCs and a complete reworking of the system.

[Kurald Galain]

Maybe 'proficency' isn't really equal to being skilled at something. Maybe it's more like being a talented amature.
Indeed. If you work out the math, then a 5E character with proficiency clearly isn't a professional or an expert.

Which is unfortunate for players who are a professional or expert in real life, and who want their fictional dream character to have equal or better skills than they have in real life.

[mgshamster]

Indeed. If you work out the math, then a 5E character with proficiency clearly isn't a professional or an expert.

Which is unfortunate for players who are a professional or expert in real life, and who want their fictional dream character to have equal or better skills than they have in real life.

There's a really easy fix for that: automatically succeeding at tasks that a non-proficient character would be unable to do. Remember, the d20 is only for when the results would be uncertain.

It seems as if to maintain the argument that proficiency means practically nothing, one must also insist that all characters roll for everything.

It also seems that what is being requested here is a mathematical model to simulate the idea of "only roll when uncertain," so one can pre-calculate whether a character will automatically succeed or not.

3.x was all about mathematical modeling. It was my old calculus teacher's favorite edition, because it was all equations. Conversely, an old friend of mine who hated math also hated 3.0 when it came out.

This edition is much less the case. There are no consistent equations, there is no internal consistency throughout the game - it's all about being flexible in the rules to allow for a GM to set up the game they want.

That's they way it is in 5e, and some people don't like that. It's ok to not like it, so for those who don't, let's focus on setting up a model to give them what they've requested rather than arguing whether our personal preference for the current system is the best.

Let's focus on a mathematical model so these players/GMs can predetermine when their characters will succeed or fail at a given task, rather than relying on the GM's and their own judgment.

[Shaofoo]

I was refuting a proposed solution by a poster, as illogical. Also, my opinion is that system is so bad adding houserules won't really help.
Also if you really want to be technical the poster asked people who are fine with 5e's skills not to post, and said nothing about those who are not.

But since I have time here goes:

If you want to have a resolution system where you use a single d20 + modifiers vs fixed DC, to have a mechanical representation of what most people expect skilled means you have to use one of two things:

A) large modifiers to the d20 roll. The difference of modifiers should be at least half the range to gain meaningful distinction, in this case a minimum difference of 10. This leads to the 3.x way of doing things.

B) A mechanical way to allow or deny characters rolls on checks. A example of this is ranking task difficulties from 0 to 7. You always roll against DC 10, and add only your ability modifier to the roll.



Task difficulty - proficiency
action


-2
succeed without rolling


-1
roll with advantage


0
roll normally


+1
roll with disadvantage


+2
cannot succeed



This means for example that a character with a proficiency 4, automatically succeeds at tasks difficulty 0-2, rolls with advantage against task difficulty 3, rolls normally against task difficulty 4, rolls with disadvantage against task difficulty 5 and cannot succeed against tasks with difficulty 6 and 7.

Both approaches require a clear description of tasks/DCs and a complete reworking of the system.

A sounds a lot like you wanting to go back to 3.x where you could have such a high bonus that the d20 didn't matter.

I can't even make heads or tails about how to approach B, everything is DC 10 but then there is a separate stat that governs the true difficulty of the action.


But I sincerely wonder, if the d20 is such anathema to people about using skills, why aren't playing a Rogue or a Bard valid choices. Why not use the classes made for skills?

A Rogue at level 11 can have up to THIRTEEN skills or tools (Five from the class itself, Three from his background, Three if he chooses the Skilled feat and two if he chooses Half Elf as his race) where he can't roll below a 10. That is a normal DC is an automatic success just by virtue of the d20 and basically your proficiency and stat bonuses set the floor of these skills. At level 13 and with expertise, max stat and the most uncooperative d20 the worst he can do is a check of 25, which means that Very Hard checks are guaranteed for a Rogue at level 13 in stuff that he is good for and nearly impossible is more than reachable.

A Bard not only has expertise but also has ways to boost his skill checks with inspiration and a College Bard in later levels he can actually double dip on inspiration and roll 2d10 or even 2d12 for a single check. Plus the spells that he can have (not to mention that he can have ANY spell in the game) and if you want an example then try Glibness where he can't roll below a 15 on any Charisma check.

I am sure people won't understand what I am going with this but since some people seems to treat the d20 as if all the sides are 1 there are classes that can ignore the d20 and do ignore the d20 when it is misbehaving. There are ways to ignore the d20 in a skill if you choose to go the route.

It seems that the highest levels of skill users can perform admirably and can easily reach the highest levels of difficulty. Everyone else will probably not have everything guaranteed for skills but personally I would think the skill based classes should get the upper hand over those who aren't focused on skills and even then not all classes are left in the dust (Barbarians get advantage on strength skills while raging and can use their score in place of a skill if it is bad, Rangers have their favored enemies and enviroments for tracking, Fighters get half proficiency on all physical checks).

[Tehnar]

A sounds a lot like you wanting to go back to 3.x where you could have such a high bonus that the d20 didn't matter.

I can't even make heads or tails about how to approach B, everything is DC 10 but then there is a separate stat that governs the true difficulty of the action.

I don't want to go back to 3.x, but if you want to use a single resolution d20+modifiers vs DC, you have to have big modifiers or a mechanical way to exclude certain characters from rolls.

The way 5e does it either leads to either essentially random outputs, or outputs being decided by DM fiat. I don't mind that in certain systems (like Dark Heresy where skills are essentially random but then again is the chance to be eaten by demons) or completely narrative games, but DnD is not that type of game.

And thirdly I don't know why you are bothered by the fact that in 3.x the d20 sometimes did not matter. Yes the ninja didn't have to roll stealth when sneaking by bored town guards, but he had to roll when sneaking by a beholder or dragon. Why would a paladin in heavy plate have a chance sneaking by a beholder, especially if he didn't invest any resources into stealth? 3.x had many problems with its skill system, but that was not it.

Its ok for characters to be awesome at some things, as long as other characters are awesome at other things.




But I sincerely wonder, if the d20 is such anathema to people about using skills, why aren't playing a Rogue or a Bard valid choices. Why not use the classes made for skills?

A Rogue at level 11 can have up to THIRTEEN skills or tools (Five from the class itself, Three from his background, Three if he chooses the Skilled feat and two if he chooses Half Elf as his race) where he can't roll below a 10. That is a normal DC is an automatic success just by virtue of the d20 and basically your proficiency and stat bonuses set the floor of these skills. At level 13 and with expertise, max stat and the most uncooperative d20 the worst he can do is a check of 25, which means that Very Hard checks are guaranteed for a Rogue at level 13 in stuff that he is good for and nearly impossible is more than reachable.

A Bard not only has expertise but also has ways to boost his skill checks with inspiration and a College Bard in later levels he can actually double dip on inspiration and roll 2d10 or even 2d12 for a single check. Plus the spells that he can have (not to mention that he can have ANY spell in the game) and if you want an example then try Glibness where he can't roll below a 15 on any Charisma check.

I am sure people won't understand what I am going with this but since some people seems to treat the d20 as if all the sides are 1 there are classes that can ignore the d20 and do ignore the d20 when it is misbehaving. There are ways to ignore the d20 in a skill if you choose to go the route.

It seems that the highest levels of skill users can perform admirably and can easily reach the highest levels of difficulty. Everyone else will probably not have everything guaranteed for skills but personally I would think the skill based classes should get the upper hand over those who aren't focused on skills and even then not all classes are left in the dust (Barbarians get advantage on strength skills while raging and can use their score in place of a skill if it is bad, Rangers have their favored enemies and enviroments for tracking, Fighters get half proficiency on all physical checks).

Those numbers are true, but only for two classes, with a few chosen skills, in their primary attribute and at high level. Anything else and they are still victim to the rng gods.

Some people want to play skillful characters without playing high level rogues.

[pwykersotz]

So here's a questoin: why not just import the 3.5 system wholecloth?

You can make the following choices without any fallout.

Upgrade the limits of the DC to 50 or keep the limit 30 and expect that top level adventurers will reliably perform the impossible
Keep the 5e skill list or use the 3.5 one (or heck, Pathfinder too)
Use a skill point system based on classes or use an SRD variant (http://www.d20srd.org/srd/variant/buildingCharacters/alternativeSkillSystems.htm)
Use the 3.5 skill guidelines (such as a DC20 Jump/Athletics for a 20' jump) or the 5e freeform ones
Have Expertise do it's normal "double proficiency bonus" or make it the 3.5 Skill Mastery instead.

And tons of other options. It's perfectly modular, since skills don't interact mechanically with much else. About the most work you'd have to do is convert the monsters you use, but that's true of most of these proposed systems.

[Shaofoo]

Its ok for characters to be awesome at some things, as long as other characters are awesome at other things.


Everyone is awesome at something, weather it be skills, fighting, magic or other, the problem here is that we are basiclaly being myopic and the only thing that matters is skills. If you have a concept then I am sure there is a class for it. I am not sure why should the fighter be so jealous of the rogue when he is the best swordsmaster.



Those numbers are true, but only for two classes, with a few chosen skills, in their primary attribute and at high level. Anything else and they are still victim to the rng gods.

Some people want to play skillful characters without playing high level rogues.

I am not sure what to say because you basically are saying that you don't want to be the class that gives you a lot of skills when you want a lot of skills. It seems like you are just being contrarian for the sake of being contrarian. It is as well as saying I want to cast spells as a Barbarian so give me spells.

If you don't want to be victim to the RNG gods there are two choices given to you, if you refuse them I think you have no right to complain when you passed the solution.

Being so skilled does exist in 5e, it is just limited to specific classes like everything else. How much does skill matter to you really?

[Solusek]

One problem, for me, is that many iconic characters from fantasy do have at least some non-combat things they are very good at. Even those characters who are obviously not rogues or bards. One of the things I think a good D&D system should do is let people make PCs that remind them of the cool characters they read about in fiction. A skill system that only lets rogues and bards actually be any good at a skill is a big problem for me.

I wouldn't mind it so much if everyone got a couple skills that they were markedly better than the untrained person in, and then rogues and bards were even a level above that with their expertise skills. I don't mind rogues and bards getting a class perk to be the absolute best at something skill-wise. My problem is that with the basic proficiency rules every other class currently uses they are only a small bit better than the untrained person. Even at something you want to be a core part of that character - like the acrobatic monk, or the scholarly cleric, or (in my case) the sorcerer who knows a lot about things arcane. It's just doesn't happen with the 5E rules.

[Shaofoo]

One problem, for me, is that many iconic characters from fantasy do have at least some non-combat things they are very good at. Even those characters who are obviously not rogues or bards. One of the things I think a good D&D system should do is let people make PCs that remind them of the cool characters they read about in fiction. A skill system that only lets rogues and bards actually be any good at a skill is a big problem for me.

I wouldn't mind it so much if everyone got a couple skills that they were markedly better than the untrained person in, and then rogues and bards were even a level above that with their expertise skills. I don't mind rogues and bards getting a class perk to be the absolute best at something skill-wise. My problem is that with the basic proficiency rules every other class currently uses they are only a small bit better than the untrained person. Even at something you want to be a core part of that character - like the acrobatic monk, or the scholarly cleric, or (in my case) the sorcerer who knows a lot about things arcane. It's just doesn't happen with the 5E rules.

Everyone can be good at skills. The fact that between untrained and trained can be +6 already and this is discounting stats which can push that over +5. I am not sure how is +11 little?

And your examples gives me a good example as to how everyone is acting myopic about skills. You can totally be an acrobatic Monk and forgo all skills because most acrobatic skills are class features for the monk. You don't need skills to jump or climb or anything Acrobatic, being a monk lets you do that automatically (with some expenditure in ki but the option exists). A scholar cleric has the added advantage of knowing divination spells to boot so if you don't know anything you could ask your god. I also assume that for your character you are both trained in Arcana and have a decent Intelligence as well, is being the best student the most important part of your character lore?

It just feels that people just want a sure thing, for skills to become spells. To remove both DM and RNG from the equation and just do what you want to do without anything holding you back.

Being good is not being infallible. Being good is not being the best. If you want to be the best there is a path that you must take, I think that is how it should be.