Skill Checks shouldn't use a d20

[MadBear]

With the conversation around skill checks and what is a reasonable DC to set, it brought me back to a fundamental question about how we go about these skill checks. Why do we let luck have a bigger deciding factor on success then the amount of training or skill involved?

With the reduction in numbers that make it so no one is rocking a +45 to a skill check (characters in my old 3.x days) this has made it so that the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Often the "fixes" to these issues involve not letting players roll if they're not trained in a skill, or setting the bar high enough that only those trained can succeed, or having different DC's for different characters. (I'm not saying these fixes are good, only that this is what many DM's often end up doing).

This bring back the question, of how do we represent chance without it overshadowing training, and natural skill (proficiency + Ability score).

My gut is to help reduce the standard deviation via using 2d10 or 5d4's. In this way, characters are far more likely to roll the average then they are to roll extremely well or extremely low.

What are your thoughts? Should we continue using the d20 and move on, or is there a better way?

[Kane0]

I tried 2d10 ability checks for a few sessions, worked fine except for the fact that everybody kept forgetting and reaching for the d20.

D&D is very good at training you that the default for doing something is rolling a d20

[NecessaryWeevil]

Seems reasonable. If my DM did this I wouldn't complain, but I might ask, "Doesn't the same logic hold for attacks and saves?"

[JonBeowulf]

I just tweak the DC if a check should be more difficult for a non-proficient character BASED ON HOW THE CHARACTER HAS BEEN PLAYED. Only for checks that the character does not have proficiency.

If the Barb has been showing interest in arcane stuff, then they get the same DC as the Wiz. If the Wiz has been testing their physical limits, then they get the same DC as the Barb.

[PhantomSoul]

One thing that came to mind in one of the many recent threads is that a more PF2-like rendition (where you actually have proficiency categories not just a proficiency bonus), you could "cheat" and give disadvantage below the requisite proficiency category (or at the minimal proficiency category) and get advantage if your proficiency category is good enough.

E.g. using the PF2 categories I googled, proficiency goes 0=untrained, 2=trained, 4=expert, 6=master, 8=legendary. If the task is rated at "expert" level, perhaps experts roll normally, masters+ roll with advantage, trained can attempt at disadvantage, and untrained can't meaningfully attempt the task. (In principle those could be independent, but for simplicity this is probably plenty of starting complexity!)

[animorte]

Provide a massive table for a ton of different DCs that would require the use of skills, separated by each skill individually. Try to account for as many things (or more likely a number of vague things) that would apply to each of those skills. Don't tell me its unreasonable because they've made an awful lot of pages full of tables for random items (magical and otherwise), not to mention the Wild Magic tables that not many people seem to care for. Though it would certainly be task for the DM to become familiar with.

You know how we have a passive Perception? Make everything a set standard. Any skill usage functions exactly the same way, it would always add the exact same things except it would look more like spell save DCs or AC.

So if you're a level 4 Rogue with 18 Dex making a Dexterity (Stealth) check and you have expertise. This might look like: 10 + Dex + (PB x 2) = 18
That would go against whoever's Perception might detect you. So on and so forth.

We could bring back the advantage/disadvantage when it is relevant (that took hold in 3.5e I believe) in which you merely gain a +2/-2. Either that or add/subtract one instance of your proficiency bonus.

TL;DR: Never roll the dice again!

[Psyren]

With the conversation around skill checks and what is a reasonable DC to set, it brought me back to a fundamental question about how we go about these skill checks. Why do we let luck have a bigger deciding factor on success then the amount of training or skill involved?

With the reduction in numbers that make it so no one is rocking a +45 to a skill check (characters in my old 3.x days) this has made it so that the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Often the "fixes" to these issues involve not letting players roll if they're not trained in a skill, or setting the bar high enough that only those trained can succeed, or having different DC's for different characters. (I'm not saying these fixes are good, only that this is what many DM's often end up doing).

This bring back the question, of how do we represent chance without it overshadowing training, and natural skill (proficiency + Ability score).

My gut is to help reduce the standard deviation via using 2d10 or 5d4's. In this way, characters are far more likely to roll the average then they are to roll extremely well or extremely low.

What are your thoughts? Should we continue using the d20 and move on, or is there a better way?

1) How the heck does the 5d4 solution work with advantage? If you use "roll (2x), pick best (x)" then the higher x gets the more powerful advantage (and conversely disadvantage) becomes. Meanwhile rolling each pair of dice separately and picking the better of the two will give you a fairer result, but bog play down to molasses.

2) I'd much rather use the "fixes" (specifically the first one, though I use many more gates than simply trained vs. not trained) than change the dice distribution. Sure - getting a 5 on 5d4 is a lot less likely than getting a 5 on a 1d20. But it's going to happen, and when it does you'll be right back to the awkward situation you were trying to avoid. The gating fix meanwhile eliminates incongruity completely. And there's no need to retrain people to not reach for a d20.

[Skrum]

If the Barb has been showing interest in arcane stuff, then they get the same DC as the Wiz. If the Wiz has been testing their physical limits, then they get the same DC as the Barb.

I've heard this said by several people, and I gotta say, it makes no sense to me at all. Modifying the DC for the character is the same as giving the character a penalty or bonus to the roll. It's the same math. Isn't that already reflected in the character's skills? I mean, if this is just an ad hoc adjustment to "correct" the weird spread on skill checks, I guess whatever works for you...but maybe let all characters add twice their proficiency bonus to their proficient skills? As a player, I squirm at the idea of the DM just making stuff up like this.

[PhantomSoul]

I've heard this said by several people, and I gotta say, it makes no sense to me at all. Modifying the DC for the character is the same as giving the character a penalty or bonus to the roll. It's the same math. Isn't that already reflected in the character's skills? I mean, if this is just an ad hoc adjustment to "correct" the weird spread on skill checks, I guess whatever works for you...but maybe let all characters add twice their proficiency bonus to their proficient skills? As a player, I squirm at the idea of the DM just making stuff up like this.

Very very marginally reflected in rolls and very very easily overcome by stats and by the d20 is the recurring problem in the threads
But if it makes you feel better, it's all made up, not just individual bonuses/penalties! :)

[NichG]

Its not at all D&D-ish, but I could imagine something like 'the task has a certain number that must be met; subtract your modifier and proficiency bonus directly from that number; now, if you're still below that number, you can roll sets of 3d6 at a rate of once per round of performing the task; 5s, and 6s give you +1 to your number against the threshold, but count and accumulate how many 1s you roll - if you roll 3 ones over the course of this process, you critically fail and suffer a harm plus cannot attempt this instance of this task again. You can stop at any time, and events going on can interrupt you; in either case, you lose progress and may choose to restart the process from scratch'.

Advantage would mean that 4s, 5s, and 6s give you +1 and you can suffer 4 total ones before crit failure, or something like that.

So basically in this system no DC would be absolutely impossible, but higher DCs become exponentially harder to make and also inevitably take multiple rounds to complete.

You could also have class features that do things like let you take damage in exchange for ignoring a '1' as part of this process, or change the size of the dice pool you roll each round (smaller pool = more ability to abort before crit fail but takes longer, larger pool = faster task performance but higher risk of going over)

[KorvinStarmast]

Skill Checks shouldn't use a d20 Given that we are still in the D20 system, you may want to go and play a different game. Without the chance to fail success has no meaning.

I just tweak the DC if a check should be more difficult for a non-proficient character BASED ON HOW THE CHARACTER HAS BEEN PLAYED. Why overwork youself? As DM, you can apply adv and disadv as the circumstance indicates. That's in the rule book in chapter 7.

[Sorinth]

The skill checks are meant to represent things that are close to being 50-50, it's one reason the books say not to bother even making a check when the DC is low. It sounds like your problems could be solved simply by letting players succeed at stuff without requiring a roll. The DMG discusses many of these things and proposes several ways of handling those situations, have you tried them? For example your problem examples could quite easily be solved using the "Variant: Automatic Success" from the DMG.

[Rukelnikov]

With the conversation around skill checks and what is a reasonable DC to set, it brought me back to a fundamental question about how we go about these skill checks. Why do we let luck have a bigger deciding factor on success then the amount of training or skill involved?

With the reduction in numbers that make it so no one is rocking a +45 to a skill check (characters in my old 3.x days) this has made it so that the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Often the "fixes" to these issues involve not letting players roll if they're not trained in a skill, or setting the bar high enough that only those trained can succeed, or having different DC's for different characters. (I'm not saying these fixes are good, only that this is what many DM's often end up doing).

This bring back the question, of how do we represent chance without it overshadowing training, and natural skill (proficiency + Ability score).

My gut is to help reduce the standard deviation via using 2d10 or 5d4's. In this way, characters are far more likely to roll the average then they are to roll extremely well or extremely low.

What are your thoughts? Should we continue using the d20 and move on, or is there a better way?

For good or bad, that is 100% intended, the term "bounded accuracy" which is often lobed around means exactly this, you are not supposed to make the d20 irrelevant, and while it can still be done, it takes far longer and is done in smaller capacity than in 3e.

If you wanna replace the d20 with another set of dice 3d6 has been a thing for a while, I cant remember if the optional rule already existed in 2e, but it was there in 3e for sure. Probalby because the change to 2d10 may not be consistent enough for the intended purpose, and also because d6s are usually extremely abundant. Using 5d4s is gonna be a pain to roll, and many groups may not have enough d4s so that everyone has their own set of 5d4s, which would result in even more bogging down.

[OldTrees1]

Have you considered using 1d8 for skill checks? It significantly improves the math.

[MadBear]

Have you considered using 1d8 for skill checks? It significantly improves the math.

Simple. Elegant. I may try that with my players. (Obviously would have to adjust the DC's).


Given that we are still in the D20 system, you may want to go and play a different game.

My title is mostly tongue and cheek. I'm mostly just curious what other ideas are out there, because every proposed solution I've heard to said problem has fairly strong negative reactions.


1) How the heck does the 5d4 solution work with advantage? If you use "roll (2x), pick best (x)" then the higher x gets the more powerful advantage (and conversely disadvantage) becomes. Meanwhile rolling each pair of dice separately and picking the better of the two will give you a fairer result, but bog play down to molasses.

2) I'd much rather use the "fixes" (specifically the first one, though I use many more gates than simply trained vs. not trained) than change the dice distribution. Sure - getting a 5 on 5d4 is a lot less likely than getting a 5 on a 1d20. But it's going to happen, and when it does you'll be right back to the awkward situation you were trying to avoid. The gating fix meanwhile eliminates incongruity completely. And there's no need to retrain people to not reach for a d20.

1. I agree that 5d4 would be a nightmare with advantage/disadvantage. I'm in no way sold on my current proposal, so much as I'm looking for better ideas.
2. While you would get some of those awkward situations, it would be less frequent hence getting us closer to the goal.


Seems reasonable. If my DM did this I wouldn't complain, but I might ask, "Doesn't the same logic hold for attacks and saves?"

You'd think so, but we never hear the same complaint to the same degree with attacks. My guess is because it's not a huge deal if in a 6 second span someone luckily hits/misses, but a 10 minute climb up a rock face being controlled by a single roll just feels worse. To bring those situations into closer parity, it'd be like if a sword fight between a wizard and a fighter was decided on a single roll.....

That actually inspires me a bit. I know skill challenges are more common now (multiple rolls to see if a group succeeds/fails), so what if we used similar system instead. Now this would take some refinement, but one thing to keep in mind is that in a battle there is a depleting resource your trying to overcome (HP's) before they overcome yours. What if instead, players had a pool of rerolls connected to skills their trained in. Lets make it equal to their proficiency bonus. If a rogue fails a stealth roll and is trained in it, they can reroll the dice a number of times for each skill equal to their proficiency bonus. However this reroll pool only refreshes on a long rest.

This means that a high level rogue can reroll up to 6 stealth checks that they've failed (after advantage/disadvantage), but these rerolls are a depleting resource over time. Characters not trained in a skill can absolutely attempt the check, but if they're not trained, they get the one chance.

So as a character's level increases they reliably become even better at succeeding on checks they are trained in.

Again, I'm sure this solution has issues, and I'd love to hear criticism or other iterations.

[Rukelnikov]

My title is mostly tongue and cheek. I'm mostly just curious what other ideas are out there, because every proposed solution I've heard to said problem has fairly strong negative reactions.

Tbh, I'd be curious as to how many people find this to be a problem...

[Psyren]

My title is mostly tongue and cheek. I'm mostly just curious what other ideas are out there, because every proposed solution I've heard to said problem has fairly strong negative reactions.

Including yours :smallwink::smallbiggrin:



While you would get some of those awkward situations, it would be less frequent hence getting us closer to the goal.

As stated, you can't get less frequent than zero so that's what I'd go with.


You'd think so, but we never hear the same complaint to the same degree with attacks. My guess is because it's not a huge deal if in a 6 second span someone luckily hits/misses, but a 10 minute climb up a rock face being controlled by a single roll just feels worse.

That depends on what the single roll does. If you fall to your splattery death with one fail, yeah that's pretty crappy, but the superior approach is to not do that.

The short version is that if any of the outcomes of your d20 are not interesting, make them interesting. If you can't do that, that's the ideal time to dispense with the d20.

[JonBeowulf]

<snip>
Why overwork yourself? As DM, you can apply adv and disadv as the circumstance indicates. That's in the rule book in chapter 7.
Don't you just hate it when someone comes a long and makes your life easier? I want to be mad at you but I'm too busy SMH at my own over-complication.

[KorvinStarmast]

Don't you just hate it when someone comes a long and makes your life easier? I want to be mad at you but I'm too busy SMH at my own over-complication. We live to serve. :smallsmile:
(We= Me...and the voices in my head! :smallbiggrin: )

[Keravath]

I think it makes more sense to just use d20s for attack rolls, skill checks and saves. There is a certain streamlined elegance to just using the one type of die to resolve all of these situations.

However, I do recognize the problem and the dissonance caused when a character with a high skill fails while a character with a low skill fails just because of the die roll.

I tend to mitigate this by using two techniques
1) Passive skills
2) Proficiency gating

If a wizard has a passive arcana score of 18 doing research in a library then they will most likely succeed automatically on anything less than a DC18 unless there is a time constraint or other consequence. There won't even be a die roll since I will just narrate the success of the character with the high passive score (which also helps validate their choice of skills).

Second, if the barbarian knows nothing about Arcana then it may actually be impossible for them to succeed at a task researching the arcane a library. They might not even get a roll. Consider a character that is illiterate, if they can't read, certain tasks just won't be possible and the same can apply for certain checks in skills without proficiency.

For the few cases that are left, where the wizard can't succeed passively, then folks can make a die roll and it is possible that a less proficient (or in other situations non-proficient) character might be successful. However, because of the use of passive skills and proficiency gating this is usually infrequent enough to not be an issue.

P.S. Rolling more than one die would just slow down the game immensely. Some folks are very slow at addition, and have issues doing it in their head while keeping track of a running total. It just isn't one of their strengths. Even adding 2d10 would slow things, adding 5d4 would be unimaginably slow based on how long it has taken some folks to add up 8d6 when rolling a fireball or sneak attack or even just adding weapon damage+modifier+hunter's mark or hex. More than one die to resolve a skill check will slow the game down (unless you have a group good at math).

[NichG]

I think it makes more sense to just use d20s for attack rolls, skill checks and saves. There is a certain streamlined elegance to just using the one type of die to resolve all of these situations.

However, I do recognize the problem and the dissonance caused when a character with a high skill fails while a character with a low skill fails just because of the die roll.

I tend to mitigate this by using two techniques
1) Passive skills
2) Proficiency gating

If a wizard has a passive arcana score of 18 doing research in a library then they will most likely succeed automatically on anything less than a DC18 unless there is a time constraint or other consequence. There won't even be a die roll since I will just narrate the success of the character with the high passive score (which also helps validate their choice of skills).

Second, if the barbarian knows nothing about Arcana then it may actually be impossible for them to succeed at a task researching the arcane a library. They might not even get a roll. Consider a character that is illiterate, if they can't read, certain tasks just won't be possible and the same can apply for certain checks in skills without proficiency.

For the few cases that are left, where the wizard can't succeed passively, then folks can make a die roll and it is possible that a less proficient (or in other situations non-proficient) character might be successful. However, because of the use of passive skills and proficiency gating this is usually infrequent enough to not be an issue.

P.S. Rolling more than one die would just slow down the game immensely. Some folks are very slow at addition, and have issues doing it in their head while keeping track of a running total. It just isn't one of their strengths. Even adding 2d10 would slow things, adding 5d4 would be unimaginably slow based on how long it has taken some folks to add up 8d6 when rolling a fireball or sneak attack or even just adding weapon damage+modifier+hunter's mark or hex. More than one die to resolve a skill check will slow the game down (unless you have a group good at math).

There's a meta-argument here that one of the problems with the way skills are handled is that the system straddles the fence as to whether they're high-abstraction things like 'resolve this negotiation' or low-abstraction things like 'move through this 5ft square without falling down'. If skills are acting as high-abstraction things, having them be a bit slower isn't necessarily a bad thing. Makes the non-combat parts of the game feel like there's more to them. The trick would be to do it in a way that isn't just 4e skill challenges 'generate a probability distribution in a more tedious fashion', but to actually have changing circumstances and meaningful decision points during the multi-roll process.

That's why I sort of liked the idea of moving more towards an 'accumulate sufficient score' versus 'pass/fail' model for things. So that numbers are more about determining the cost of success (in time, side-effects, resources, etc) rather than whether or not you succeed 'this time'. That way there's at least a bit of sunk cost fallacy mini-game to be had - do I keep trying this check even if the consequences for failure get worse and worse as I retry, or do we give up and start over doing something else?

[Sorinth]

I think it makes more sense to just use d20s for attack rolls, skill checks and saves. There is a certain streamlined elegance to just using the one type of die to resolve all of these situations.

However, I do recognize the problem and the dissonance caused when a character with a high skill fails while a character with a low skill fails just because of the die roll.

I tend to mitigate this by using two techniques
1) Passive skills
2) Proficiency gating

If a wizard has a passive arcana score of 18 doing research in a library then they will most likely succeed automatically on anything less than a DC18 unless there is a time constraint or other consequence. There won't even be a die roll since I will just narrate the success of the character with the high passive score (which also helps validate their choice of skills).

Second, if the barbarian knows nothing about Arcana then it may actually be impossible for them to succeed at a task researching the arcane a library. They might not even get a roll. Consider a character that is illiterate, if they can't read, certain tasks just won't be possible and the same can apply for certain checks in skills without proficiency.

For the few cases that are left, where the wizard can't succeed passively, then folks can make a die roll and it is possible that a less proficient (or in other situations non-proficient) character might be successful. However, because of the use of passive skills and proficiency gating this is usually infrequent enough to not be an issue.

P.S. Rolling more than one die would just slow down the game immensely. Some folks are very slow at addition, and have issues doing it in their head while keeping track of a running total. It just isn't one of their strengths. Even adding 2d10 would slow things, adding 5d4 would be unimaginably slow based on how long it has taken some folks to add up 8d6 when rolling a fireball or sneak attack or even just adding weapon damage+modifier+hunter's mark or hex. More than one die to resolve a skill check will slow the game down (unless you have a group good at math).

Another good option is to treat what success/failure looks like based on the characters. Need some arcane runes translated, when the wizard succeeds the check they get a full translation, and a partiall translation on failure. The Barbarian succeeding means they recognize the symbols from a book their tribe's shaman had, they get even less of a translation then the wizard's failed check but they produced a lead, if they go see that shaman they'll be able to get the full translation. On failure the Barb gets nothing.

[RSP]

It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.


The Str 8 Wizard has a max lift of 230 lbs.

The Str 20 Barbarian has a max lift of 600 lbs.

How heavy is the portcullis? If it’s 600 lbs or less, why is the Barbarian rolling at all? And if it’s 600 lbs, are you okay with the Str 8 Wizard more than doubling their MAX lift weight?

The two characters shouldn’t have the same DC because they aren’t trying the same thing. If it’s 700 lbs, the Barb is trying to increase their max lift by about 17%. The Wizard is trying to increase their max lift by more than 200%.

That’s not the same task.

Nor is it the same task to recall arcane knowledge that you’ve studied before; versus trying to come up with arcane knowledge that you’ve never seen or studied before. the first is a memory thing, the second is trying to pull information from a picture for which you have no context.

It may be possible to get information from the latter exercise, but it’s not trying the same thing as remembering something: I can attempt to recall the words to a song I’ve heard before, but I can’t recall the words to a song I’ve never heard before (assuming I haven’t read the lyrics).

By understanding what the tasks are that are trying to be accomplished, you can eliminate the element of random die roll than you’re not liking.

[Chronos]

This is a complaint of mine, too. Honestly, I think that bounded accuracy is a mistake in general, but it's much worse for skills than it is for attack rolls.

Let's say that a rogue and a barbarian are both trying to complete an obstacle course side-by-side, and one of the obstacles is to pick a lock on a door, and another obstacle is to use a greataxe to defeat an orc. That's fair, right? Each class gets one obstacle that they're good at, and one obstacle that they're bad at.

Except that the door is one roll. Even though the rogue has a higher bonus to lockpicking, it's still not unlikely that the rogue will fail, or that the barbarian will succeed, or both, because the d20 is so much bigger than the bonus.

Meanwhile, in the fight against the orc, the barbarian might miss or the rogue might hit... on any single attack roll. But the fight is going to involve many attack rolls. Over enough rolls, the law of averages will eventually take over, and while it's still possible that the rogue wins the fight while the barbarian loses, it's much less likely.

[OldTrees1]

The Str 8 Wizard has a max lift of 230 lbs.

The Str 20 Barbarian has a max lift of 600 lbs.

The two characters shouldn’t have the same DC because they aren’t trying the same thing. If it’s 700 lbs, the Barb is trying to increase their max lift by about 17%. The Wizard is trying to increase their max lift by more than 200%.

That’s not the same task.

If you are ignoring the Str modifier to justify it not being the same task, then it is consistent to also ignore the Str modifier when calling for the check. Lifting that 700 lbs portcullis is a 1d20 check, not a Str check, if the DC already accounts for the different Str scores.

Alternatively you can acknowledge the Str modifier difference. Even at the same DC it is not the same task because the Str 8 has a -6 penalty relative to the Str 20. If you acknowledge the Str modifier difference, it already accounts for the difference in the tasks by having the required roll be higher.


Of course, if the ability check math is broken, then you can't rely on the math and you have to make excuses and fixes for it. For example inconsistently ignoring/acknowledging a modifier, or reducing the size of the 1d20 to a 1d8. (both are flawed answers, and one is mine)

[MadBear]

The Str 8 Wizard has a max lift of 230 lbs.

The Str 20 Barbarian has a max lift of 600 lbs.

How heavy is the portcullis? If it’s 600 lbs or less, why is the Barbarian rolling at all? And if it’s 600 lbs, are you okay with the Str 8 Wizard more than doubling their MAX lift weight?

The two characters shouldn’t have the same DC because they aren’t trying the same thing. If it’s 700 lbs, the Barb is trying to increase their max lift by about 17%. The Wizard is trying to increase their max lift by more than 200%.

That’s not the same task.

Nor is it the same task to recall arcane knowledge that you’ve studied before; versus trying to come up with arcane knowledge that you’ve never seen or studied before. the first is a memory thing, the second is trying to pull information from a picture for which you have no context.

It may be possible to get information from the latter exercise, but it’s not trying the same thing as remembering something: I can attempt to recall the words to a song I’ve heard before, but I can’t recall the words to a song I’ve never heard before (assuming I haven’t read the lyrics).

By understanding what the tasks are that are trying to be accomplished, you can eliminate the element of random die roll than you’re not liking.

I hear what you're saying, but if you listen to posters on the board, you'll hear people repeatedly decry the unfairness of having different DC's for different PC's.

[Sorinth]

I hear what you're saying, but if you listen to posters on the board, you'll hear people repeatedly decry the unfairness of having different DC's for different PC's.

If it's something that will work for your table then who cares what strangers on the internet have to say.

[Sorinth]

This is a complaint of mine, too. Honestly, I think that bounded accuracy is a mistake in general, but it's much worse for skills than it is for attack rolls.

Let's say that a rogue and a barbarian are both trying to complete an obstacle course side-by-side, and one of the obstacles is to pick a lock on a door, and another obstacle is to use a greataxe to defeat an orc. That's fair, right? Each class gets one obstacle that they're good at, and one obstacle that they're bad at.

Except that the door is one roll. Even though the rogue has a higher bonus to lockpicking, it's still not unlikely that the rogue will fail, or that the barbarian will succeed, or both, because the d20 is so much bigger than the bonus.

Meanwhile, in the fight against the orc, the barbarian might miss or the rogue might hit... on any single attack roll. But the fight is going to involve many attack rolls. Over enough rolls, the law of averages will eventually take over, and while it's still possible that the rogue wins the fight while the barbarian loses, it's much less likely.

Where does it say that if the Rogue fails to pick a lock that they can't retry?

[OldTrees1]

Where does it say that if the Rogue fails to pick a lock that they can't retry?

I believe they were talking about how a locked door requires 1 success while a fight takes multiple successes.

The Rogue needs, for example, 1 success vs the door and 5 successes with the greataxe
The Barbarian needs, for example, 1 success vs the door and 3 successes with the greataxe
Bounded Accuracy's downsides are exacerbated for skill checks because repetitive checks compensate for some of the downsides ("Over enough rolls, the law of averages will eventually take over").

[Sorinth]

I believe they were talking about how a locked door requires 1 success while a fight takes multiple successes.

The Rogue needs, for example, 1 success vs the door and 5 successes with the greataxe
The Barbarian needs, for example, 1 success vs the door and 3 successes with the greataxe
Bounded Accuracy's downsides are exacerbated for skill checks because repetitive checks compensate for some of the downsides ("Over enough rolls, the law of averages will eventually take over").

You can make a lock that requires 3 successes if that's what you want and I'd argue a trapped lock already does
1) Spotting the trap
2) Disarming the trap
3) Picking the lock

[OldTrees1]

You can make a lock that requires 3 successes if that's what you want and I'd argue a trapped lock already does
1) Spotting the trap
2) Disarming the trap
3) Picking the lock

@Chronos
I believe they meant this as an amended reply to your post

[Psyren]

I hear what you're saying, but if you listen to posters on the board, you'll hear people repeatedly decry the unfairness of having different DC's for different PC's.

If you listen to posters on the board then everything gets decried by somebody; that way lies madness. (The irony of your handle isn't lost on me.)

As Sorinth said, pick the solution that works best for you. Some us, like RSP, prefer to very finely chop the definition of "task." Some us, like me, prefer to use static DCs but gatekeep the check itself. Some of us, like OldTrees1, limit the variance of the die more sharply. And some of us don't mind incongruous results at all* and let the dice fall where they may. Pick whichever you think is the most fun and don't worry about those who disagree.

*or claim not to but then complain

[Goobahfish]

I think the title of the thread is misleading. I think this phrase:

"Tasks of any kind performed by characters of abilities which vary within a narrow range should always be determined by chance where the variation is more than double the usual range of character abilities, via the use of a D20."

Most people would probably agree that this isn't ideal. This framework, while simple creates some fairly unintuitive outcomes.

I don't think replacing 'D20' with 'insert other distribution' solves this problem, or even really improves it? There is an argument for using a lower dice with less variation (i.e. a D8, D10 or D12) which would make the difference between high score + prof and low score + no prof almost be enough to go from auto-pass to auto-fail.

However, I think that the general linearity of these mechanisms will still create unsatisfying outcomes.

Gatekeeping checks is a decent work-around.
Sliding DCs is a decent work-around.

I like the idea of consecutive successes vs consecutive failures. I.e., in combat (which is complex) there is a very chaotic interplay between successes and failures. To 'lose' combat, you basically need to roll horrendously such that the odds of it happening are pretty miniscule (outside DM error). However, because there are pretty good representations of 'degrees of success' (i.e., how maimed you becomes, how much resources is expended) there is still tension.

The problem I think is that some aspects of D20 checks, there is no 'middle ground'. There isn't really a good way of providing 'minor punishments'. Like... perhaps failing a jump check, you make it, but get scratched or injured as part of it. You fail an arcana check... (DM hurriedly works out some kind of 'interesting' mistake which isn't too punishing but still... that's it I give up, you fail).

I think, a good thing that should eventually be included in D&D is 'not codified skill challenges' but sort of 'different kinds of skill challenges'.

I.e., Some skill challenges have a score that each failure can reduce, each success can increase but if you survive 10 rounds with out getting to zero you succeed. A simple example for stealth. For this to work you need to have some 'choices' involved part way through but it can work. For climbing, survival etc this stuff can work well too.

You fail a survival check. You are attacked by a snake, one random character suffers D6 wounds and we move on. Next check. How smashed are you before you get to where you are going?

These kinds of things don't care very much about individual failures or successes, but the aggregate between them (like combat, which is nominally 'fun').

There is also some fancy maths stuff you can do with 'consecutive successes' vs 'consecutive failures' where the odds are highly non-linear.

For example, you want to open a lock. You choose how many dice you roll. You need at least two successes. If you roll three or more failures your character gives up and you can't try again. These challenges bias heavily towards proficient characters while making 'random success' for the barbarian much harder.

I guess the point is, making it a single dice roll (unless it is a very yes-no answer where both answers are interesting) is probably a bad idea, but it isn't hard to make a bit more interesting with a few more D20 rolls.


That's why I sort of liked the idea of moving more towards an 'accumulate sufficient score' versus 'pass/fail' model for things.

This gets closer to something reasonable.

[Lunali]

Honestly I feel like all of D&D would be equally well served by using coin flips instead of dice. Or, if you want to preserve the current balance, a d4, with 1 being a miss.

D&D is designed as an attrition based tactical combat simulator, everything beyond that is half-assed at best. Stick to the current rules if you just want to skip through everything but combat, switch systems if you want to give meaningful focus to anything else.

[RSP]

I hear what you're saying, but if you listen to posters on the board, you'll hear people repeatedly decry the unfairness of having different DC's for different PC's.

It’s not a problem for me: though I go with RP and prefer the character to matter more than die rolling, which hopefully comes across in how I play. I find people who argue about unfair DCs are people who want to attack DMs with rules, which isn’t how I feel the game should go. If you disagree with how a DM runs their table at such a functional level as skill checks, I’d suggest not playing at it, rather than constantly arguing over rules.


If you are ignoring the Str modifier to justify it not being the same task, then it is consistent to also ignore the Str modifier when calling for the check. Lifting that 700 lbs portcullis is a 1d20 check, not a Str check, if the DC already accounts for the different Str scores.

Alternatively you can acknowledge the Str modifier difference. Even at the same DC it is not the same task because the Str 8 has a -6 penalty relative to the Str 20. If you acknowledge the Str modifier difference, it already accounts for the difference in the tasks by having the required roll be higher.

It’s not ignoring the stat - in someways, quite the opposite. If you ignore what the stat represents (for instance, that the Barb can already lift 600 lbs, or that the Wiz maxes out at 240 lbs), that, to me, is ignoring the stat.

I don’t see a logical way to say “it’s DC 15 to lift the 500 lbs portcullis” for all characters, as that’s just ignoring what the stats represent. Would it be DC 15 for a Fire Giant to lift?

It seems odd to ignore facts about the characters as well to ignore that the tasks aren’t actually the same tasks.

And I imagine that those who are well trained and more advanced in a certain area, are more likely to successfully push their limits; than ones who aren’t. For instance, my experience is if you’re in shape and regularly lifting weights, you’re a lot more likely to succeed in increasing your lift, than someone who doesn’t (less likely to get injured too). So in that sense, to me, it makes sense that the modifier still helps.



Of course, if the ability check math is broken, then you can't rely on the math and you have to make excuses and fixes for it. For example inconsistently ignoring/acknowledging a modifier, or reducing the size of the 1d20 to a 1d8. (both are flawed answers, and one is mine)

Sure, though I don’t think the ability check math is broken: I like that it’s open for DMs to use as they see fit. Likewise, if a d8 works better for your table, that’s great and in-line with how I feel the game is meant to played.

[Psyren]

"Tasks of any kind performed by characters of abilities which vary within a narrow range should always be determined by chance where the variation is more than double the usual range of character abilities, via the use of a D20."

"Tasks of any kind" is the problem here. The d20 should really only come into play for a fairly narrow possibility space of actions, which is why the range of variance is also narrow. You don't roll a d20 to get out of bed or walk across a room, nor do you roll a d20 to jump to the moon or conquer every nation in the world. The vast majority of tasks don't need a d20 to resolve, which is a far cry from "tasks of any kind."

Even the UA says it: "When the outcome is uncertain and narratively interesting, the dice determine the results." Most things characters do in a day fail one or both of these litmus tests. And even for the things that clear both, what is both uncertain and narratively interesting for one character may not be both (or either) for another.

[Goobahfish]

"Tasks of any kind" is the problem here. The d20 should really only come into play for a fairly narrow possibility space of actions, which is why the range of variance is also narrow. You don't roll a d20 to get out of bed or walk across a room, nor do you roll a d20 to jump to the moon or conquer every nation in the world. The vast majority of tasks don't need a d20 to resolve, which is a far cry from "tasks of any kind."

Even the UA says it: "When the outcome is uncertain and narratively interesting, the dice determine the results." Most things characters do in a day fail one or both of these litmus tests. And even for the things that clear both, what is both uncertain and narratively interesting for one character may not be both (or either) for another.

Ha ha, perhaps I gave the false impression of a Strawman. The point I'm making is that the premise of the original post kind of has this as its underlying strawman. I.e. it is easily attackable.

The issue with 5e is that some tasks have different 'variability' to others but are all tacitly beholden to a D20 check (which has an interesting skill-vs-luck balance). The work-around is to have sliding DCs, gating checks (your example) etc.

This is all a way of the DM saying... I want you to always succeed, you to have a 1 in 10 chance of failure, you a 50/50 and you a 9 in 10 chance of failure (despite your apparent stat spread). To make that work, I need to give you a +4 bonus to the check, you are fine and you get a -7 penalty. It is a pity that D&D doesn't make this super easy (it relies on a lot of DM experience to get right).

Other systems with multiple dice/mechanics can give the DM more tools to do this more 'automatically' by requiring higher DCs OR more successes etc.

The biggest problem with gatekeeping checks (i.e. you auto-succeed) is it does take a bit of the wind out of the specialists' sails. If everyone auto-succeeds, having those extra points feels like a waste. If you alone auto-succeed but don't get anything 'extra' (like super-successes like rolling a nat 20) it can feel a little underwhelming too.

[Damon_Tor]

At my table, most skills that have any real importance require multiple rolls.

For example, when you talk about lifting a portcullis, that's not (necessarily) a one-roll task. The portcullis has 30 "DC Points" and it recovers 10 DC Points per round (as it lowers back down). So the barbarian will likely struggle for several rounds to get the thing open, slowly making progress on the task. But the wizard, even if he gets lucky a few rounds in a row, is far less likely to actually accomplish the task.

On the flip side, searching a room for clues to a mystery will similarly advantage the wizard, with 40 DC points, but rolls of 5 or lower turn up a red herring and a false solution. So the wizard with his high int score and investigation proficiency has no issue solving the puzzle, but the barbarian is likely to draw the wrong conclusion at some point in the task as he rolls for several rounds.

[GloatingSwine]

At my table, most skills that have any real importance require multiple rolls.

For example, when you talk about lifting a portcullis, that's not (necessarily) a one-roll task. The portcullis has 30 "DC Points" and it recovers 10 DC Points per round (as it lowers back down). So the barbarian will likely struggle for several rounds to get the thing open, slowly making progress on the task. But the wizard, even if he gets lucky a few rounds in a row, is far less likely to actually accomplish the task.

On the flip side, searching a room for clues to a mystery will similarly advantage the wizard, with 40 DC points, but rolls of 5 or lower turn up a red herring and a false solution. So the wizard with his high int score and investigation proficiency has no issue solving the puzzle, but the barbarian is likely to draw the wrong conclusion at some point in the task as he rolls for several rounds.

This sort of roll-per-round starts feeling a bit clunky if the passage of rounds doesn't matter to the task and there's no way to change what you're rolling.

If the Barbarian can work on the gate as long as they like, then do you sit there rolling until they do?

If the result of one roll can cause the players to change what they're doing in between rolls, then it works out, if not you might want to hack the whole system out and replace it with some kind of dice pool that lets you do the whole roll at once but still doesn't have the swinginess of a single D20.

[stoutstien]

This sort of roll-per-round starts feeling a bit clunky if the passage of rounds doesn't matter to the task and there's no way to change what you're rolling.

If the Barbarian can work on the gate as long as they like, then do you sit there rolling until they do?

If the result of one roll can cause the players to change what they're doing in between rolls, then it works out, if not you might want to hack the whole system out and replace it with some kind of dice pool that lets you do the whole roll at once but still doesn't have the swinginess of a single D20.

Basically what i use. 3d10 needing to pass 2 to succeed normally and and add/subtract one for disadvantage/advantage accordingly.
It differs from the usual roll 3 and take the middle one in that you are still checking to see if each one pass/fails so natural 1/20 still a thing.
Have to to keep DC low because of compounding odds but it's no different than those who use muti stage checks.

Need at least a 10 three different times on the D20 to pass? That's ~12% chance to be successful.

[GloatingSwine]

An alternative would be doing something like attribute bonus + proficiency bonus as your dice pool, every even roll is a success*, and you then set the difficulty of the task by how many successes they need.

Use a coin flip calculator to work out the success percentages so you know how many successes to call for for different difficulties of task.

Instead of flat advantage or disadvantage you would have stages of it which let you reroll fails or make you reroll successes.

* This means you don't need to be picky about what dice you roll as long as they have an even number of faces, uniform smaller dice like D6 will be easier though.

[Amechra]

This is one of the reasons that I've become increasingly unsatisfied with skills just being numerical bonuses to doing stuff.

One thing I've considered doing is flavoring success/failure based off of whether or not the character in question is proficient in what they're trying.

The Wizard fails an Arcana check to while researching something in a library and the Barbarian succeeds? Someone miss-filed all of the relevant books, and the Barbarian happened to stumble on them while browsing.

The Wizard passes the Athletics check to lift a gate that the Barbarian failed earlier? Well, obviously the Barbarian loosened it for them (https://www.youtube.com/watch?v=oa8i9m_1N9I).

[PhoenixPhyre]

An alternate to requiring multiple successes for a single task that provides much more granularity while working with the current system is...stop doing single-point-of-failure/success situations.

Just like combat isn't made of a single attack roll (success = win, failure = lose), interesting situations involving ability checks are generally made of more than one ability check. Ie more than one "task". Each one moves the situation's "progress bar" along in narrative ways on both a success or failure, but no single check is make/break all by itself.

Basically zoom out a bit. Yes, each individual sub-component is relatively even probabalistically (without gatekeeping, etc). The non-linearity comes from having the combination of all of those checks, succeeded or failed, mean something. And the set of tasks isn't determined in advance--after each attempt (which either moves the party notably forward or causes a complication), the party can choose something else. But can't really retry that individual attempt without cost, as the situation has changed.

Consider "getting through the vault door" as the situation. If the only thing standing in the way is the lock itself...that's not an interesting situation. It's binary, one success is enough and it can be retried. Failing it doesn't meaningfully change the situation. Now consider the situation if
* there are multiple access controls that need to be hit in a particular order, based on clues elsewhere. Taking too long to figure it out means guards will patrol by.
* Failing one access control sets off an alarm, reducing the time until guards arrive and increasing the response.
* You can take extra time with one part, making another easier to bypass...at the cost of burning time.
* No individual access control can be retried in isolation, and one person can't do all the access controls themselves.

Now you have a situation where multiple actions (small a) are needed by the entire party to attempt this, and there are multiple ways to get through it.

------

Or instead of "task is convince the king to send help", you zoom out and consider the actual situation, where the king isn't a sole dictator and so you can make things easier by
* convincing the head general that it's a serious issue
* soothing the finance minister (who really wants the money involved with mustering the troops for something else)
* getting the king's mistress to put pressure on the king (whether by blackmailing her or by buttering her up), etc.
And each of these, depending on how you approach them, modify the scene with the king. And all of which may require some behind-the-scenes research (either skullduggery or book research) that can involve the non-faces. And the general might respond much better to the gruff, no-nonsense, no-patience-for-small-talk fighter or barbarian vs the smooth-talking bard (who's much better at convincing the more court-oriented folks including the king).

------

Basically, think of things as scenes, not as individual tasks. Tasks are components of scenes, but there are many possible ones. And anemic scenes (those with only one or very few tasks) are boring scenes. They may happen, but it's fine to just narrate through them/auto-succeed (giving the success to the most narratively-suited person).

Counting "event HP" or "successes needed" (ie must succeed X times to succeed at a task) just slows things down and focuses on the individual tasks, which are inherently boring (just like individual actions in combat are generally fairly boring by nature). The interesting things happen when each task is quick and produces a small result and those small results aggregate into interesting results as you react to them/as the situation snowballs one way or another.

[OldTrees1]

An alternate to requiring multiple successes for a single task that provides much more granularity while working with the current system is...stop doing single-point-of-failure/success situations.

Just like combat isn't made of a single attack roll (success = win, failure = lose), interesting situations involving ability checks are generally made of more than one ability check. Ie more than one "task". Each one moves the situation's "progress bar" along in narrative ways on both a success or failure, but no single check is make/break all by itself.

Basically zoom out a bit. Yes, each individual sub-component is relatively even probabalistically (without gatekeeping, etc). The non-linearity comes from having the combination of all of those checks, succeeded or failed, mean something. And the set of tasks isn't determined in advance--after each attempt (which either moves the party notably forward or causes a complication), the party can choose something else. But can't really retry that individual attempt without cost, as the situation has changed.

Good, possibly already implemented, advice. It does not change how luck dominates each task, however it does salvage the situation.

Even for those that still have issues with the sub-components (tasks), we can (and possibly are do) use this advice.

[PhoenixPhyre]

Good, possibly already implemented, advice. It does not change how luck dominates each task, however it does salvage the situation.

Even for those that still have issues with the sub-components (tasks), we can (and possibly are do) use this advice.

A thought--luck (random chance) dominates everything if you take too granular a view. Even in the real world. Just like statistical mechanics focuses on the macrostates, not the microstates (where lots of really weird things can happen, just with low probability), D&D works better if you just let tasks be and not put too much weight on any one of them. Or saving throws (one reason save-or-die/save-or-lose is so awful is that it breaks this). Or attack rolls.

D&D works best (in my experience) on the model of "lots of little rolls, each with small but real effects". Turns flow better if people are generally just doing a simple action and then letting someone go, with complexity arising out of the combination of actions, not out of any individual action.

[OldTrees1]

A thought--luck (random chance) dominates everything if you take too granular a view. Even in the real world. Just like statistical mechanics focuses on the macrostates, not the microstates (where lots of really weird things can happen, just with low probability), D&D works better if you just let tasks be and not put too much weight on any one of them. Or saving throws (one reason save-or-die/save-or-lose is so awful is that it breaks this). Or attack rolls.

D&D works best (in my experience) on the model of "lots of little rolls, each with small but real effects". Turns flow better if people are generally just doing a simple action and then letting someone go, with complexity arising out of the combination of actions, not out of any individual action.

Sigh. I understand your intent, but you came across as not listening to why some dislike how much the 1d20 dominates the ability check math. Can you articulate why I (or someone else) dislikes how much luck dominates 5E checks at the task level despite using multiple tasks to salvage the situation level?

Saying "D&D works better if you assume your characters can't be skilled at tasks, only parties being skilled at situations" is not helpful. (Not that you said that, but that is what your comment about everything is complete luck at a small enough scale translates to when the concern was about luck vastly dominating skill at the task level). I like to play characters that are characterized by skilled competence. That extends to tasks where the outcome is uncertain but the growing skill of the character makes a significant impact letting them get significantly better at old tasks and grow into new tasks. Not "No, tasks are dominated by luck. Skill is negligible." When success/failure at a task consistently reflects luck vastly more than it reflects skill, it taints both outcomes for a skilled character when it comes to their contribution (pass or fail) to the situation level. Their pass is "lucky, but not good" and their fail is "unlucky, and still not good".

I agree D&D works best on a model of "lots of little rolls, each with small but real effects", but those little rolls are important too. This is why I agreed that your advice of having situations be several checks is good (possibly already utilized) advice even for those with concerns about the task level that persist after applying the advice.

[PhoenixPhyre]

Sigh. I understand your intent, but you came across as not listening to why some dislike how much the 1d20 dominates the ability check math. Can you articulate why I (or someone else) dislikes how much luck dominates 5E checks at the task level despite using multiple tasks to salvage the situation level?

Saying "D&D works better if you assume your characters can't be skilled at tasks, only parties being skilled at situations" is not helpful. I like to play characters that are characterized by skilled competence. That extends to tasks where the outcome is uncertain but the growing skill of the character makes a significant impact letting them get significantly better at old tasks and grow into new tasks. Not "No, tasks are dominated by luck. Skill is negligible." When success/failure at a task consistently reflects luck vastly more than it reflects skill, it taints both outcomes for a skilled character when it comes to their contribution (pass or fail) to the situation level. Their pass is "lucky, but not good" and their fail is "unlucky, and still not good".

I agree D&D works best on a model of "lots of little rolls, each with small but real effects", but those little rolls are important too. This is why I agreed that your advice of having situations be several checks is good (possibly already utilized) advice even for those with concerns about the task level that persist after applying the advice.

Two things. First, the point you replied to was an extension, a springboard. Not really attempting to contradict.

Second, it seems that you're taking "skilled competence" to mean "basically never fails at things I'm good at" and additionally that anyone who isn't skilled in that thing basically never succeeds (in your preferred model). That's...binary beyond my belief. Anything of that worth actually pulling out dice for is something uncertain at roughly a 50% level (+- a lot). Anything that a character (based on the fiction) shouldn't plausibly fail at (or should only plausibly fail very infrequently) isn't actually something that the Ability Check is designed to model. Those are thresholded out way earlier in the process as auto-success. Anything that a character can't (in keeping with the fiction) realistically succeed at have also been thresholded out as auto-failures. The only tasks for which it's even worth considering a DC and an ability check at all are ones where there's already been determined to be a substantial, meaningful, real chance of both failure and success AND where both of those states are interesting and valid in the fiction.

Trying to use the existing system to simulate all tasks is where things seem to be going wrong. It's not designed for that. It's an uncertainty resolution system, not a "simulate the state of the world" system. And yes, the same goes for attack rolls and saving throws. If the fiction says that success is impossible or overwhelmingly likely, the attack roll/saving throw just doesn't happen. Those states are much rarer than similar things for ability checks, but they do happen and that's playing as intended.

For me, skilled competence is more about
* a baseline of stuff that just happens, no check, no invocation of rules. The fiction says that the only plausible outcome is your success. So it just happens.
* reliability averaged over lots of small tasks. Individual tasks don't matter--everyone bar everyone can stuff something up and pull a dumb move. Even the brightest and most skilled. But the most skilled do it less when you look at the whole stream of things. Even grandmaster chess players make boneheaded mistakes at times. Just...a lot less than non-grandmasters.

And for both of those, D&D 5e works just fine as written.

Beyond that, D&D 5e characters are not specialists, despite specializing. They're all generalists with a bit of focus. Everyone is an adventurer first and foremost, a <X> second. And the reverse is true--5e does not support "I'm illiterate and know nothing about anything academic" as a character type. You can play it, but it's bucking the system's expectations. Nor does it support "I'm the foremost expert in <X>". Or even an academic expert in anything. Except adventuring. Yes, even PC wizards aren't academics. They may have started out as such, but that's in the past.

On a scale of 1-100, PCs range from 40-70 at just about everything. Neither world-class experts (in narrow niches) OR bumbling fools.

[OldTrees1]

Two things. First, the point you replied to was an extension, a springboard. Not really attempting to contradict.
I apologize. I thought it was a reply to concerns about luck being too dominant at the task level. I did not understand you spring boarded further away.


Second, it seems that you're taking "skilled competence" to mean "basically never fails at things I'm good at"
...
and additionally that anyone who isn't skilled in that thing basically never succeeds (in your preferred model).
Nope, that is a completely incorrect summary. They can still fail a check anywhere from 5% to 95% depending on the task (assuming we ignore the auto pass/fail cases since they are not checks). It is all about the numbers that communicate skill and growth in skill being too small relative to the 1d20. 19 levels is worth roughly 33%-50% of the RNG. Why have a 20th level character try the task instead of a 1st level character (they can probably do it too). They are basically the same and the 1st level character costs less. The difference from unskilled and negative talent vs skilled and talented is also roughly 33% of the RNG. That is much better because they would be closer, but it can be a concern too. It really is more the former (lack of growth) than the latter.

Honestly it is like Flaming Sphere being a 9th level spell.



Ultimately your answer was "5E is not about skilled characters. Any concerns raised from trying to play one in 5E are concerns to be ignored." Well I guess that is better than coming off as not listening. It does not resolve my concern but it concludes your need to consider it.

[Goobahfish]

Ya... variance of a D20 seems high some times.

I'm really leaning towards systems where you choose how many D20s you roll and having a 'you need X successes' but 'avoid Y problems' for some tasks.

With the safe example... before considering how that works, how does it work out in real life.

#1: They open the safe easily.
#2: They open the safe but waste a lot of time.
#3: They fail to open the safe because they give up
#4: They fail to open the safe because they are insufficiently skilled

With the alternate conditions

#5: They make it really obvious that the safe was tampered with
#6: They make it impossible for anyone to open the safe ever

There might be a few others I'm missing but I'll start here.

#1 (??) & #4 (You would need a '21') are pretty easy to adjudicate in 5e. I left the ?? because how you do this is still a bit up in the air.

If you assume infinite time, you could replace ?? with a passive 'take 20' style check.
You could adjudicate a #1 vs #2 via a check.

#3... is a lot harder? It is a real phenomenon. The problem is, at table it is a pretty boring... keep rolling. That is it, I give up. Also, players knowing the difference between #3 & #4 is pretty dubious. That should be hidden knowledge.

If you let the players choose how many dice, you are letting them have some control over 'desperation to do the thing' and at the same time speeding up play (counting 16+'s is a pretty straightforward).

From this setup, you can then easily adjudicate:
1) How many successes were necessary at what DC
2) How 'long' it took the player
3) How many 'complications' occurred
4) When the player 'gave up'.

I'm toying with complications being '1s and 2s' for non proficient and '1s' for proficient. Complications can be... autofail (cause you break the lock) or 'loud noises' => extra encounters or 'broken tools' etc.

Stats wise, this will really benefit trained characters because the relevant metric is the ratio between complications and their 'success chance'. I.e. if you need a 11+, that is a 1 to 10 ratio of success. Whereas the barbarian with lock pick tools gets... 16+ with a double fail chance getting you a 2 to 5 ratio... which is significantly worse.

You could even have 1 complication =>, 2 complications => etc where 2 complications is 'it is now broken'.

This way we can keep the D20, decrease the 'luck factor' and let the proficient characters 'shine' without massively slowing down play.

[KorvinStarmast]

Anything of that worth actually pulling out dice for is something uncertain at roughly a 50% level (+- a lot). {snip} The only tasks for which it's even worth considering a DC and an ability check at all are ones where there's already been determined to be a substantial, meaningful, real chance of both failure and success AND where both of those states are interesting and valid in the fiction. Using an older term, prevent over reliance on roll playing.

It's an uncertainty resolution system, not a "simulate the state of the world" system. And yes, the same goes for attack rolls and saving throws. If the fiction says that success is impossible or overwhelmingly likely, the attack roll/saving throw just doesn't happen. Those states are much rarer than similar things for ability checks, but they do happen and that's playing as intended. FWIW, a bog standard Orc, or any PC with a wisdom of 11 and not proficiency in Wisdom saves, cannot pass the DC 21 Wis Save for the Ancient Red Dragon's frightening presence. And that makes sense from a fiction perspective. Ancient Red Dragons are supposed to be scary.

Frightful Presence. Each creature of the dragon’s choice that is within 120 feet of the dragon and aware of it must succeed on a DC 21 Wisdom saving throw or become frightened for 1 minute A creature can repeat Blacks and greens are a 19, a blue is a 20, so while it's unlikely for the sample creatures to save, it's possible. (And I am not sure an ancient blue shouldn't also have a 21, TBH).

5e does not support "I'm illiterate and know nothing about anything academic" as a character type.
Correct. My barbarian player on Wednesday night questioned me on "can I actually read that?" and I replied with "unless modified, having a language proficiency includes reading and writing in that language." (He can speak Common and Giantish).

Ya... variance of a D20 seems high some times. That's right. Swinginess is embraced. See the forward to the PHB, written by Mike Mearls. The Dice Can Be Fickle.

[ruy343]

I'm stepping away from the current discussion direction to ask: Why are we assuming that player characters fail when they don't manage to hit the arbitrary DC? If my time with PBTA games has taught me anything, it's that the GM should come up with a way to take a failure and let it continue the story forward, rather than prevent the character from accomplishing something meaningful.

Example: Rogue is picking a locked door that the party needs to get through in a dungeon (no other way in). Rogue fails the check. Now what? The rogue rolls again? No! Else there's really no point in rolling! The GM should allow the door to open, but it takes long enough in-game that something else happens. Instead, the failed roll means that a wandering monster comes by, the monsters on the other side set an ambush, The BBEG is alerted through a spell, etc.

Example: The Wizard rolls Arcana to understand the runes in the sanctum. He rolls and fails abysmally. Now, the GM was kind of hoping that the players would get this information to provide a hook for next session but... now what? Make another character roll? Instead, let the Wizard translate, following the words and the flow of power between the runes with his finger, muttering "The path to the land of the dead is in the... Swamp...?" and right as he says the last word POP, he rolls constitution to avoid being turned into a frog for an hour... or he takes 3d6 lightning damage for accidentally merging the power of the runes at the tip of his finger... or maybe he has to expend a spell slot to power on the runes in order to understand then by triggering their power.

I think that skills, as written, are fine. We, as GMs, need to simply get more creative than "you fail".

[Easy e]

As a player, it really does piss me off when I spend Expertise in a certain area, and then when I get to use it a bad roll makes me fail. If it is an unskilled check and I fail, no big deal; but when I fail to live up to my own character archetype ..... why did I bother specializing again?

Plus, often times failure on a skill check just halts the game. If you are looking for clues or leads, and you fail to find any..... ummmm.... now what? You need to get past a door and you fail..... ummm..... now what? You fail to see an illusion and believe there is no path forward...... ummmm...... now what?

D&D focuses way too much on failure in an effort to create challenge for the players, but the challenges this leads to are super shallow. These sorts of "failure prone" mechanics thanks to swingy d20s leads to players taking a 3 hour sessions to cross a stream, I know because I have played in such sessions.

As a GM, I have moved to if you have the skill and make a check it is not if you succeed or not as you will succeed. The question becomes how long and how well is it done. For example, you are making an arcana check to determine what the runes say. You fail. That doesn't mean you can not translate them, it means it takes you half a day to figure it out instead of an instant. Now, what could have happened in that half a day?

[Rukelnikov]

As a player, it really does piss me off when I spend Expertise in a certain area, and then when I get to use it a bad roll makes me fail. If it is an unskilled check and I fail, no big deal; but when I fail to live up to my own character archetype ..... why did I bother specializing again?

Plus, often times failure on a skill check just halts the game. If you are looking for clues or leads, and you fail to find any..... ummmm.... now what? You need to get past a door and you fail..... ummm..... now what? You fail to see an illusion and believe there is no path forward...... ummmm...... now what?

D&D focuses way too much on failure in an effort to create challenge for the players, but the challenges this leads to are super shallow. These sorts of "failure prone" mechanics thanks to swingy d20s leads to players taking a 3 hour sessions to cross a stream, I know because I have played in such sessions.

As a GM, I have moved to if you have the skill and make a check it is not if you succeed or not as you will succeed. The question becomes how long and how well is it done. For example, you are making an arcana check to determine what the runes say. You fail. That doesn't mean you can not translate them, it means it takes you half a day to figure it out instead of an instant. Now, what could have happened in that half a day?

If the DM is not willing to deal with the PCs failing, then he shouldn't be asking for a roll. More than once players at my games have had to abandon pursuing a given dungeon because they just couldn't get past a barrier, they sometimes come back afterwards with a plan, but not necessarily, I remember a riddle statue they couldn't figure, and a door they couldn't open, but there were likely others that don't come to mind.

Do not be afraid of the PCs failing. Failures make successes more relevant.

[stoutstien]

As a player, it really does piss me off when I spend Expertise in a certain area, and then when I get to use it a bad roll makes me fail. If it is an unskilled check and I fail, no big deal; but when I fail to live up to my own character archetype ..... why did I bother specializing again?

Plus, often times failure on a skill check just halts the game. If you are looking for clues or leads, and you fail to find any..... ummmm.... now what? You need to get past a door and you fail..... ummm..... now what? You fail to see an illusion and believe there is no path forward...... ummmm...... now what?

D&D focuses way too much on failure in an effort to create challenge for the players, but the challenges this leads to are super shallow. These sorts of "failure prone" mechanics thanks to swingy d20s leads to players taking a 3 hour sessions to cross a stream, I know because I have played in such sessions.

As a GM, I have moved to if you have the skill and make a check it is not if you succeed or not as you will succeed. The question becomes how long and how well is it done. For example, you are making an arcana check to determine what the runes say. You fail. That doesn't mean you can not translate them, it means it takes you half a day to figure it out instead of an instant. Now, what could have happened in that half a day?

That is how the rules work already. That's One reason the whole 'fail to pick a lock' meme is so bad. Once the DM determines you can do it it's not a matter of pass/fail as much of how long will it take and how smooth the attempt is. You don't fail to knock down a door or just takes you 2 kicks instead of one. Sure occasionally something will come up where you really only have one instant that you can make an attempt but hopefully your DM understands probability and doesn't set the DC in the stratosphere.

[Damon_Tor]

If the Barbarian can work on the gate as long as they like, then do you sit there rolling until they do?

If there's no time pressure and there's no consequence of failure, the roll can be waived. Sometimes I'll ask for one roll and use the result to tell them how long the task takes them, but this is performative, again, unless there's time pressure.

[Sorinth]

If there's no time pressure and there's no consequence of failure, the roll can be waived. Sometimes I'll ask for one roll and use the result to tell them how long the task takes them, but this is performative, again, unless there's time pressure.

This is a good approach that I use as well, failure just means it will take 10 times longer then normal and I give the player a choice on whether to spend that time or not. For example failed to pick the lock "You are really struggling with this lock, you know you can do it, but it's going to take some time. Do you want to keep at it?" They don't know how/if taking the extra time will impact things, sometimes it does, sometimes it doesn't.

Another option is the success at a cost model for failing the skill check. So a failed pick locks might just mean they end up making some noise, if there are people on the other side of the door they've almost certainly been alerted to the fact someone was doing something with the door.


It's fine to make failure a failure when it's not important, ie a failed Investigation check means they don't find the secret compartment that has some nice loot. But when the McGuffin is hidden in that secret compartment then my view is that it's the DMs job to push the story forward in someway such as it took so long that now there's an extra encounter as you make your way out.

[Psyren]

That is how the rules work already. That's One reason the whole 'fail to pick a lock' meme is so bad. Once the DM determines you can do it it's not a matter of pass/fail as much of how long will it take and how smooth the attempt is. You don't fail to knock down a door or just takes you 2 kicks instead of one. Sure occasionally something will come up where you really only have one instant that you can make an attempt but hopefully your DM understands probability and doesn't set the DC in the stratosphere.

^ On top of this, failure doesn't have to mean you make no progress; it can just as easily be "progress combined with a setback determined by the DM" (PHB 174). Your lock specialist failing vs. a tough lock might mean they got through the lock but damaged their pick, so subsequent locks take longer until you can replace your thieves' tools. Or it can mean you get through the lock but there's a loud noise and now you have to find a place for the party to hide until a nearby sentry leaves. Or you got through the lock by damaging it, making it impossible to lock behind you, so there's a chance a patrol will come across the party while they're searching the next room. The possibilities are as endless as the DM's imagination and none of the examples I proposed will make the lock specialist feel like they wasted their time becoming a lock specialist.


If there's no time pressure and there's no consequence of failure, the roll can be waived. Sometimes I'll ask for one roll and use the result to tell them how long the task takes them, but this is performative, again, unless there's time pressure.

^ This is in the books too. "Roll first, fail, and then decide to take 10x as long to automatically succeed" is right on DMG 237. You can even do this without the initial failure.

[Raven777]

the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Just chiming in to mention that Dungeons & Dragons is still, fundamentally, a dice rolling game. Letting dice rolls "procedurally" determine story beats is a feature, not a bug. Maybe the Wizard legitimately never heard about this ritual, while the Barbarian heard a campfire tale about it. Maybe the Barbarian is legitimately lifting the wrong way, while the Wizard spots the protrusion in the frame where the porticulis gets stuck and is able to shake it free. The result of the dices are fundamental; but so is how you frame them in the collaborative storytelling act that is playing the game.

[Psyren]

Just chiming in to mention that Dungeons & Dragons is still, fundamentally, a dice rolling game. Letting dice rolls "procedurally" determine story beats is a feature, not a bug. Maybe the Wizard legitimately never heard about this ritual, while the Barbarian heard a campfire tale about it. Maybe the Barbarian is legitimately lifting the wrong way, while the Wizard spots the protrusion in the frame where the porticulis gets stuck and is able to shake the door free. The result of the dices are fundamental; but so is how you frame them in the collaborative storytelling act that is playing the game.

Alternatively, deciding not to procedurally determine story beats with dice is a feature too. The beauty of D&D is that it can accommodate both styles.

[RSP]

Just chiming in to mention that Dungeons & Dragons is still, fundamentally, a dice rolling game. Letting dice rolls "procedurally" determine story beats is a feature, not a bug. Maybe the Wizard legitimately never heard about this ritual, while the Barbarian heard a campfire tale about it. Maybe the Barbarian is legitimately lifting the wrong way, while the Wizard spots the protrusion in the frame where the porticulis gets stuck and is able to shake it free. The result of the dices are fundamental; but so is how you frame them in the collaborative storytelling act that is playing the game.

Why is noticing that something is blocking the portcullis a Str check?

[Psyren]

Why is noticing that something is blocking the portcullis a Str check?

This is exactly why I hate procedural generation :smallbiggrin: again though, I recognize that it can be a valid playstyle for some.

[Raven777]

Why is noticing that something is blocking the portcullis a Str check?

Because I had limited time to craft the perfect example. Would you like to help me find a better one? I'm sure we can think of something. Maybe the Barbarian pulled at an angle that got stuck, while the Wizard was lucky and pulled just right on their own attempt? They do not have to notice the protrusion, just be the one who managed to pull around it. And that element of luck is the dices.

But Psyren is right, these justifications do not have to be the only way to resolve check results. Maybe the failure still opens the portculis, but it gets jammed halfway and the group has to crawl underneath; that's gonna be an interesting extra wrinkle when they run away from the hydra hiding in the next room. Maybe the failure doesn't open the portculis, but there's another way around in the dungeon.

My point is that we're telling a story. The dices are there to add some variation to the way similar situations can resolve. If it was a video game, I'd say they add replay value.

[Rukelnikov]

Just chiming in to mention that Dungeons & Dragons is still, fundamentally, a dice rolling game. Letting dice rolls "procedurally" determine story beats is a feature, not a bug. Maybe the Wizard legitimately never heard about this ritual, while the Barbarian heard a campfire tale about it. Maybe the Barbarian is legitimately lifting the wrong way, while the Wizard spots the protrusion in the frame where the porticulis gets stuck and is able to shake it free. The result of the dices are fundamental; but so is how you frame them in the collaborative storytelling act that is playing the game.


Alternatively, deciding not to procedurally determine story beats with dice is a feature too. The beauty of D&D is that it can accommodate both styles.


Why is noticing that something is blocking the portcullis a Str check?


This is exactly why I hate procedural generation :smallbiggrin: again though, I recognize that it can be a valid playstyle for some.

But this is not prodcedural generation, this is retcon generation, you are rewriting the world to adjust for the roll, that feels like cheating in a way.

[Psyren]

But this is not prodcedural generation, this is retcon generation, you are rewriting the world to adjust for the roll, that feels like cheating in a way.

I view procedural generation as the umbrella term for that approach to world design, and proactive vs. retroactive generation as subsets of that term.

To play devil's advocate, the benefit of retroactive generation is that you can leave things up in the air until after the players try certain things. There's no need to roll to determine the state of any (let alone every) portcullis or set of manacles in the world until the players try to bypass them. This plays into one of the strengths of tabletop - a TTRPG is not like a CRPG where all the code has to be programmed in and compiled before the game is booted up, you can wait until a given challenge is "on-screen" before making a determination as to possibility.

The difference is that I view retroactive generation as being too hands-off. If I'm determining whether it's even possible for the wizard to lift a portcullis, I'd rather determine that prior to them rolling, than to call for a roll and justify the number after the fact. That allows me to avoid a cascade of "well the wizard lifted it while your barbarian failed because... uh... they noticed a protrusion you didn't." "Wait, but I have perception proficiency and they don't, how did they notice something like that while I didn't?" "Uh, wait, did I say protrusion? I meant to say... you lifted it at a bad angle and they lifted it at the right one, yeah that's it." "Bad angle? I didn't say what angle I was lifting it at. And it's a portcullis, it can only go up and down anyway?" "Look, they rolled better than you, deal with it, we're moving on!"

[Rukelnikov]

I view procedural generation as the umbrella term for that approach to world design, and proactive vs. retroactive generation as subsets of that term.

For me pro-cedural comes from pro-active, but w/e I never checked the definition.


To play devil's advocate, the benefit of retroactive generation is that you can leave things up in the air until after the players try certain things. There's no need to roll to determine the state of any (let alone every) portcullis or set of manacles in the world until the players try to bypass them. This plays into one of the strengths of tabletop - a TTRPG is not like a CRPG where all the code has to be programmed in and compiled before the game is booted up, you can wait until a given challenge is "on-screen" before making a determination as to possibility.

But you don't do that in procedural generation either, I won't roll to decide the state of the manacles, until the state of the manacles comes into question. Once they are gonna be interacted with, I roll to determine their state, in an old unkempt dungeon the variance will be higher than in a small town's barrack's jail, and in a major city's dungeon, I may not roll at all and just decide that they are in pristine condition. Once the state of the object to interact with has been decided, now the object is interacted with. Yes, it is slower than retcon generation, but its also fairer, since if the manacles were in poor condition, or coated in acid, the creature interacting with it, could maybe notice such things before attempting whatever it is they were gonna attempt.


The difference is that I view retroactive generation as being too hands-off. If I'm determining whether it's even possible for the wizard to lift a portcullis, I'd rather determine that prior to them rolling, than to call for a roll and justify the number after the fact. That allows me to avoid a cascade of "well the wizard lifted it while your barbarian failed because... uh... they noticed a protrusion you didn't." "Wait, but I have perception proficiency and they don't, how did they notice something like that while I didn't?" "Uh, wait, did I say protrusion? I meant to say... you lifted it at a bad angle and they lifted it at the right one, yeah that's it." "Bad angle? I didn't say what angle I was lifting it at. And it's a portcullis, it can only go up and down anyway?" "Look, they rolled better than you, deal with it, we're moving on!"

Exactly, which is why it feels like cheating.

[NichG]

Even within the sphere of 'procedural generation' or 'retcon generation', I think it makes a difference to the coherency of the result what the causal factors are that prompt things to be generated. The potential issue with using one character's dice roll to alter an element of the scene is that the act of one character rolling can change the affordances of other characters in interacting with that same element. If you are 100% sure that no other character will ever interact with that element again, this basically doesn't matter, but for things which are going to be re-visited multiple times it can create unintended consequences and even weird meta-game strategies which if you're running a 4th wall breaking game can be really cool peeks around the curtain, but would be annoying in a more serious campaign.

For example, if that portcullis falls back down again and the party has to get through it a second time, the fact that you introduced a particular special handhold to explain away the weak character lifting it when the strong character failed means that handhold should still be there, meaning that the strong character should be able to trivially lift it this time. Or, the fact that the strong character failed to lift something that is similar to things they lifted in the past might mean the GM introduces extra factors like 'this gate is rusty' which then make it harder for other characters to succeed. So from a metagame perspective, this means that it matters who tries first. It's safer to have the characters with a lower chance of success try first because if they succeed it will introduce factors to make the task easier for everyone, whereas if they fail it doesn't suggest the need for explanatory factors that would increase the difficulty for others.

Note this can also be an issue even without multiple characters, but when rolls are used to determine things where the variance of the roll is mismatched to the narrative variance in ability the character would be expected to demonstrate. Even if we just have the strong character trying to lift an object - if in some cases that is just look up their lift weight and see whether the object is heavier or lighter, but in other cases its seen as an obstacle which must be rolled past, then in the latter case that added variance can end up being injected into the environment. Meaning that again there may be a metagame consideration about whether or not to back down upon being told there is a roll involved.

On the other hand, if retcon/procedural generation is not used reactively to justify a dice result, but is instead used when something about the direction of play necessitates those details being created, then the effects of metagame causality are much more controlled, because that procedural generation can prioritize self-consistency with other established details rather than being pinned to a particular end that now needs to be explained. You can still get in trouble if for example you're filling in those details in a way that is informed by the current state of the PCs - entering a zone and making the GM detail it at low levels, then returning at higher levels to force it to not rubber-band to the party's power level for example. But as long as the procedural generation isn't being based off of the PCs, that sort of meta-causality will be pretty weak.

[Segev]

With the conversation around skill checks and what is a reasonable DC to set, it brought me back to a fundamental question about how we go about these skill checks. Why do we let luck have a bigger deciding factor on success then the amount of training or skill involved?

With the reduction in numbers that make it so no one is rocking a +45 to a skill check (characters in my old 3.x days) this has made it so that the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Often the "fixes" to these issues involve not letting players roll if they're not trained in a skill, or setting the bar high enough that only those trained can succeed, or having different DC's for different characters. (I'm not saying these fixes are good, only that this is what many DM's often end up doing).

This bring back the question, of how do we represent chance without it overshadowing training, and natural skill (proficiency + Ability score).

My gut is to help reduce the standard deviation via using 2d10 or 5d4's. In this way, characters are far more likely to roll the average then they are to roll extremely well or extremely low.

What are your thoughts? Should we continue using the d20 and move on, or is there a better way?If something is so obvious that it should go one way or another, you shouldn't roll for it at all. Just declare it goes the obvious way. If something has to go one way to have the game be interesting, don't roll; it goes that way. If these are in conflict, then you probably should go with the obvious way rather than the interesting one, and chalk it up as a learning experience on what your PCs can do. Or you can go ahead and roll when those are in conflict and let the dice - random as they are - sort it out. Maybe the inobvious happened and made the interesting thing happen. Maybe the obvious happened and the boring solution saved the day.

But you should only be rolling when the randomness of the dice is in some way desirable. Whether for verisimilitude, for intrigue, to resolve a conflict of verisimilitude and intrigue, or just because the outcome is truly unclear and could be interesting either way (the classic time to call for a check).

Because this may sound puzzling to those following my debates with Psyren on the subject, it is important to note that, once you've decided to roll and set a DC, you should let everybody try it. You have clearly decided that either outcome is interesting, that there is some doubt as to the success of it, or that you have to pick between "obvious outcome" and "interesting outcome" and want to let the dice decide, so yes, you should let the untalented, untrained person try the DC 18 check just as much as the talented expert is allowed to try it. If you only want the talented expert to succeed, make him succeed without needing a roll. Once you decide a task's DC, that's how hard it is. That is not how hard it is for a given person; how hard it is for a given person is determined by their bonuses or penalties to the roll.

[Rukelnikov]

Even within the sphere of 'procedural generation' or 'retcon generation', I think it makes a difference to the coherency of the result what the causal factors are that prompt things to be generated. The potential issue with using one character's dice roll to alter an element of the scene is that the act of one character rolling can change the affordances of other characters in interacting with that same element. If you are 100% sure that no other character will ever interact with that element again, this basically doesn't matter, but for things which are going to be re-visited multiple times it can create unintended consequences and even weird meta-game strategies which if you're running a 4th wall breaking game can be really cool peeks around the curtain, but would be annoying in a more serious campaign.

For example, if that portcullis falls back down again and the party has to get through it a second time, the fact that you introduced a particular special handhold to explain away the weak character lifting it when the strong character failed means that handhold should still be there, meaning that the strong character should be able to trivially lift it this time. Or, the fact that the strong character failed to lift something that is similar to things they lifted in the past might mean the GM introduces extra factors like 'this gate is rusty' which then make it harder for other characters to succeed. So from a metagame perspective, this means that it matters who tries first. It's safer to have the characters with a lower chance of success try first because if they succeed it will introduce factors to make the task easier for everyone, whereas if they fail it doesn't suggest the need for explanatory factors that would increase the difficulty for others.

But that is not the case in procedural, the roll to determine the state of the object is character agnostic, if the porticullis is determined to be rusty or not has nothing to do with which character prompted the interaction.


Note this can also be an issue even without multiple characters, but when rolls are used to determine things where the variance of the roll is mismatched to the narrative variance in ability the character would be expected to demonstrate. Even if we just have the strong character trying to lift an object - if in some cases that is just look up their lift weight and see whether the object is heavier or lighter, but in other cases its seen as an obstacle which must be rolled past, then in the latter case that added variance can end up being injected into the environment. Meaning that again there may be a metagame consideration about whether or not to back down upon being told there is a roll involved.

How can they "back down"? The roll is representing the character making the attempt, once they say "I pry the porticullis open", then the skill check roll is called.


On the other hand, if retcon/procedural generation is not used reactively to justify a dice result, but is instead used when something about the direction of play necessitates those details being created, then the effects of metagame causality are much more controlled, because that procedural generation can prioritize self-consistency with other established details rather than being pinned to a particular end that now needs to be explained. You can still get in trouble if for example you're filling in those details in a way that is informed by the current state of the PCs - entering a zone and making the GM detail it at low levels, then returning at higher levels to force it to not rubber-band to the party's power level for example. But as long as the procedural generation isn't being based off of the PCs, that sort of meta-causality will be pretty weak.

That's a feature in my eyes, scaling things numbers to "level up" with the party defeats the purpose of leveling in the first place. And going back to a dungeon you couldn't clear a couple levels ago and being able to clear it now, is 100% working as intended.

[Psyren]

For me pro-cedural comes from pro-active, but w/e I never checked the definition.

Procedural comes from procedure, which simply means you have a method or process you use to generate the world. That method can be proactive or retroactive.


But you don't do that in procedural generation either, I won't roll to decide the state of the manacles, until the state of the manacles comes into question. Once they are gonna be interacted with, I roll to determine their state, in an old unkempt dungeon the variance will be higher than in a small town's barrack's jail, and in a major city's dungeon, I may not roll at all and just decide that they are in pristine condition. Once the state of the object to interact with has been decided, now the object is interacted with. Yes, it is slower than retcon generation, but its also fairer, since if the manacles were in poor condition, or coated in acid, the creature interacting with it, could maybe notice such things before attempting whatever it is they were gonna attempt.

My point is that rolling to determine the manacles condition when placing them in the world, using the player's roll to break them determine their state, and simply deciding what stat they are without rolling for them at all, are all valid ways to play 5e. I happen to prefer #3, raven777 appears to prefer #2, neither of us is wrong.

(What would be wrong, is picking #2, and then complaining that the wizard broke them while the barbarian failed hurts your immersion. You chose that possibility.)


Exactly, which is why it feels like cheating.

Like I said, I was playing devil's advocate there. I'm not going to go out of my way to defend a method I dislike.

What I will defend is 5e's ability to encompass all three approaches.

[NichG]

But that is not the case in procedural, the roll to determine the state of the object is character agnostic, if the porticullis is determined to be rusty or not has nothing to do with which character prompted the interaction.

In the portcullis example, the state of the object was determined in order to explain the success or failure of the characters who tried - why the strong character failed to lift it but the weak character succeeded.



How can they "back down"? The roll is representing the character making the attempt, once they say "I pry the porticullis open", then the skill check roll is called.


'How heavy does this gate look? Is it within my maximum lift capacity?' -> 'Well you don't know, you could try to lift it though' -> 'Nevermind, lets find a lever somewhere'


That's a feature in my eyes, scaling things numbers to "level up" with the party defeats the purpose of leveling in the first place.

This example would be like reverse rubber-banding, where the party intentionally sticks their head into a zone that seems like it should be higher level than they can deal with, the GM down-scales it, then the party leaves and comes back later. Like rushing to get to the first Oblivion gate in Oblivion before Lv12 or so, so that the NPCs can still provide useful support against the daedra that spawn rather than just being totally slaughtered.

[Rukelnikov]

My point is that rolling to determine the manacles condition when placing them in the world, using the player's roll to break them determine their state, and simply deciding what stat they are without rolling for them at all, are all valid ways to play 5e. I happen to prefer #3, raven777 appears to prefer #2, neither of us is wrong.

(What would be wrong, is picking #2, and then complaining that the wizard broke them while the barbarian failed hurts your immersion. You chose that possibility.)

Like I said, I was playing devil's advocate there. I'm not going to go out of my way to defend a method I dislike.

What I will defend is 5e's ability to encompass all three approaches.

Playing without dice is also valid, my point is that retcon generation robs the players of the ability to make educated guesses like we are use to doing in the real world, if I see that water is leaking out of a bottle I can infer that if I pick it up it will wet the floor. If the Dm decides the bottle is punctured after the PC lifted it, the PC never got a chance to notice it was punctured in the first place because when they picked it up it wasn't punctured yet.


In the portcullis example, the state of the object was determined in order to explain the success or failure of the characters who tried - why the strong character failed to lift it but the weak character succeeded.

Yeah, which is why I call it retcon generation, becuase its generating the state of the world retroactively, the porticullis wan't rusty when the barbarian attempted to lift it, it only became rusty when the barbarian failed. In procedural generation, the porticullis is either rusty or not independant of the barbarian's roll.


'How heavy does this gate look? Is it within my maximum lift capacity?' -> 'Well you don't know, you could try to lift it though' -> 'Nevermind, lets find a lever somewhere'

Perfect, they took the time to gauge if its within their capabilities (which is when the state of the object would have to be decided), and that's ok, I can look at something and get an idea if I'm gonna be able to lift it or not, or I'm unsure.


This example would be like reverse rubber-banding, where the party intentionally sticks their head into a zone that seems like it should be higher level than they can deal with, the GM down-scales it, then the party leaves and comes back later. Like rushing to get to the first Oblivion gate in Oblivion before Lv12 or so, so that the NPCs can still provide useful support against the daedra that spawn rather than just being totally slaughtered.

Eh... I don't know, I don't usually scale things to players level, iron is iron whether the PCs are lvl 2, 5 or 18.

Expanding upon this though, I do create some stuff specifically outisde of procedurally generating things, and those I do take the level of the PCs into account, but those I have already created beforehand, so the players interactions will not prompt any major changes, if its an encounter designed towards lvl 4 PCs (which is not necessarilly the levels my players are), then whether its iron bars or adamantine bars, can be easily inferred from the rest of the dungeon. If the doors are made of Iron, and there's no reason why the porticullis would be specifically harder than any other door, then it will also be iron. If the doors are adamantine, then the porticullis will be too.

[Segev]

To me, either you're going to determine that the barbarian can break them and the wizard cannot, because the outcomes are just that obvious, or you're going to determine that it's possible the barbarian could break them, and set a DC. If you set a DC, you're making a statement about the difficulty of breaking the manacles. If the wizard can hit the DC, then he should succeed if he does. If you feel the manacles should not be breakable by somebody with a strength of 10 and no proficiency in athletics, then you should set the DC to be at least 21. You're not doing this because "the wizard shouldn't be able to break them," but because people of particular strength and skill shouldn't. The wizard, if he finds a way to best that anyway - perhaps he has bonuses you don't recall - should be allowed to try at the same DC. If you're setting the permissions based on talent and skill, set the DCs rather than the permissions based on that. Otherwise, just give it to the one that you feel should succeed.

But I would posit that you're setting the DCs "wrong" if you've set them such that an unrealistic outcome is possible.

[Psyren]

Playing without dice is also valid

I'm not talking about resolving challenges though, I'm strictly talking about building the world.

As mentioned, determining which challenges are possible without using dice to do so is exactly what I do.


To me, either you're going to determine that the barbarian can break them and the wizard cannot, because the outcomes are just that obvious, or you're going to determine that it's possible the barbarian could break them, and set a DC. If you set a DC, you're making a statement about the difficulty of breaking the manacles. If the wizard can hit the DC, then he should succeed if he does. If you feel the manacles should not be breakable by somebody with a strength of 10 and no proficiency in athletics, then you should set the DC to be at least 21. You're not doing this because "the wizard shouldn't be able to break them," but because people of particular strength and skill shouldn't. The wizard, if he finds a way to best that anyway - perhaps he has bonuses you don't recall - should be allowed to try at the same DC. If you're setting the permissions based on talent and skill, set the DCs rather than the permissions based on that. Otherwise, just give it to the one that you feel should succeed.

But I would posit that you're setting the DCs "wrong" if you've set them such that an unrealistic outcome is possible.

I don't think every bonus is created equal. If the wizard can do it because they shapeshifted into a Storm Giant and became much stronger, fine. If instead they got +10 to their check via lucky rolls from guidance + bardic song, that could likely get them out of the manacles (Acrobatics) but not by physically sundering them.

[Rukelnikov]

I'm not talking about resolving challenges though, I'm strictly talking about building the world.

As mentioned, determining which challenges are possible without using dice to do so is exactly what I do.

I'm talking about everything, you can perfectly play without dice at all.

[RSP]

Because I had limited time to craft the perfect example. Would you like to help me find a better one? I'm sure we can think of something.
…My point is that we're telling a story. The dices are there to add some variation to the way similar situations can resolve. If it was a video game, I'd say they add replay value.

I’m a little confused. Is the first part what you’d tell your players?

It seems to me you’ve created a situation in-game where it doesn’t matter who rolls, or what they roll, so long as the die says a certain number.

Why didn’t the Wiz roll Perception? If it wasn’t a perception problem before they rolled, how did it become a perception problem after they made a Str roll?

Did the Barb have a chance to use Perception (passive or with a roll) on the portcullis? Did the Rogue?

Why is noticing something a Strength check?

[Psyren]

I'm talking about everything, you can perfectly play without dice at all.

I know that, I'm saying that's a non sequitur when we're talking about worldbuilding. Yes, you can do combat exploration etc. without dice, but very few people actually do. Establishing the world without dice is much more common than playing the game without dice.

[Rukelnikov]

I know that, I'm saying that's a non sequitur when we're talking about worldbuilding. Yes, you can do combat exploration etc. without dice, but very few people actually do. Establishing the world without dice is much more common than playing the game without dice.

Sure, but dice make it more impartial IMO, metagame is not just a PCs thing, and using dice to decide stuff I hadn't decided prior to PCs prompt ensures that I'm not just deciding how things play. Its a matter of choice.

[Psyren]

Sure, but dice make it more impartial IMO, metagame is not just a PCs thing, and using dice to decide stuff I hadn't decided prior to PCs prompt ensures that I'm not just deciding how things play. Its a matter of choice.

Right, that's one of the three approaches (https://forums.giantitp.com/showthread.php?650488-Skill-Checks-shouldn-t-use-a-d20&p=25609215&viewfull=1#post25609215) I mentioned (#1, to be specific.) To reiterate, none of them is badwrongfun.

[Rukelnikov]

Right, that's one of the three approaches (https://forums.giantitp.com/showthread.php?650488-Skill-Checks-shouldn-t-use-a-d20&p=25609215&viewfull=1#post25609215) I mentioned (#1, to be specific.) To reiterate, none of them is badwrongfun.

And what I meant with you can play without dice if you want to was that nothing is ever badwrongfun. However, retcon generation is robbing the players of agency, by them not having the ability to perceive the world until they have already interacted with it, as with the example of the leaking bottle.

[Psyren]

And what I meant with you can play without dice if you want to was that nothing is ever badwrongfun. However, retcon generation is robbing the players of agency, by them not having the ability to perceive the world until they have already interacted with it, as with the example of the leaking bottle.

Yet again, I don't like that method either.

I get the feeling we're in violent agreement so I'll bow out.

[MadBear]

To me, either you're going to determine that the barbarian can break them and the wizard cannot, because the outcomes are just that obvious, or you're going to determine that it's possible the barbarian could break them, and set a DC. If you set a DC, you're making a statement about the difficulty of breaking the manacles. If the wizard can hit the DC, then he should succeed if he does. If you feel the manacles should not be breakable by somebody with a strength of 10 and no proficiency in athletics, then you should set the DC to be at least 21. You're not doing this because "the wizard shouldn't be able to break them," but because people of particular strength and skill shouldn't. The wizard, if he finds a way to best that anyway - perhaps he has bonuses you don't recall - should be allowed to try at the same DC. If you're setting the permissions based on talent and skill, set the DCs rather than the permissions based on that. Otherwise, just give it to the one that you feel should succeed.

But I would posit that you're setting the DCs "wrong" if you've set them such that an unrealistic outcome is possible.

The issue I'd point out with your example, runs counter to how you set DC's in the game though.

Let's say we put the Breaking Manacles DC at 21. That way a strength 10 person with no skill finds it impossible to break the bonds. A level 5 barbarian with an 18 strength and athletics profciency will have a +7 bonus to their roll. This means they'd need to roll a 14 or higher to succeed. So now I have a DC that the wizard can't overcome, but my extremely strong barbarian can.... 35% of the time that they roll.

I totally get that this is why people will set different DC's for different characters. For verisimilitude, you'd think that the extremely strong player would have a higher chance then 1/3 of breaking bonds that the weak character couldn't even do at their best.

Again, I have no concrete solutions to this problem, only a suspicion that there are probably better ideas out there (I've seen a few in this thread I'll be trying with my players next Thursday).

[RSP]

Let's say we put the Breaking Manacles DC at 21. That way a strength 10 person with no skill finds it impossible to break the bonds. A level 5 barbarian with an 18 strength and athletics profciency will have a +7 bonus to their roll.

Obviously play it as you want, but I wanted to point out a common misapplication of the skill: Athletics doesn’t apply to checks to break stuff, RAW. Breaking manacles, or breaking down a door would just be a Str check.

[Psyren]

Let's say we put the Breaking Manacles DC at 21. That way a strength 10 person with no skill finds it impossible to break the bonds.

But all they need is the most beginner/least talented bard or cleric in the next room to make it... right?

[NichG]

But all they need is the most beginner/least talented bard or cleric in the next room to make it... right?

Well, this basically comes down to what exactly it is you envision the bard or cleric's magic doing. If its making them feel a bit happier, ridiculous. Inducing a state of hysterical strength, the kind that lets someone lift a car to save their child? Maybe more plausible. Invoking the legend of a mythic strongman and allowing that legend to overlay its existence with the weak character for just a moment, permitting them to borrow strength not their own? Well, why not, if Hercules could break the manacles then magic summoning the vestige of Hercules may as well do so as well.

A lot of stuff comes down to whether you feel like being impressed enough to go through the trouble to imagine how it might be in a way that would convince you 'yeah okay this could work'...

[Psyren]

The whole "lift a car to save your child" thing is generally less than 5 feet, which D&D doesn't really measure. You can free one with a couple of inches if clearance even. It's not quite the mythical feat of strength an ability score difference would imply.

[animorte]

The whole "lift a car to save your child" thing is generally less than 5 feet, which D&D doesn't really measure. You can free one with a couple of inches if clearance even. It's not quite the mythical feat of strength an ability score difference would imply.

You mean, with proper leverage? (https://www.youtube.com/watch?v=0Hn6FHh7NOA)

[Segev]

The issue I'd point out with your example, runs counter to how you set DC's in the game though.

Let's say we put the Breaking Manacles DC at 21. That way a strength 10 person with no skill finds it impossible to break the bonds. A level 5 barbarian with an 18 strength and athletics profciency will have a +7 bonus to their roll. This means they'd need to roll a 14 or higher to succeed. So now I have a DC that the wizard can't overcome, but my extremely strong barbarian can.... 35% of the time that they roll.

I totally get that this is why people will set different DC's for different characters. For verisimilitude, you'd think that the extremely strong player would have a higher chance then 1/3 of breaking bonds that the weak character couldn't even do at their best.

Again, I have no concrete solutions to this problem, only a suspicion that there are probably better ideas out there (I've seen a few in this thread I'll be trying with my players next Thursday).

If you want the barbarian to make the check reliably, do you really want him to fail? Or do you want, narratively, the barbarian to succeed, and you're looking to make the numbers work so that he has such a good chance to that his failure would be a surprise? Do you allow retries, perhaps with consequences?

Either just let the Barbarian succeed without a roll because you are uninterested in his failure, or set the DC where you realistically believe it should be based on who you think has even a slight chance of making it without specialized aid.

You could also set a strength score threshold: if the combined strength score of those trying to break the chain exceeds a certain number, it breaks. Set that at 18 and the barbarian breaks it by himself, and three weedy wizards can also break it working together.

But if you are calling for rolls, you should be expecting the possibility that the barbarian could fail, and that the wizard might find a way to actually succeed. Maybe with magic helping, maybe with aid from another, but if you already have the mental image that the barbarian succeeds and the wizard fails, simply leave the dice out of it and give the barbarian the success and tell the wizard "no."

I know a GURPS GM who consistently talks about how he wants victories to be impressive because they are so rare on the dice, but is consistently surprised when they never happen in his games and players either start "munchkinning" to get high numbers or lose interest because they fail so much. He genuinely doesn't seem to get that the mathematically low probability of success that makes the victory so amazing in a novel or a movie is a lie: the probability was always 100%, because the author was always going to choose that outcome.

I feel like people who want to rig the probabilities so the chosen PC succeeds are making the same kind of mistake, albeit not to the same degree: they want a particular character to succeed, but want to believe the dice made a difference. And are disappointed and say the dice are the problem when the dice don't cooperate. Well, they are right: the dice were the problem. Or, rather, the choice to use them at all was the problem. 5e stipulates at you should not call for a roll when you don't consider both outcomes narratively interesting. If you are disappointed to the point that you feel the dice were a problem for your verisimilitude, you should not have allowed randomness into the result, and simply declared the result you wanted.

[Telok]

Again, I have no concrete solutions to this problem, only a suspicion that there are probably better ideas out there (I've seen a few in this thread I'll be trying with my players next Thursday).

One thing I proposed a couple years ago was to just use the monsters. Reskin them as a "challenge", figure out what to replace hp & damage & ac with. Suddenly that combat system that works is making your non-combat stuff work just as well. The only things you solve with single ability rolls is stuff you'd solve with a single attack roll, everything else is a significant challenge or auto-success/auto-fail.

Nobody like it for some reason.

[Raven777]

I’m a little confused. Is the first part what you’d tell your players?

It seems to me you’ve created a situation in-game where it doesn’t matter who rolls, or what they roll, so long as the die says a certain number.

Why didn’t the Wiz roll Perception? If it wasn’t a perception problem before they rolled, how did it become a perception problem after they made a Str roll?

Did the Barb have a chance to use Perception (passive or with a roll) on the portcullis? Did the Rogue?

Why is noticing something a Strength check?

Barbarian and Wizard are on the run from an angry guild of thieves in the sewers. Time is of the essence. They come across an iron portculis that bars their escape. They hear the thieves coming from around the corner. If the two adventurers would rather escape than make a stand, DM informs the players that the portculis can be lifted with a strength check. Barbarian tries first, rolls low, and fails. Wizard tries second, rolls high, and succeeds. They make their escape. Back at the inn, out of character, Barbarian's player complains that it was not realistic for Wizard to lift the portculis while Barbarian couldn't just because Wizard's player got lucky with the dice. At this point, DM tells the players that Wizard got lucky and happened to lift at an angle that overcame some protrusion in the frame, while Barbarian did not. What would you tell them?

I am beginning to believe that the best answer from DM, rather than justifying the result a posteriori, would have been to just say "Yes... Peculiar, isn't it?", grin mysteriously, not elaborate, and move on with the game.

[animorte]

What do you tell them?

In haste, your hands slipped off the rail, pulling a great deal of the built up grime with you. The Wizard gripped that same spot, now mostly clear and dry of the nasty. Must have gotten a much better grip.

Just to be clear, I’m not terribly fond of needing to make up for silly instances like that. I’m also very open though, to imagine the odd chance. I do prefer a slightly more consistent representation of legitimate skill proficiency, or training.

In situations, it might also be worth mentioning the many times you have succeeded with the thing you’re good at, especially under duress.

[Psyren]

Barbarian and Wizard are on the run from an angry guild of thieves in the sewers. Time is of the essence. They come across an iron portculis that bars their escape. They hear the thieves coming from around the corner. If the two adventurers would rather escape than make a stand, DM informs the players that the portculis can be lifted with a strength check. Barbarian tries first, rolls low, and fails. Wizard tries second, rolls high, and succeeds. They make their escape. Back at the inn, out of character, Barbarian's player complains that it was not realistic for Wizard to lift the portculis while Barbarian couldn't just because Wizard's player got lucky with the dice. At this point, DM tells the players that Wizard got lucky and happened to lift at an angle that overcame some protrusion in the frame, while Barbarian did not. What would you tell them?

Not aimed at me, but my approach:

1) If the portcullis is one the wizard could not have lifted, there would be no opportunity for them to "get lucky with the dice."
2) If the portcullis is one the wizard could have lifted, then that means it would likely be an easier challenge for the barbarian. Either it would be automatic success for them, or else a failure from the barbarian would likely mean "progress + setback."


I am beginning to believe that the best answer from DM, rather than justifying the result a posteriori, would have been to just say "Yes... Peculiar, isn't it?", grin mysteriously, not elaborate, and move on with the game.

While this would certainly work, I tend to try and maintain a very tight budget of these.

[RSP]

Barbarian and Wizard are on the run from an angry guild of thieves in the sewers. Time is of the essence. They come across an iron portculis that bars their escape. They hear the thieves coming from around the corner. If the two adventurers would rather escape than make a stand, DM informs the players that the portculis can be lifted with a strength check. Barbarian tries first, rolls low, and fails. Wizard tries second, rolls high, and succeeds. They make their escape. Back at the inn, out of character, Barbarian's player complains that it was not realistic for Wizard to lift the portculis while Barbarian couldn't just because Wizard's player got lucky with the dice. At this point, DM tells the players that Wizard got lucky and happened to lift at an angle that overcame some protrusion in the frame, while Barbarian did not. What would you tell them?

I am beginning to believe that the best answer from DM, rather than justifying the result a posteriori, would have been to just say "Yes... Peculiar, isn't it?", grin mysteriously, not elaborate, and move on with the game.

I tell them the 550 lbs. portcullis is too heavy for the character with a max lift of 240 lbs. to lift it. But the 600 lbs max lift Barb can lift it.

Why do you not take the information available into account?

[GloatingSwine]

I tell them the 550 lbs. portcullis is too heavy for the character with a max lift of 240 lbs. to lift it. But the 600 lbs max lift Barb can lift it.

Why do you not take the information available into account?

If the Barbarian will consistently lift the portcullis, why is it there?

[RSP]

If the Barbarian will consistently lift the portcullis, why is it there?

What are the strategic uses of a 550 lbs portcullis? Are the goblins chasing them able to lift it? If the Barb lifts it during combat, that’s their action, and they have to hold it for the group to get through. Did the Barb get knocked out during combat meaning they’re not capable of lifting it at the moment?

Combat, and the game in general I’d imagine, becomes rather boring if the world is devoid of features other than DC X.

[GloatingSwine]

What are the strategic uses of a 550 lbs portcullis? Are the goblins chasing them able to lift it? If the Barb lifts it during combat, that’s their action, and they have to hold it for the group to get through. Did the Barb get knocked out during combat meaning they’re not capable of lifting it at the moment?

Combat, and the game in general I’d imagine, becomes rather boring if the world is devoid of features other than DC X.

The point is that in that scene as described (tension provided by a chase) the portcullis is supposed to be an obstacle with uncertainty attached to produce tension, "can you successfully do this or will you have to fight". If you present it such that the barbarian always succeeds then the tension is gone and actually the thing that was supposed to be an obstacle is the win condition now.

The world also becomes boring if it is devoid of uncertainty, and that's what the dice are there to provide. The job of the DM is to provide situations where the players can't automatically pass, or where an optional route that uses an automatic pass has some other downside that they can avoid by succeeding at a check.

(And that leads back to the thing the thread is talking about originally, single roll D20 checks are maybe not the best way to do that, especially not in 5e bounded accuracy).

[Segev]

If the Barbarian will consistently lift the portcullis, why is it there?

Are only things that impede heroes extant in your world?

[GloatingSwine]

Are only things that impede heroes extant in your world?

No, but things that are intended to impede heroes, as the portcullis in the posited example was should actually have a chance to do so.

[RSP]

The point is that in that scene as described (tension provided by a chase) the portcullis is supposed to be an obstacle with uncertainty attached to produce tension, "can you successfully do this or will you have to fight". If you present it such that the barbarian always succeeds then the tension is gone and actually the thing that was supposed to be an obstacle is the win condition now.

The world also becomes boring if it is devoid of uncertainty, and that's what the dice are there to provide. The job of the DM is to provide situations where the players can't automatically pass, or where an optional route that uses an automatic pass has some other downside that they can avoid by succeeding at a check.

(And that leads back to the thing the thread is talking about originally, single roll D20 checks are maybe not the best way to do that, especially not in 5e bounded accuracy).


No, but things that are intended to impede heroes, as the portcullis in the posited example was should actually have a chance to do so.

This seems very railroady, to an extent I don’t usually do.

The portcullis is a portcullis, regardless of the situation. I don’t change facts I’ve determined about the portcullis based on what the Players do: it’s there and it’s properties are the same regardless of whether it’s a benefit to the Players or a hindrance.

If the Players get through and prop it open, that may allow certain creatures to get through after them. Or if they close it behind them, it may be a hindrance if they get forced back that way. They could also use it to their advantage in a combat.

The idea that the Wizard gets to functionally increase their Str from 8 to 20 by rolling a high number on a d20 doesn’t fly at my table; anymore than the Barb gets to cast spells by rolling high on a d20.

[Sorinth]

No, but things that are intended to impede heroes, as the portcullis in the posited example was should actually have a chance to do so.

This is portcullis example actually shows why the skill checks shouldn't always have binary success/fail outcomes. If the Barbarians failed strength check isn't they couldn't lift it but instead something like they lift it a foot or two off the ground then the party can squeeze through but they'll have to do something like brace the portcullis to allow the barbarian to also get through. Maybe the cleric uses their shield to brace it which works but the shield is damaged/not retrievable so the next set of combats is going to be a little tougher for the cleric. So the portcullis has still created an impediment/challenge but the in-world logic of this being something the barbarian should reliably be able to do is maintained.

[Jervis]

With the conversation around skill checks and what is a reasonable DC to set, it brought me back to a fundamental question about how we go about these skill checks. Why do we let luck have a bigger deciding factor on success then the amount of training or skill involved?

With the reduction in numbers that make it so no one is rocking a +45 to a skill check (characters in my old 3.x days) this has made it so that the roll of the dice is can very often be a bigger predictor of a result then anything else. It leads to akward scenarios where the wizard doesn't know something about the arcane that the barbarian whose somehow does know. Or that the barbarian is not able to lift a portcullis, that the wizard was somehow able to.

Often the "fixes" to these issues involve not letting players roll if they're not trained in a skill, or setting the bar high enough that only those trained can succeed, or having different DC's for different characters. (I'm not saying these fixes are good, only that this is what many DM's often end up doing).

This bring back the question, of how do we represent chance without it overshadowing training, and natural skill (proficiency + Ability score).

My gut is to help reduce the standard deviation via using 2d10 or 5d4's. In this way, characters are far more likely to roll the average then they are to roll extremely well or extremely low.

What are your thoughts? Should we continue using the d20 and move on, or is there a better way?

I mean just using 3d6 would make sense. Same average and low deviation

[GloatingSwine]

This is portcullis example actually shows why the skill checks shouldn't always have binary success/fail outcomes. If the Barbarians failed strength check isn't they couldn't lift it but instead something like they lift it a foot or two off the ground then the party can squeeze through but they'll have to do something like brace the portcullis to allow the barbarian to also get through. Maybe the cleric uses their shield to brace it which works but the shield is damaged/not retrievable so the next set of combats is going to be a little tougher for the cleric. So the portcullis has still created an impediment/challenge but the in-world logic of this being something the barbarian should reliably be able to do is maintained.

In the example given both success and failure were obviously acceptable outcomes though. Lift the portcullis and don't have a fight, or fail to lift it and fight the pursuers.

The players don't have to get through the portcullis, they can fight the goons they just decided to try and avoid it.


The portcullis is a portcullis, regardless of the situation. I don’t change facts I’ve determined about the portcullis based on what the Players do: it’s there and it’s properties are the same regardless of whether it’s a benefit to the Players or a hindrance.

Right, and if there's ever going to be a question about whether the players should be able to get through it under pressure or not you should define those properties such that an ability check takes place when they try, so when you were writing the scenario and you knew the barbarian could lift 600lb you made the portcullis 700, and that if they fail something else interesting happens like they have to keep running, or they have to have a fight, or they have to find another way into the place on the other side.

But once you've called for an ability check and decided on a DC you've made that call and that's the DC no matter who is rolling.

And again, the "wrong person succeeds" problem is the point of this thread which is calling for a better check resolution system than a single D20 roll.

[LudicSavant]

is there a better way?

I'd say yes. I feel lots of other systems handle skills better than D&D does.

[GloatingSwine]

NB: also in the "strong barbarian tries to lift a thing, fails, then weak wizard lifts same thing and succeeds" scenario, that shouldn't occur.

If there's time to try more than once the barbarian should try more than once, their arms didn't fall off when they failed, and if there's room for both to try at the same time the barbarian should do it with advantage because that's how shared checks work.

[Sorinth]

But once you've called for an ability check and decided on a DC you've made that call and that's the DC no matter who is rolling.

Nowhere in the rules does it say the above and I'd point out the tasks aren't even the same. One task is trying lift something that is 16% above your normal limit (700lbs when max is 600lbs) and the other is trying to lift something 192% above your normal limit (700lbs when max is 240lbs). It's perfectly reasonable to say that lifting an extra 16% is a Moderate Difficulty but lifting an extra 192% is Nearly Impossible if not outright impossible.

[GloatingSwine]

Nowhere in the rules does it say the above and I'd point out the tasks aren't even the same. One task is trying lift something that is 16% above your normal limit (700lbs when max is 600lbs) and the other is trying to lift something 192% above your normal limit (700lbs when max is 240lbs). It's perfectly reasonable to say that lifting an extra 16% is a Moderate Difficulty but lifting an extra 192% is Nearly Impossible if not outright impossible.

That's true, but as noted unless they're trying to lift two different things at the same time (and you wouldn't design that challenge for one strong and one weak character unless the weak character has a resource to assist them making their task different anyway), the situation should never arise.

Either the strongest character keeps trying until they succeed (if something else interesting is happening whilst they try), or they try together and you use the highest strength with advantage, or there's only time for one attempt in the first place.

[Telok]

This is portcullis example actually shows why the skill checks shouldn't always have binary success/fail outcomes. If the Barbarians failed strength check isn't they couldn't lift it but instead something like they lift it a foot or two off the ground then the party can squeeze through but they'll have to do something like brace the portcullis to allow the barbarian to also get through. Maybe the cleric uses their shield to brace it which works but the shield is damaged/not retrievable so the next set of combats is going to be a little tougher for the cleric. So the portcullis has still created an impediment/challenge but the in-world logic of this being something the barbarian should reliably be able to do is maintained.

Open heavy thing in my games is... translate to 5e... something like 50 points over dc 15 to completely open. Then since I know how big the opening is I can map the amount of success to how far its open. This benefits the people with good scores by giving them better successes and you can further limit the ability of low bonus characters while only slightly inconveniencing the high bonus characters by making it more like; dc 12, needs 50 points over, loses 1d4 points at the end of the turn.

Really to only things you ask of the dice are "can they do it at all", "how well can they do it", and "how fast can they do it". Lots of the time people make the mistake of rolling success/failure when they really want the how well/how fast. In addition putting success/failure on a multi-roll framework screws with your probabilities in ways that people keep getting wrong (example: first & second versions of 4e skill challenge math). But D&D for a long time has put all d20 rolls abd the various multi-roll frameworks as steaight success/failure tests, which conditions the users to think only in terms of success/failure.

[Raven777]

Nowhere in the rules does it say the above and I'd point out the tasks aren't even the same. One task is trying lift something that is 16% above your normal limit (700lbs when max is 600lbs) and the other is trying to lift something 192% above your normal limit (700lbs when max is 240lbs). It's perfectly reasonable to say that lifting an extra 16% is a Moderate Difficulty but lifting an extra 192% is Nearly Impossible if not outright impossible.

Wizard's lifting capacity being lower is already taken into acount in the roll: their strength bonus will be lower, maybe even negative. Tweaking the DC per character based on their stats would effectively be double dipping on the effects of ability scores.

[Sorinth]

Wizard's lifting capacity being lower is already taken into acount in the roll: their strength bonus will be lower, maybe even negative. Tweaking the DC per character based on their stats would effectively be double dipping on the effects of ability scores.

And why is that a bad thing if it leads to more logical results? Keep in mind it's 100% RAW legal for the DM to set a different DC, it's possible they've even codified it by doing something like going up to 10% over it's DC 10, going up to 20% over, DC 15, going 30% over DC 20, etc...

[Segev]

No, but things that are intended to impede heroes, as the portcullis in the posited example was should actually have a chance to do so.

And they do; if the barbarian is the only one who can lift it, the party doesn't advance until the barbarian is there to lift it.

The rogue scouting ahead can't lift it, for example.

And that assumes your goal is strictly gamist, and not merely verisimilitude that castles have portcullises because they want to keep people out.

[Amechra]

I'd say yes. I feel lots of other systems handle skills better than D&D does.

This, really — though this probably has a lot to do with a lot of other systems treating skills as your core measure of competency instead of D&D's wobbly tripod of attack rolls/saving throws/skill checks.

Honestly, a lot of groups could probably get better results than the current skill system by dumping it in the trash can and free-forming stuff like "can you lift the grating?"

[MadBear]

Obviously play it as you want, but I wanted to point out a common misapplication of the skill: Athletics doesn’t apply to checks to break stuff, RAW. Breaking manacles, or breaking down a door would just be a Str check.

You're right, I forgot about that. That makes the situation worse. now with only a +4, the barbarian is only succeeding at things the wizard can't 20% of the time that rely on raw strength.

and while yes you can say "just let the barbarian succeed" or "don't let the wizard roll", that is kinda missing the point of this thread. Yes, one solution is to just not use dice at all in these scenarios. This is something I already do. But it answers the issue while missing the main point that the d20 has a bigger impact on the result then any amount of training/raw skill.

I'm definitely aware of other systems with better utilization of skill systems, I'm just curious what other ideas exist out there that mesh well with D&D.

[animorte]

I'm definitely aware of other systems with better utilization of skill systems, I'm just curious what other ideas exist out there that mesh well with D&D.

Ultimately under this thought process, we could redirect the thread (or start a new one, which would probably fit better on homebrew) to hunt down through each of our varied experiences to discover what are the better skill systems we could plug in for 5e.

[GloatingSwine]

And they do; if the barbarian is the only one who can lift it, the party doesn't advance until the barbarian is there to lift it.

The rogue scouting ahead can't lift it, for example.

And that assumes your goal is strictly gamist, and not merely verisimilitude that castles have portcullises because they want to keep people out.

Again, read the posited example. There was a barbarian and a wizard, they were running away, the portcullis may allow them to avoid a fight if they can lift it.

So you need to make sure the "if" is preserved. That needs to be a point of uncertainty, the game is built on points of uncertainty because they generate tension and release when that tension is resolved.

And you do that by making sure that this portcullis is outside of the guaranteed ability of either character to lift, so they have to make a check to do it.

[Psyren]

If the Barbarian will consistently lift the portcullis, why is it there?

What if the party has no barbarian? Do all your traps and locks cease to exist just because the rogue is good at dealing with them?

Yes, you should make sure that the party you're DMing for is challenged, but obstacles that a specialist can bypass easily are a good thing too, it lets them feel their specialization is valuable. A mix of both is a good thing.

[RSP]

But once you've called for an ability check and decided on a DC you've made that call and that's the DC no matter who is rolling.

And again, the "wrong person succeeds" problem is the point of this thread which is calling for a better check resolution system than a single D20 roll.

No. If the portcullis is 700 lbs then a Str 20 character would need to extend their max lift by 1/6th, or 16%.

The Str 8 character needs to increase their max lift by just under 300%.

The first may be a DC 15 Str check. The latter is impossible (in my games it’s impossible to do stuff like effectively double you stats, much less triple them).

You’re saying I need to see those tasks as the same, but they’re very much not.


Wizard's lifting capacity being lower is already taken into acount in the roll: their strength bonus will be lower, maybe even negative. Tweaking the DC per character based on their stats would effectively be double dipping on the effects of ability scores.

No: it’s using the info we have about the character to effectively determine what the challenge is. The stats mean something other than a modifier. Specific to this example, we know what the characters’ max lifts are. Ignoring that info is ignoring that they have those stats.

Moreover, saying DC 15 for either the Barb or Wiz isn’t accurately defining the task.

The Wizard shouldn’t have a 20% chance to more than double their effective Str, regardless of what the difficulty is for the Barb to increase their max lift by 16%.


You're right, I forgot about that. That makes the situation worse. now with only a +4, the barbarian is only succeeding at things the wizard can't 20% of the time that rely on raw strength.

and while yes you can say "just let the barbarian succeed" or "don't let the wizard roll", that is kinda missing the point of this thread. Yes, one solution is to just not use dice at all in these scenarios. This is something I already do. But it answers the issue while missing the main point that the d20 has a bigger impact on the result then any amount of training/raw skill.


The d20 works in 5% increments. If you don’t concede to a “everyone gets to roll, even though they aren’t equal tasks” then you can quite easily set a DC for a character.

Should the 20 Str Barbarian lift the portcullis ~50% of the time? DC 15. 25% of the time? DC 21.

Your question is flawed. There shouldn’t be a roll that allows the Str 8 character to succeed lifting 500 lbs. There shouldn’t be a roll that allows the Str 20 person to fail lifting 500 lbs. There’s no one DC that works, if you’re trying to force that both have to have a chance of success and of failure, while working against the same DC.

[Segev]

You're right, I forgot about that. That makes the situation worse. now with only a +4, the barbarian is only succeeding at things the wizard can't 20% of the time that rely on raw strength.

and while yes you can say "just let the barbarian succeed" or "don't let the wizard roll", that is kinda missing the point of this thread. Yes, one solution is to just not use dice at all in these scenarios. This is something I already do. But it answers the issue while missing the main point that the d20 has a bigger impact on the result then any amount of training/raw skill.

I'm definitely aware of other systems with better utilization of skill systems, I'm just curious what other ideas exist out there that mesh well with D&D.

The thing is, the d20 having that big an effect is the system working as intended.

The reasons not to call for rolls are the intended, designed-for solution to the problem of this straining verisimilitude.

You are free to disagree with the entire notion of bounded accuracy, just as I am free to disagree with 4th Edition's design philosophy of making every class a martial adept, but it isn't a problem as far as the system is concerned. Depending on how much of a deal-breaker bounded accuracy is for you, and how mush work you're willing to put in, more or less just for your own games, it may mean 5e is not for you. Or it may mean you just have to live with it, and apply the system's provided solutions for the actual problem you have (the verisimilitude-breaking nature of the d20 roll being so impactful if you do not engage the care to only call for rolls when and where they will not result in that problem).

I don't think you're going to get the help in this subforum to make a new ability check/skill system that you wants just because this subforum is for discussing 5e and people will likely see the topic and keep pointing out how 5e expects these things to be handled. You may have more luck in the homebrew subforum.

As a bit of a guideline, I would find existing systems that do skills in a way you find better. Did 3e or 4e D&D handle them better, to you? How about TURPS?

What sorts of ranges of possible results aldo you want, and how often do you want to call for rolls rather than assuming success or failure without one?

[Psyren]

The thing is, the d20 having that big an effect is the system working as intended.

The reasons not to call for rolls are the intended, designed-for solution to the problem of this straining verisimilitude.

You are free to disagree with the entire notion of bounded accuracy, just as I am free to disagree with 4th Edition's design philosophy of making every class a martial adept, but it isn't a problem as far as the system is concerned. Depending on how much of a deal-breaker bounded accuracy is for you, and how mush work you're willing to put in, more or less just for your own games, it may mean 5e is not for you. Or it may mean you just have to live with it, and apply the system's provided solutions for the actual problem you have (the verisimilitude-breaking nature of the d20 roll being so impactful if you do not engage the care to only call for rolls when and where they will not result in that problem).

I don't think you're going to get the help in this subforum to make a new ability check/skill system that you wants just because this subforum is for discussing 5e and people will likely see the topic and keep pointing out how 5e expects these things to be handled. You may have more luck in the homebrew subforum.

All of this. +100



As a bit of a guideline, I would find existing systems that do skills in a way you find better. Did 3e or 4e D&D handle them better, to you? How about TURPS?

What sorts of ranges of possible results aldo you want, and how often do you want to call for rolls rather than assuming success or failure without one?

There's actually something I liked about the 4e skill system that I wish 5e would have tried - level scaling for off-skills. In 4e, proficiency bonus was a flat number (I think +5?), and every skill or ability check went up passively by 1/2 your level - representing the odd collection of various knowhow that adventurers in a D&D setting would be likely to pick up just by adventuring long enough. Those numbers are far too high to work for bounded accuracy - but I could see something like 1/5 your level passively and a flat +3 for proficiency yielding similar-ish results to current 5e while also allowing a level 5 character and a level 15 character with the same ability scores to feel very different from one another at performing skills they're not formally trained in.

[Kane0]

There's actually something I liked about the 4e skill system that I wish 5e would have tried - level scaling for off-skills. In 4e, proficiency bonus was a flat number (I think +5?), and every skill or ability check went up passively by 1/2 your level - representing the odd collection of various knowhow that adventurers in a D&D setting would be likely to pick up just by adventuring long enough. Those numbers are far too high to work for bounded accuracy - but I could see something like 1/5 your level passively and a flat +3 for proficiency yielding similar-ish results to current 5e while also allowing a level 5 character and a level 15 character with the same ability scores to feel very different from one another at performing skills they're not formally trained in.

Consider my interest piqued

[Catullus64]

Some reasons to prefer the d20 over some of the lower-variance methods:

First off, the d20 is iconic to the game. The meaning of rolling a Natural 1 or a Natural 20 is perhaps one of the game's most deeply penetrating cultural memes. It feels like what D&D ought to feel like in our shared imagination, and therefore has real value. I believe some cows really are sacred.

Second off, the icosahedron, being the most complex of the platonic solid is a geometrically beautiful object. When they invent a series of d4s that can lock together into an icosahedron before breaking up into d4s when you roll it (maybe with magnets?) we'll talk.

And thirdly, moments that emerge from those extreme swings of chance are often memorable, at least when they are reacted to intelligently and with a gaming spirit; they often provoke or challenge roleplaying. I don't personally feel that my games would be enhanced by deliberately reducing the frequency with which those moments occur.

[MoiMagnus]

I mean just using 3d6 would make sense. Same average and low deviation

Small note for peoples who don't like to make additions in their head, "3d20 take the middle one" is very similar to "3d6 take the sum". The spread is slightly bigger (1-20 instead of 3-18), but is has the same average and the same shape mathematically speaking.

Or, if you want to stray even less from 5e's ruleset, there is a simple solution as a GM to avoid the variance of a d20:
Avoid calls for a skill check that are a single point of failure or a single point of success, and rely as much as possible on requiring multiple skill checks to make a task, with varying degree of success depending on how many of the individual tests succeeded.

For example for the Barbarian lifting a rock that block an entry, you can ask for 3 consecutive checks with two success required to work, which can be justified in various ways, like:
(? ? ?) First try of the Barbarian from what looks like the most practical position
(V ? ?) In case of success for lifting, the Barbarian needs to hold for long enough.
(V V ?) In case of success for lifting long enough, the rock is lifted quickly. No need for the third check.
(V X ?) In case of failure for lifting long enough, the Barbarian gets another try since the position is not too tiring
(V X V) The rock is lifted long enough, but it took more time than expected.
(X ? ?) In case of failure of lifting, the Barbarian can try another position that looks less practical but might just work.
(X V ?) This new position works for lifting the work, but is much more tiring than the first one so one get only one try to lift it long enough.
(X V V) The rock is lifted long enough, but it took more time than expected.
(X V X) The Barbarian just cannot lift it long enough. Maybe a single person can manage to sneak past if they're quick, but there is a risk they end up squished.
(X X ?) It's just a failure to lift the rock. No need for the third check.

[Psyren]

Consider my interest piqued

The math isn't too far off. Instead of proficiency bonus scaling from +2 to +6 before ability scores as it does now, it would scale from +3 to +7. And skills/weapons you're not proficient in (and flat ability checks) would scale from +0/+0 to +0/+4.

Expertise would get a little wonky though. Instead of expertise skills scaling +4/+12, they would scale +6/+10.

[OldTrees1]

Small note for peoples who don't like to make additions in their head, "3d20 take the middle one" is very similar to "3d6 take the sum". The spread is slightly bigger (1-20 instead of 3-18), but is has the same average and the same shape mathematically speaking.

Adding on, here is an anydice writeup (https://anydice.com/articles/interpolating-distributions/) about 3d20 take the middle.

[Keravath]

I hear what you're saying, but if you listen to posters on the board, you'll hear people repeatedly decry the unfairness of having different DC's for different PC's.

The DM decides the DCs. The DM knows the capabilities of the characters. If the DM wants a skill check that one character can pass and another can't then they can set a DC that one character could roll and the other can't. (I don't like the d20 auto success suggested in One D&D for that reason - there are some tests that a specific character should not be able to succeed at no matter what they roll. Too many DMs give the characters a roll when they really shouldn't and when that 20 comes up the DM is going to have to deal with what they consider success vs what the player was thinking of as success.)

In addition, the DM can assign advantage to the barbarian and disadvantage to the wizard. So, make lifting the 600lb portcullis a DC 20 athletics check which the wizard with -1 str and no athletics proficiency can't pass while the barbarian with +5 str and +3 proficiency and advantage stands a very good chance of passing. Perhaps make the DC easier for the barbarian if there are others helping to compensate for them already having advantage if the DM really wants them to succeed.

Anyway, there are ways for the DM to use even the same DCs for different characters to make it almost impossible for one character to succeed while fairly easy for another character to succeed.

[Psyren]

Adding on, here is an anydice writeup (https://anydice.com/articles/interpolating-distributions/) about 3d20 take the middle.

How does that work with advantage/disadvantage? The article doesn't appear to say.

[Segev]

There's actually something I liked about the 4e skill system that I wish 5e would have tried - level scaling for off-skills. In 4e, proficiency bonus was a flat number (I think +5?), and every skill or ability check went up passively by 1/2 your level - representing the odd collection of various knowhow that adventurers in a D&D setting would be likely to pick up just by adventuring long enough. Those numbers are far too high to work for bounded accuracy - but I could see something like 1/5 your level passively and a flat +3 for proficiency yielding similar-ish results to current 5e while also allowing a level 5 character and a level 15 character with the same ability scores to feel very different from one another at performing skills they're not formally trained in.

In theory, bounded accuracy is meant to enable you to have no need to scale up all of your values. I know you're not aiming to incorporate this particular aspect from 4e - at least, I think you're not - but one of the problems with 4e was that its scaling bonuses were largely illusory: you were expected to be on a treadmill wherein an "average difficulty" task DC was literally based on the level of the party, and thus you needed the same number on the d20 regardless of your level. Now, I'm hopeful that some of the design wasn't 100% faithful to that, so you could at least try "below level" or "above level" challenges and face the lower or higher DCs so that sometimes, your 1/2 level bonus made a difference. But 4e's design philosophy was very much to leave that treadmill running constantly.

5e's bounded accuracy is almost the same thing, just without bothering with the treadmill. Instead, you don't improve numbers...outside of a few things. Things you're proficient in, you get better at. You get some ASIs to bolster rolls. Some magic items help, too. And you can train to gain proficiencies (though skill training is a house rule, I think; it's one I heartily recommend, though, as narrative development of new skill sets can be very satisfying).

I think having proficiency training for skills is probably the best way to achieve what I think is your goal in the quoted section: representing PCs getting better at stuff over time and expanding into new areas. If they just flat-out all get something akin to Jack of All Trades, I think that will just encourage more treadmilling by DMs and will diminish the desired spread of competency. Heck, you, yourself, run the game such that you'll not even allow one player to make a check you'll let another make because you feel that it is impossible for the first player's PC but merely "somewhat difficult" for the other's, despite their numeric spread being such that they actually are close enough that the first player could possibly make it if you let him try at the same DC you let the other one try. This suggests to me that your DMing style would actually prefer larger gaps in competency between PCs in areas they're "good at" vs "not good at," and having even a small general level-based boost to nonproficient things would exacerbate the problem that the 8 strength wizard might have a shot at making a DC 18 strength check.

[PhantomSoul]

Small note for peoples who don't like to make additions in their head, "3d20 take the middle one" is very similar to "3d6 take the sum". The spread is slightly bigger (1-20 instead of 3-18), but is has the same average and the same shape mathematically speaking.


2@3d20 (roll 3d20 take middle) is a great curve and the average lineup vs. 3d6 and 1d20 is great, but "same shape" is stretching things a bit for typical usage (cf. https://anydice.com/program/2b923) unless being VERY broad about what "same shape mathematically speaking" means. The ranges are different (already a pretty big thing, particularly in the game context!) and the shapes aren't the same: 3d6 can't be modelled as part of a simple quadratic equation (it curves again on both edges, as shown through the "dips" in anydice), whereas 2@3d20 can be (it's y = -0.0752380952380952*(x^2)+1.57952380952381*x-0.791428571428572 for the density [frequency] plot), and 2@3d20 has much less of its probability space towards the center of the distribution compared to 3d6.


# Disclaimer: meant to be quick and easy to confirm formula visually, rather than to be elegant or pretty!

library(ggplot2)

df1 <- data.frame(expand.grid(
c(1:6),
c(1:6),
c(1:6)
))
df1$Model <- "3d6"
df1$Value <- df1$Var1 + df1$Var2 + df1$Var3
df2 <- data.frame(expand.grid(
c(1:20),
c(1:20),
c(1:20)
))
df2$Model <- "2@3d20"
df2$Value <- ifelse(
df2$Var1 > df2$Var2 & df2$Var1 > df2$Var3,
ifelse(
df2$Var2 > df2$Var3,
df2$Var2,
df2$Var3
),
ifelse(
df2$Var2 > df2$Var1 & df2$Var2 > df2$Var3,
ifelse(
df2$Var1 > df2$Var3,
df2$Var1,
df2$Var3
),
ifelse(
df2$Var2 > df2$Var1,
df2$Var2,
df2$Var1
)
)
)
df <- rbind(df1, df2)
dfval <- data.frame(expand.grid(
Value = c(1:20),
Model = c("2@3d20", "3d6")
))

for (val in unique(dfval$Value)) {
for (mod in unique(dfval$Model)) {
if (mod == "2@3d20") {
dfval$Freq[dfval$Value == val & dfval$Model == mod] <- length(df2$Value[df2$Value == val])/nrow(df2)*100
} else if (mod == "3d6") {
dfval$Freq[dfval$Value == val & dfval$Model == mod] <- length(df1$Value[df1$Value == val])/nrow(df1)*100
}
}
}

ggplot(dfval, aes(x=Value, y=Freq, color=Model)) +
geom_point() +
geom_line() +
stat_function(fun = function(x) -0.0752380952380952*(x**2)+1.57952380952381*x-0.791428571428572, aes(color="Formula"))




---
EDIT for new post since I went down the rather lovely and spacious rabbit hole:


How does that work with advantage/disadvantage? The article doesn't appear to say.

If turning it into basically super-advantage and super-disavantage by taking the highest (advantage) or lowest (disadvantage) of the 3d20 instead of the middle value:
https://anydice.com/program/2b928
(Yikes -- technically a curve, but pretty steep!)

If you instead don't roll one die and do regular advantage and disadvantage:
https://anydice.com/program/2b929
(Still quite different, but this is just the existing system for when advantage/disadvantage occur)

[MoiMagnus]

2@3d20 (roll 3d20 take middle) is a great curve and the average lineup vs. 3d6 and 1d20 is great, but "same shape" is stretching things a bit for typical usage (cf. https://anydice.com/program/2b923) unless being VERY broad about what "same shape mathematically speaking" means.

Indeed, my bad. I messed up my math and compared "3d20 take the average" to "3d6 take the sum" (which do have the same shape) instead of "3d20 take the middle one" to "3d6 take the sum" (which are indeed different from one another).

[PhantomSoul]

Indeed, my bad. I messed up my math and compared "3d20 take the average" to "3d6 take the sum" (which do have the same shape) instead of "3d20 take the middle one" to "3d6 take the sum" (which are indeed different from one another).

That makes more sense!

[Composer99]

5e's bounded accuracy is almost the same thing, just without bothering with the treadmill.


That might be true in theory; in practice it's more that if you aren't proficient in things you get worse at them over time, instead of there not being a treadmill at all. Saving throws are a pretty obvious example of that. It's likely less pronounced for ability checks, but I suspect that as adventures become more epic in scale and scope, DCs are naturally going to creep up... meanwhile, PC capabilities aren't keeping up if they aren't proficient.

Also, the thing about downtime training for skills is... well, that it takes downtime. Some campaigns might not have time for that kind of downtime.

[stoutstien]

That might be true in theory; in practice it's more that if you aren't proficient in things you get worse at them over time, instead of there not being a treadmill at all. Saving throws are a pretty obvious example of that. It's likely less pronounced for ability checks, but I suspect that as adventures become more epic in scale and scope, DCs are naturally going to creep up... meanwhile, PC capabilities aren't keeping up if they aren't proficient.

Also, the thing about downtime training for skills is... well, that it takes downtime. Some campaigns might not have time for that kind of downtime.

That's more of a case of a DM, or in some instances the designers, knowing how little the values increase over the character's career. Which is fair because the explanation is spread out amongst three or four books and twice as many sections with little to no indexing.

In theory any saving throw or ability check over 20 should be extremely rare even if the characters are level 20.

[PhantomSoul]

That might be true in theory; in practice it's more that if you aren't proficient in things you get worse at them over time, instead of there not being a treadmill at all. Saving throws are a pretty obvious example of that. It's likely less pronounced for ability checks, but I suspect that as adventures become more epic in scale and scope, DCs are naturally going to creep up... meanwhile, PC capabilities aren't keeping up if they aren't proficient.

I don't really agree with that characterisation (the "worse at them") for two reasons:

1. Facing a given threat, you perform no worse as you increase in level. That matters because a core goal for 5e (and bounded accuracy) is that you actually can face those threats at higher levels. For skills that might partly be masked by conceptual "not worth rolling" tasks changing even when your bonus doesn't change, but it's also where it's most obvious assuming the task DC is actually held constant for an identical task (which seems to be a reasonable assumption from my experience at least).

2. (I don't remember where I was going with there being two... I probably just merged them together in 1? xD I think those two first sentences might have been the two points?)

I do think the non-increase does cause a problem in one spot: ACs. Attack Rolls scale by Attacker Level/CR, but AC doesn't, meaning that there's an asymmetry in part of the equation. But in that case, I see the problem as being that (martial) characters should have scaling, not that you're actually performing worse (since you perform equally well against a given Attack Bonus). This is a big difference between Spell Save DCs and ACs for me, where there's no real "I'm proficient" category at all.

[Segev]

While AC factors into CR, I think the design intent for 5e was that hit points pile up higher and higher as you go up in level, and things just get hit more often but last through more hits. How well this holds up is a matter of some debate, I think.

Yes, saves are the area where the lack of scaling bonuses is most felt, I believe, since DCs do tend to go up as monster stats and proficiencies increase. Paladins are really valuable for this reason, as would be spells like resistance if it weren't so impossible to have it up when you need it. (Unlike guidance, which I prefer as-is even if it might be fine as a reaction, I think resistance absolutely needs to move to a reaction casting time in 5.1e.)

5e also combats this a bit by having all the save-or-suck effects grant a save every round on the victim's turn, and often things that let you save additional times (like when you take damage). It doesn't always, though; hypnotic pattern is a really potent save-or-suck spell for that reason.

[Amechra]

I don't really agree with that characterisation (the "worse at them") for two reasons:

1. Facing a given threat, you perform no worse as you increase in level. That matters because a core goal for 5e (and bounded accuracy) is that you actually can face those threats at higher levels. For skills that might partly be masked by conceptual "not worth rolling" tasks changing even when your bonus doesn't change, but it's also where it's most obvious assuming the task DC is actually held constant for an identical task (which seems to be a reasonable assumption from my experience at least).

This is slightly incorrect... but mostly because how good you perceive your character being at something has less to do with how good you are at specific tasks, and more to do with how good you are at general tasks

If I'm a Monk with proficiency in Acrobatics, my chances of passing "moderate" DC 15 checks, "hard" DC 20 checks, and "very hard" DC 25 checks are as follows:



Level
DC 15
DC 20
DC 25


1-3
55%
30%
5%


4 (+2 Dex)
60%
35%
10%


5-7
65%
40%
15%


8 (+2 Dex)
70%
45%
20%


9-12
75%
50%
25%


13-16
80%
55%
30%


17-20
85%
60%
35%



Now, that chart looks like I'm continually getting better, right? Now, look at what the chart looks like if my DM starts off mostly asking for DC 15 checks, switches to asking for a lot of DC 20 checks at 8th level, and then "graduates" to DC 25 checks at 17th level:



Level
DC 15
DC 20
DC 25


1-3
55%
30%
5%


4 (+2 Dex)
60%
35%
10%


5-7
65%
40%
15%


8 (+2 Dex)
70%
45%
20%


9-12
75%
50%
25%


13-16
80%
55%
30%


17-20
85%
60%
35%




In that kind of game, I'd be well within my rights to describe myself as getting worse at something I used to be good at. Now, granted, this does require a specific kind of failure state (it wouldn't show up if the DM mostly called for Moderate checks regardless of what level the party is), but I feel like it's a very​ common failure state.

[Segev]

In that kind of game, I'd be well within my rights to describe myself as getting worse at something I used to be good at. Now, granted, this does require a specific kind of failure state (it wouldn't show up if the DM mostly called for Moderate checks regardless of what level the party is), but I feel like it's a very​ common failure state.

While it likely is a very common failure state, this is the result of either the DM failing to understand, or the RAW failing to explain, or both, that this is not how the DM is supposed to run the game. Yes, yes, "the DM can run it however he wants," but the game is designed with certain things in mind. Calling for more high-DC checks just because the party is higher level is not one of them. The converse of even high-level PCs having values that struggle with DCs above 15 on certain checks is that there need not be super-high DC obstacles to protect even high-level things.

I think this is yet another thing that would be helped by the DMG providing some examples of what qualifies as "a hard check" under various genres in which the DM could run the game.

[NichG]

Honestly, what I've learned from the discussion in the last page or so is that I think framing things as 'the DM calling for checks' is the root of a lot of problems, and directions towards bounded accuracy and the like have exposed (a different) part of that problem than what was exposed before. E.g. attitudes like 'why have the portcullis if you know the Barbarian can lift it' view the game as first and foremost a sequence of checks against a probability distribution of outcomes, where what is important to preserve is the specific probabilities in that sequence. However that basically reduces the game to chutes and ladders - it means that the consequences of what characters the players choose to play, what actions they choose to take, etc end up compartmentalized in some other part of the game that isn't about this statistical challenge element.

This is as opposed to a framing of the game as including the set of counterfactual ways that play could have gone, if people had chosen differently. That is to say, instead of the one party of specific PCs and everything being about how the world relates to them, imagine a thousand different variations of the party - all reasonable builds and strategic choices and so on, but different ones. The portcullis then is an element that splits that sheaf of alternate possible lines of play into ones in which the party has invested in whatever elements allow them to lift or bypass it, and lines of play in which the party did not make those preparations or investments or particular decisions to spend resources. If there is randomness, that split is a fuzzy one. If there is no randomness, that split is a sharp one. In any one line of play, even if there was no randomness involved in determining whether the portcullis is lifted, the portcullis still serves a purpose in telling people 'if you didn't have a high strength character, you would have been stuck here'.

Another example would be something like the concept of suppressing fire. Its not the case that suppressing fire would be any more lethal to enemy forces than just shooting them directly, or a more effective way to actually land a bullet. Rather, by creating an area that clearly communicates 'if you step into this area, you will likely die', the viable area to maneuver can be controlled, pinning a force down or inducing them to go a certain way. If not a single bullet lands, even if those bullets didn't do what they're directly for, the existence of that communicated consequence still mattered. In the same sense, even if a given party finds a certain thing impossible to do or trivial to do right now, that does not make that element meaningless.

Furthermore, because characters grow, there is an important relationship in showing such things even if they are currently impossible - it presents a long-term strategic choice of 'if you were to spend these resources, then you could make this impossible thing possible'. Whether that choice describes a single instance or a whole class of anticipated future instances, that sort of thing has weight on players' perception of the game.

If you took that philosophy to heart rather than the 'the DM's job is to set mechanical challenges' philosophy, then having something where you do improve by a little bit over 20 levels like in that Monk table would be fine. Because you would be able to say before you got there 'this is the kind of thing that getting my +4 will let me do that I would have struggled with before'. On the other hand, that sort of bounded accuracy / narrow range forms a very bad combination with DMs who do view their job as being about ensuring the right difficulties or failure chances are present.

[KorvinStarmast]

Honestly, what I've learned from the discussion in the last page or so is that I think framing things as 'the DM calling for checks' is the root of a lot of problems

No, it isn't. It's not the problem. What you have seen wasted pages on is a bunch of solutions running about looking for a problem.

[Composer99]

That's more of a case of a DM, or in some instances the designers, knowing how little the values increase over the character's career. Which is fair because the explanation is spread out amongst three or four books and twice as many sections with little to no indexing.

In theory any saving throw or ability check over 20 should be extremely rare even if the characters are level 20.


I don't really agree with that characterisation (the "worse at them") for two reasons:

1. Facing a given threat, you perform no worse as you increase in level. That matters because a core goal for 5e (and bounded accuracy) is that you actually can face those threats at higher levels. For skills that might partly be masked by conceptual "not worth rolling" tasks changing even when your bonus doesn't change, but it's also where it's most obvious assuming the task DC is actually held constant for an identical task (which seems to be a reasonable assumption from my experience at least).

2. (I don't remember where I was going with there being two... I probably just merged them together in 1? xD I think those two first sentences might have been the two points?)

I do think the non-increase does cause a problem in one spot: ACs. Attack Rolls scale by Attacker Level/CR, but AC doesn't, meaning that there's an asymmetry in part of the equation. But in that case, I see the problem as being that (martial) characters should have scaling, not that you're actually performing worse (since you perform equally well against a given Attack Bonus). This is a big difference between Spell Save DCs and ACs for me, where there's no real "I'm proficient" category at all.


While it likely is a very common failure state, this is the result of either the DM failing to understand, or the RAW failing to explain, or both, that this is not how the DM is supposed to run the game. Yes, yes, "the DM can run it however he wants," but the game is designed with certain things in mind. Calling for more high-DC checks just because the party is higher level is not one of them. The converse of even high-level PCs having values that struggle with DCs above 15 on certain checks is that there need not be super-high DC obstacles to protect even high-level things.

I think this is yet another thing that would be helped by the DMG providing some examples of what qualifies as "a hard check" under various genres in which the DM could run the game.

I can't agree that this is DMs doing DMing wrong. Let me repeat something I said in my previous post...


It's likely less pronounced for ability checks, but I suspect that as adventures become more epic in scale and scope, DCs are naturally going to creep up... meanwhile, PC capabilities aren't keeping up if they aren't proficient.


Low level... navigate the local trackless swamp or rugged alpine lands.
High level... navigate the depths of the Abyss or the caverns of Pandemonium.

Low level... break into the strongbox of a human noble with a more-or-less conventional lock.
High level... break into the vaults contained within the palatial estate of an efreeti noble in the City of Brass, with a lock made by a master dao finesmith.

Low level... convince Terzo the Rat to snitch on his thieves' guild buddies.
High level... convince Medea to help you steal the Golden Fleece.

In short, as you go from low level to high level, the challenges you face are expected to grow increasingly epic in scope - at least, that's a reasonable inference from the description of the tiers of play in the PHB.

But a corollary to this is that it's also perfectly reasonable for a DM to assume that these more epic challenges ought to be more difficult to accomplish, and therefore that the DCs ought to be higher. But a natural consequence of this line of thinking is that challenges scale while capabilities lacking proficiency don't, meaning you really do get worse at things as you gain levels if you're not proficient with them.

DMG pg. 238-239 "Setting the DC" is (not surprisingly) entirely unhelpful in setting the record straight (assuming the designer intent runs against the above line of thinking).

[Goobahfish]

There's actually something I liked about the 4e skill system that I wish 5e would have tried - level scaling for off-skills.
One thing that 5e does not capture well is non-proficient high-level characters.

When I was designing my game, the following all had to be a true:

A proficient level 1 character > A non-proficient level 1 character
A proficient level 10 character > A non-proficient level 10 character

AND

A proficient level 10 character > A proficient level 1 character
A non-proficient level 10 character > A non-proficient level 1 character.

5e does... 3 of these, but not the last one (aside from a level 20 character having potentially higher stats).

I find this a bit unsatisfying. While you aren't getting worse (in a strict sense), there is a good chance the tide is rising and you are 'falling behind by default'. 3e and Pathfinder are pretty bad for this too. When you start, as a Wizard you can 'sort of' use a crossbow, but as the game progresses, it becomes less and less viable (which kind of makes level 1 more fun in a way because you can 'try things'). Likewise, with skill checks, the difference between 'proficient' and 'non-proficient' ballooned from a 4 point difference to a 20 point difference.

I resolved it via the following:

Check bonuses = (Skill ranks + Level)/2 + Bonus. So basically, going from 1-10 (I only have 10 levels) you get a default +5 bonus. If you are proficient you get the full +10.

For 5e... I think maybe having it that when other players get +4 for Proficiency you get a +1 general proficiency (i.e. level 9), and when other players get +6 you get +2 (i.e. level 17). It's a bit like 'jack of all trades' but strictly worse. I think it is about enough to 'cover the difference'.

[Telok]

Maybe if this part of the game is so sensitive to DM errors (be they math mistakes, style differences, genre expectations, or textual misunderstandings) in estimating success chances and such, then whats really needed is a bit at the beginning of the DMG saying "take this math test to see if you will screw up ability checks".

My group waa joking about it yesterday. Drowning isn't a thing in combat because it takes so long. But ask for a half dozen swim checks and a PC is sure to drown without magic. WATER IS DEATH!

[Amechra]

While it likely is a very common failure state, this is the result of either the DM failing to understand, or the RAW failing to explain, or both, that this is not how the DM is supposed to run the game. Yes, yes, "the DM can run it however he wants," but the game is designed with certain things in mind. Calling for more high-DC checks just because the party is higher level is not one of them. The converse of even high-level PCs having values that struggle with DCs above 15 on certain checks is that there need not be super-high DC obstacles to protect even high-level things.

I think this is yet another thing that would be helped by the DMG providing some examples of what qualifies as "a hard check" under various genres in which the DM could run the game.

The bolded bits is kinda why I brought this up, honestly (and why I referred to it as a "failure state").

Just to clarify... this isn't suggesting that the DMs in question are going "Oh, hey, you know that thing I let you do at level 1 with a DC 15 check? Well, now it's a DC 20 check!"

This is about DMs that set their early adventures in the kinds of places where the DC to open a lock is 15, and then set later adventures in places where it's a DC 20 check (because the average lock in the Palace of Uun is logically going to be better quality than the ones you'd find in some podunk village, right?). Yes, if you went back to those earlier places, you'd find that the locks were pretty simple to open... but that's small consolation for your Rogue who didn't take Expertise in Thieves' Tools.

[Segev]

The bolded bits is kinda why I brought this up, honestly (and why I referred to it as a "failure state").

Just to clarify... this isn't suggesting that the DMs in question are going "Oh, hey, you know that thing I let you do at level 1 with a DC 15 check? Well, now it's a DC 20 check!"

This is about DMs that set their early adventures in the kinds of places where the DC to open a lock is 15, and then set later adventures in places where it's a DC 20 check (because the average lock in the Palace of Uun is logically going to be better quality than the ones you'd find in some podunk village, right?). Yes, if you went back to those earlier places, you'd find that the locks were pretty simple to open... but that's small consolation for your Rogue who didn't take Expertise in Thieves' Tools.

There's some of that, but there's also the notion that maybe you shouldn't be setting the expectations of "locks that will impede the party" with those in "a podunk village." Locks are rarely "a thing" in poorer medieval areas; they're more likely to hide things or use other methods to secure them. Because locks beyond some basic latches can be pretty darned expensive by peasant standards. This is actually one of the flaws of the "well, just set it at 15, whatever it is, if you don't have a solid idea what DC it should be." This leads to "well, I made this lock in Podunkia DC 15, so obviously all other locks are better and must be 20 or 25!" because you wind up with literally everything you run into at Tier 1 being "the first time" you run into it, and the DM has no inkling what the DC "should" be for most things, so everything is DC 15. Even stuff that maybe shouldn't be rolled at all, because it maybe isn't something that should have had "interesting consequences" or, if it does, should still have been easier to achieve. Maybe. Who knows? The books sure don't give you any useful advice!

[Sorinth]

The bolded bits is kinda why I brought this up, honestly (and why I referred to it as a "failure state").

Just to clarify... this isn't suggesting that the DMs in question are going "Oh, hey, you know that thing I let you do at level 1 with a DC 15 check? Well, now it's a DC 20 check!"

This is about DMs that set their early adventures in the kinds of places where the DC to open a lock is 15, and then set later adventures in places where it's a DC 20 check (because the average lock in the Palace of Uun is logically going to be better quality than the ones you'd find in some podunk village, right?). Yes, if you went back to those earlier places, you'd find that the locks were pretty simple to open... but that's small consolation for your Rogue who didn't take Expertise in Thieves' Tools.

You probably won't like this but when the PCs are level 10 and visit the Palace of Uun you can keep it at DC 15 and if they ever go back to podunk village reduce the DC from the 15 when they were level 1 to 5 or 10 now that they are higher level. There's nothing that says the DC has to be the same throughout your campaign, and there's no reason not to consider player character level when setting the DC. The DMG even has a variant where it does take level into account when determining auto-success, so factoring it into DC is just as logical.

[Amechra]

You probably won't like this but when the PCs are level 10 and visit the Palace of Uun you can keep it at DC 15 and if they ever go back to podunk village reduce the DC from the 15 when they were level 1 to 5 or 10 now that they are higher level. There's nothing that says the DC has to be the same throughout your campaign, and there's no reason not to consider player character level when setting the DC. The DMG even has a variant where it does take level into account when determining auto-success, so factoring it into DC is just as logical.

Nah, that's basically how I run things.

The issue is that I've played with quite a few groups where the DM didn't run things that way.

[MadBear]

This line of questioning brings about an interesting topic. Is the DC supposed to be static and represent the difficulty of anyone in the world from attempting it. Or put another way, if the DC is 15, is that the DC for every PC, NPC, and level throughout the game. Or, is the DC supposed to represent how difficult a given task is for a given player.

If it's the former, you run into Amechra's issue where the DC's will go up as you move to harder and harder areas. If it's the latter, then you have varying DC's between the players (barbarian only need roll a 15 to lift the portcullis, while the wizard needs a 20".

I point this out, because I see both sides here and both sides of this have merit for how to make the system better.


No, it isn't. It's not the problem. What you have seen wasted pages on is a bunch of solutions running about looking for a problem.

what a completely uncharitable and unhelpful statement.

[Segev]

This line of questioning brings about an interesting topic. Is the DC supposed to be static and represent the difficulty of anyone in the world from attempting it. Or put another way, if the DC is 15, is that the DC for every PC, NPC, and level throughout the game. Or, is the DC supposed to represent how difficult a given task is for a given player.

If it's the former, you run into Amechra's issue where the DC's will go up as you move to harder and harder areas. If it's the latter, then you have varying DC's between the players (barbarian only need roll a 15 to lift the portcullis, while the wizard needs a 20".

I point this out, because I see both sides here and both sides of this have merit for how to make the system better.

The DC is the difficulty of the task, no matter who is attempting it. How hard it is for a given creature to pull off the task is represented by a couple of things: primarily by its bonuses or penalties to the d20 roll, and also by whether it's physically capable of the necessary activity (a creature without hands cannot normally do things that requires hands, and a DM may either decide a different task entirely is being attempted if they need to do the same thing, or assign massive penalties; this falls under "exigent circumstances" and is generally outside the scope of standard discussions about setting DCs). Circumstances altering the difficulty are abstracted to the Advantage and Disadvantage system, generally, though the rules for Cover do suggest that there may be times to grant characters actual bonuses or penalties to the roll for such external circumstances (or to just change the DC because you're viewing "circumstances" as part of the task itself - a lock that is burning hot might have a higher DC to pick than a lock that doesn't burn you if you leave your picks in too long or brush its casing...or it might just grant disadvantage to the same DC, depending on the DM's judgment of how important "burning hot" difficulties are to the task at ahnd).

But the DC is the DC. If Bob and Alice are both attempting to break the same manacles, they use the same mechanics and (if those mechanics are a Strength check of some variety) the same DC, because they are the same manacles. By the time you're including, "The task is, 'Bob is breaking the manacles,'" you're effectively giving Bob a bonus or penalty to his roll based on how high his bonus or penalty is. That is, if he's got a high strength and you want high strength to succeed, you're adding MORE bonus to his roll by letting Bob have a lower DC.

So, if you set different DCs just based on who's doing it, you're essentially saying that each character's bonuses or penalties aren't what's on their character sheet, but change for every task.

Because, again, the character's abilities are on his sheet. The task's DC is the task's DC. Different characters have easier or harder times because they have different bonuses or penalties to the roll.

[Elves]

No need for a fancy solution.

5e overcorrected for modifier bloat. In 6e, this simply needs to be corrected so that modifiers max out in the ~+20 or even ~+30 range rather than the ~+10 range.

5e correctly realized there's no need for arbitrary scaling like 4e's +1/2 level mod. It went a little too far is all.

[Sorinth]

This line of questioning brings about an interesting topic. Is the DC supposed to be static and represent the difficulty of anyone in the world from attempting it. Or put another way, if the DC is 15, is that the DC for every PC, NPC, and level throughout the game. Or, is the DC supposed to represent how difficult a given task is for a given player.

If it's the former, you run into Amechra's issue where the DC's will go up as you move to harder and harder areas. If it's the latter, then you have varying DC's between the players (barbarian only need roll a 15 to lift the portcullis, while the wizard needs a 20".

I point this out, because I see both sides here and both sides of this have merit for how to make the system better.

Nothing in the books say the DCs are supposed to be static nor do they say it should change, so it's very much up to the DM. For sure RAW it's perfectly legal for the DC to be different because the DC is only set once a character decides to make the attempt. Keep in mind you probably aren't setting a different DC just because it's fun, there's likely a very logical reason behind it such as treating lifting the portcullis as the % over your normal lift limits.

[Psyren]

No need for a fancy solution.

5e overcorrected for modifier bloat. In 6e, this simply needs to be corrected so that modifiers max out in the ~+20 or even ~+30 range rather than the ~+10 range.

5e correctly realized there's no need for arbitrary scaling like 4e's +1/2 level mod. It went a little too far is all.

You want to double or triple PC modifiers and you think that will solve bloat? :smallconfused:



5e does... 3 of these, but not the last one (aside from a level 20 character having potentially higher stats). I find this a bit unsatisfying.

Concur, that's what got me thinking about it. 5e patches this by essentially encouraging your DM to let you apply proficiency to nearly anything you ask, so that there as few non-proficient-character comparisons as possible - yet they go on to list a bunch of things under each ability where they suggest proficiency shouldn't apply, contradicting themselves.



While you aren't getting worse (in a strict sense), there is a good chance the tide is rising and you are 'falling behind by default'. 3e and Pathfinder are pretty bad for this too. When you start, as a Wizard you can 'sort of' use a crossbow, but as the game progresses, it becomes less and less viable (which kind of makes level 1 more fun in a way because you can 'try things'). Likewise, with skill checks, the difference between 'proficient' and 'non-proficient' ballooned from a 4 point difference to a 20 point difference.

Bounded Accuracy solved this one thankfully (or at least, mitigated it way down.)


There's some of that, but there's also the notion that maybe you shouldn't be setting the expectations of "locks that will impede the party" with those in "a podunk village." Locks are rarely "a thing" in poorer medieval areas; they're more likely to hide things or use other methods to secure them. Because locks beyond some basic latches can be pretty darned expensive by peasant standards. This is actually one of the flaws of the "well, just set it at 15, whatever it is, if you don't have a solid idea what DC it should be." This leads to "well, I made this lock in Podunkia DC 15, so obviously all other locks are better and must be 20 or 25!" because you wind up with literally everything you run into at Tier 1 being "the first time" you run into it, and the DM has no inkling what the DC "should" be for most things, so everything is DC 15. Even stuff that maybe shouldn't be rolled at all, because it maybe isn't something that should have had "interesting consequences" or, if it does, should still have been easier to achieve. Maybe. Who knows? The books sure don't give you any useful advice!

But if the books had said "DC 10 for cheap/Podunkia locks" this would have only delayed the problem. Because as you rightly mentioned, the DM should be thinking about whether "pauper locks" are even a thing in the first place, and if they are, whether they even need a roll.

[Segev]

No need for a fancy solution.

5e overcorrected for modifier bloat. In 6e, this simply needs to be corrected so that modifiers max out in the ~+20 or even ~+30 range rather than the ~+10 range.

5e correctly realized there's no need for arbitrary scaling like 4e's +1/2 level mod. It went a little too far is all.

+20 should be a very extreme maxing out of modifiers, I think.

5e's solution for "you really shouldn't be failing this" is for the DM to make that call and just give you the success, not to try to make the d20 a non-factor. (It still can happen at the highest levels - even a +10 can make certain saves auto-successes - but it shouldn't be designed with the intent that you NEED these high numbers to not have clownish outcomes of failure on things characters shouldn't be failing at. If the task is one that shouldn't be failed, then it doesn't need a roll: people succeed at it.)

[GloatingSwine]

+20 should be a very extreme maxing out of modifiers, I think.

5e's solution for "you really shouldn't be failing this" is for the DM to make that call and just give you the success, not to try to make the d20 a non-factor. (It still can happen at the highest levels - even a +10 can make certain saves auto-successes - but it shouldn't be designed with the intent that you NEED these high numbers to not have clownish outcomes of failure on things characters shouldn't be failing at. If the task is one that shouldn't be failed, then it doesn't need a roll: people succeed at it.)

5e's solution for "you really shouldn't be failing this (but still might and something interesting happens if you do)" is Advantage. Just giving people success is for when they can't fail at all or can keep trying until they succeed.

[RSP]

The DC is the difficulty of the task, no matter who is attempting it.

Disagree. Think of this: what’s the DC to recall the holy symbol of Bane? Is that the DC for someone who’s spent years studying the religions of FRs? Is that the DC for a High Priest of Bane? Is that the DC for someone who’s never seen the holy symbol and never heard of Bane?

If it’s DC X to recall the holy symbol of Bane and “The DC is the difficulty of the task, no matter who is attempting it”; then all those characters have to roll to see if they can recall the symbol.

[Keltest]

Disagree. Think of this: what’s the DC to recall the holy symbol of Bane? Is that the DC for someone who’s spent years studying the religions of FRs? Is that the DC for a High Priest of Bane? Is that the DC for someone who’s never seen the holy symbol and never heard of Bane?

If it’s DC X to recall the holy symbol of Bane and “The DC is the difficulty of the task, no matter who is attempting it”; then all those characters have to roll to see if they can recall the symbol.

If its a commonly held symbol, theres no DC at all, everyone automatically recognizes it by default. If its some obscure alternate or older symbol, then the DC is the same for everyone, and the years of study are represented by the character's bonus to the check.

[GloatingSwine]

Disagree. Think of this: what’s the DC to recall the holy symbol of Bane? Is that the DC for someone who’s spent years studying the religions of FRs? Is that the DC for a High Priest of Bane? Is that the DC for someone who’s never seen the holy symbol and never heard of Bane?

If it’s DC X to recall the holy symbol of Bane and “The DC is the difficulty of the task, no matter who is attempting it”; then all those characters have to roll to see if they can recall the symbol.

Only the first one can make a check, the second doesn't need to and the third never knew the information and so can't recall or even guess it (as they've never heard of Bane to guess at him). The DC of the check for the only person who *can* roll depends on how the DM is running their game and how commonplace and prominent the faith of Bane is.

If there are two characters who are trying to recall a holy symbol at the same time and both are within the range where it is theoretically possible for them to succeed or fail (eg. a cleric with Knowledge(Religion) and A.N. Other character who is sufficiently travelled that they have heard of Bane and may have encountered the holy symbol before) then the person with the highest bonus for the task makes the check with Advantage (because that's how co-operating on a task works).

[RSP]

Only the first one can make a check, the second doesn't need to and the third never knew the information and so can't recall or even guess it (as they've never heard of Bane to guess at him).


So you agree with my post that it’s incorrect to say it’s the same DC for every character for a given task?


If its a commonly held symbol, theres no DC at all, everyone automatically recognizes it by default.

Are you stating this as RAW, or how you’d rule? As I understand it, Bane is not commonly referenced in what is the standard time period of 5e FR.



If it’s some obscure alternate or older symbol, then the DC is the same for everyone, and the years of study are represented by the character's bonus to the check.

So the High Priest of Bane must make a check to see if they recall the symbol, and that DC is the same for the person who has never seen the symbol or heard of Bane?

[Psyren]

I agree that each task should have a set DC (that some characters can bypass entirely - e.g. the high priest of Moander recognizing Moander's* symbol) - but I also agree that two approaches that look to be the same task at a surface glance ultimately might be two completely different ones.

For example "gain audience with the king who is being manipulated by his evil vizier" is a very different task for a bard from a foreign land vs. one of the king's trusted knights, but it's not automatic for either of them. When you drill down, that task becomes two different tasks: "vassal with close relationship tries to gain audience with his king" and "complete stranger from other nation tries to gain audience with their king." Thus different DCs would be warranted. And a third task entirely might be "sneak into king's bedroom late at night to speak with him there instead of trying to gain an audience at all", which gets a third DC.

*swapping Moander for Bane as he is more obscure

[MadBear]

For example "gain audience with the king who is being manipulated by his evil vizier" is a very different task for a bard from a foreign bard vs. one of the king's trusted knights, but it's not automatic for either of them. When you drill down, that task becomes two different tasks: "vassal with close relationship tries to gain audience with the king" and "complete stranger tries to gain audience with the king." Thus different DCs would be warranted. And a third task entirely might be "sneak into king's bedroom late at night to speak with him there instead of trying to gain an audience", which gets a third DC.

Could that exact same line of thinking not be used to have different DC's for a variety of other tasks?

The barbarian who can regularly lift 600lbs, should have a lower DC to lift a 500lb portcullis then the strength 14 ranger. Because the former is someone doing something they've done on a regular basis and well within their wheelhouse while the latter is someone trying to lift something heavier then what they could normally lift? As you get more granular, doesn't that imply that the task you're asking both to do is different enough to warrant a different DC? The same could be said of any skill as well if we decide to break it down far enough.

(i'm not in necessary disagreement with you btw, more I'm curious how you could parse one without including the other).

[RSP]

I agree that each task should have a set DC (that some characters can bypass entirely - e.g. the high priest of Moander recognizing Moander's* symbol) - but I also agree that two approaches that look to be the same task at a surface glance ultimately might be two completely different ones.

For example "gain audience with the king who is being manipulated by his evil vizier" is a very different task for a bard from a foreign bard vs. one of the king's trusted knights, but it's not automatic for either of them. When you drill down, that task becomes two different tasks: "vassal with close relationship tries to gain audience with the king" and "complete stranger tries to gain audience with the king." Thus different DCs would be warranted. And a third task entirely might be "sneak into king's bedroom late at night to speak with him there instead of trying to gain an audience", which gets a third DC.

Def agree that how the task is framed is important. Also agree that what we know about the character can also be important (“trusted knight” vs foreigner to the land), as well as how one approaches an obstacle. These lead to differences in DCs for what one may say is the same “task”.


Could that exact same line of thinking not be used to have different DC's for a variety of other tasks?

The barbarian who can regularly lift 600lbs, should have a lower DC to lift a 500lb portcullis then the strength 14 ranger. Because the former is someone doing something they've done on a regular basis and well within their wheelhouse while the latter is someone trying to lift something heavier then what they could normally lift? As you get more granular, doesn't that imply that the task you're asking both to do is different enough to warrant a different DC? The same could be said of any skill as well if we decide to break it down far enough.

(i'm not in necessary disagreement with you btw, more I'm curious how you could parse one without including the other).

How one frames a task, in my opinion, is very important.

For instance: the party needs to get through a door. Is the task “getting past the door” DC 15? Does that mean whether they’re bashing it down using brute force, picking the lock, or convincing the guard on the other side to unlock it, it’s DC 15 regardless?

Or is it a strong door with a crap lock? Or an easily duped guard?

What one considers to be the task is, indeed, very important in determining how difficult said task is.

[GloatingSwine]

So you agree with my post that it’s incorrect to say it’s the same DC for every character for a given task?

For all characters who have a DC for the task.

Characters who automatically succeed or are incapable of attempting are not relevant to the discussion of what the DC is for those who live between those extremes.

[Rukelnikov]

I agree that each task should have a set DC (that some characters can bypass entirely - e.g. the high priest of Moander recognizing Moander's* symbol) - but I also agree that two approaches that look to be the same task at a surface glance ultimately might be two completely different ones.

For example "gain audience with the king who is being manipulated by his evil vizier" is a very different task for a bard from a foreign bard vs. one of the king's trusted knights, but it's not automatic for either of them. When you drill down, that task becomes two different tasks: "vassal with close relationship tries to gain audience with the king" and "complete stranger tries to gain audience with the king." Thus different DCs would be warranted. And a third task entirely might be "sneak into king's bedroom late at night to speak with him there instead of trying to gain an audience", which gets a third DC.

*swapping Moander for Bane as he is more obscure


Def agree that how the task is framed is important. Also agree that what we know about the character can also be important (“trusted knight” vs foreigner to the land), as well as how one approaches an obstacle. These lead to differences in DCs for what one may say is the same “task”.

How one frames a task, in my opinion, is very important.

For instance: the party needs to get through a door. Is the task “getting past the door” DC 15? Does that mean whether they’re bashing it down using brute force, picking the lock, or convincing the guard on the other side to unlock it, it’s DC 15 regardless?

Or is it a strong door with a crap lock? Or an easily duped guard?

What one considers to be the task is, indeed, very important.

The thing is, there's a point when no two tasks are ever the same.

Character tries to paint portrait of the king.

Same character attempts to paint portrait of the king again, its a different DC since he has the experience from the previous time.

Same character attempts to paint portrait of the king again, its a different DC since he has the experience from the previous time.

etc.

So, overall, I think DCs should be set, and Adv/Disadv/Auto Pass/Fail should be the "modifiers". In the example above, every try after the first could have advantage, but the DC would remain the same.

Also, I don't think using knowledge stuff for these examples is good (like the holy symbol thing), since those are pretty different from the rest of things the system tries to represent in skills, ones require the character to have done something in the past (gain the knowledge they are trying to retrieve), the others require attempting something here and now, where stuff done previously doesn't guarantee one can do it or fail to (having swam againt the river once doesn't mean they will be able to do it again), and IMO knowledge like skills are not well represented by the system.

[GloatingSwine]

What one considers to be the task is, indeed, very important in determining how difficult said task is.

By the time you're setting a DC for an ability check the players have told you what they want to try doing and how. That's how you know whether and what they're going to roll.

[Psyren]

Def agree that how the task is framed is important. Also agree that what we know about the character can also be important (“trusted knight” vs foreigner to the land), as well as how one approaches an obstacle. These lead to differences in DCs for what one may say is the same “task”.

Indeed (though, see below).


Could that exact same line of thinking not be used to have different DC's for a variety of other tasks?

The barbarian who can regularly lift 600lbs, should have a lower DC to lift a 500lb portcullis then the strength 14 ranger. Because the former is someone doing something they've done on a regular basis and well within their wheelhouse while the latter is someone trying to lift something heavier then what they could normally lift? As you get more granular, doesn't that imply that the task you're asking both to do is different enough to warrant a different DC? The same could be said of any skill as well if we decide to break it down far enough.

(i'm not in necessary disagreement with you btw, more I'm curious how you could parse one without including the other).

As noted above, RSP and I agree on the king example, because the nature of the characters' connection to the king (or lack thereof in the Bard's case) matters and can change the fundamental nature of the task, which means the task's DC changes too.

For the portcullis example specifically, I don't necessarily agree with RSP's approach - I think it's the same portcullis whether the barbarian is trying to lift it or the wizard. I have made it a static object in my world; the barbarian doesn't have a different connection to it by being a barbarian. Therefore, I would adjudicate it one of 5 ways:

1) Extremely heavy (impossible for both - functionally a wall. No roll.)
2) Extremely light (automatic for both - functionally a door. No roll.)
3) So heavy that it's impossible for the wizard, but possible for the barbarian (wizard autofails, barbarian needs to roll).
4) Light enough that it's possible for the wizard and trivial for the barbarian (wizard needs to roll, barbarian autosucceeds).
5) Medium such that they both need to roll (see below.)

Of these scenarios, #5 is the one where you run the risk of an incongruous or unsatisfying result - i.e. both of them are rolling, meaning both have a chance of success or failure, meaning there's a nonzero chance that the barbarian fails and the wizard succeeds, even if you try putting a finger on the scales via advantage/disadvantage. Rather than take that risk, the way I would resolve #5 would be by modulating a "failure" from the barbarian to mean "progress combined with setback" (PHB 173) - essentially turning #5 into #4 but with a complication of some kind. The Barbarian "fails" but lifts the portcullis anyway, but maybe they take some damage or they're restrained until they make a second check to throw it off etc. Whereas a fail from the wizard means it doesn't budge.

[Keltest]

So the High Priest of Bane must make a check to see if they recall the symbol, and that DC is the same for the person who has never seen the symbol or heard of Bane?

Sure. And the High Priest of Bane can pass the check fairly consistently because of their good bonus to Religion, whereas some rando cannot unless they happen to be a scholar of religion in some manner.

[GloatingSwine]

Of these scenarios, #5 is the one where you run the risk of an incongruous or unsatisfying result - i.e. both of them are rolling, meaning both have a chance of success or failure, meaning there's a nonzero chance that the barbarian fails and the wizard succeeds, even if you try putting a finger on the scales via advantage/disadvantage. Rather than take that risk, the way I would resolve #5 would be by modulating a "failure" from the barbarian to mean "progress combined with setback" (PHB 173) - essentially turning #5 into #4 but with a complication of some kind. The Barbarian "fails" but lifts the portcullis anyway, but maybe they take some damage or they're restrained until they make a second check to throw it off etc. Whereas a fail from the wizard means it doesn't budge.

But you don't if you actually do the situation properly. There is no set of circumstances where the Barbarian and Wizard should both be trying to lift the same portcullis and rolling separately.

Either there is room for them both to try at the same time and it's an assisted check (Barbarian with advantage) or there's time for them to try in sequence in which case the Barbarian tries twice and the Wizard does something else.

The only time it's an "incongruous result" is if they both tried to lift two identical but separate portcullises at the same time, but that would be a silly position to put those two characters in unless you expected the Wizard to be using a spell to help with their portcullis.

[Psyren]

But you don't if you actually do the situation properly. There is no set of circumstances where the Barbarian and Wizard should both be trying to lift the same portcullis and rolling separately.

Either there is room for them both to try at the same time and it's an assisted check (Barbarian with advantage) or there's time for them to try in sequence in which case the Barbarian tries twice and the Wizard does something else.

The only time it's an "incongruous result" is if they both tried to lift two identical but separate portcullises at the same time, but that would be a silly position to put those two characters in unless you expected the Wizard to be using a spell to help with their portcullis.

I agree lifting something isn't the best example but I stand by the sequence. Replace lifting a portcullis with deciphering ancient runes on a wall for instance; helping each other or the same character trying twice isn't always possible.

Also, advantage/disadvantage are not catch-all solutions here either. Your players might already have advantage or disadvantage on their own going into the challenge for instance, meaning you lose that lever.

[RSP]

For all characters who have a DC for the task.

Characters who automatically succeed or are incapable of attempting are not relevant to the discussion of what the DC is for those who live between those extremes.

DC is a range of difficultly, assessed by the DM. That assessment includes whether a task is auto fail or auto pass.

If you allow that some characters autopass or auto fail a task, but not others, why are locking in the DM’s assessment at that point? You’ve already decided that differences in the characters allow for different assessments of their ability to complete a task, why then say you can’t assess that a task is easier for one than another?

Going back to a lifting example: you have a 500 lbs portcullis.

The Str 20 Barb (600 lbs max lift) is determined to autopass lifting it - no roll required.

The Str 8 Wizard (240 lbs max lift) is determined to auto fail - no roll required).

The Str 12 character though (360 lbs max lift) do they get to roll?

The Str 15 character (450 lbs), do they get to roll?

They’re only separated by a +1 modifier, or 5% on a d20 roll. Should the difference in chances to lift 50 lbs over your max and 140 lbs over your max be 5%?

Put another way, should increasing your max lift ~11% be only 5% less difficult than increasing your max lift ~39%?

Why is it DC 15 (for example) for the Str 12 character to lift 140 lbs over their max, but also DC 15 for the Str 15 character to lift 50 lbs over their max?

[Keltest]

I agree lifting something isn't the best example but I stand by the sequence. Replace lifting a portcullis with deciphering ancient runes on a wall for instance; helping each other or the same character trying twice isn't always possible.

Also, advantage/disadvantage are not catch-all solutions here either. Your players might already have advantage or disadvantage on their own going into the challenge for instance, meaning you lose that lever.

History checks as presented by the book are the only ones where I think thats an appropriate solution, because theyre technically a check to "recall" information you already read about at one point, not a check to see if you encountered it. Now, I havent actually heard of any groups that play it the former rather than the latter because playing the former fairly requires defining more of a character's life than is really feasible even in a prepublished setting like the Realms, let alone a DM's personal setting.

If you want the untrained, untalented Barbarian to be unable to succeed, set the DC higher than 20. Otherwise, the intent of the system is that it isnt an incongruous result for the Barbarian to occasionally, rarely succeed where the Wizard cannot.

[RSP]

But you don't if you actually do the situation properly. There is no set of circumstances where the Barbarian and Wizard should both be trying to lift the same portcullis and rolling separately.

Either there is room for them both to try at the same time and it's an assisted check (Barbarian with advantage) or there's time for them to try in sequence in which case the Barbarian tries twice and the Wizard does something else.

The only time it's an "incongruous result" is if they both tried to lift two identical but separate portcullises at the same time, but that would be a silly position to put those two characters in unless you expected the Wizard to be using a spell to help with their portcullis.

Not necessarily. What if on going through the tunnel, the Barb lifted the 600 lbs portcullis. But then the group got split, and now the Wizard is trying to retreat. The portcullis is still 600 lbs and again requiring lifting.

Same DC for the Str 8 Wiz as it was the Str 20 Barb? Are they not each trying to lift 600 lbs in portcullis form, separately?


History checks as presented by the book are the only ones where I think thats an appropriate solution, because theyre technically a check to "recall" information you already read about at one point, not a check to see if you encountered it. Now, I havent actually heard of any groups that play it the former rather than the latter because playing the former fairly requires defining more of a character's life than is really feasible even in a prepublished setting like the Realms, let alone a DM's personal setting.

I think it’s more common than you've encountered.

Our current campaign is this way (I’m not the DM). Checks like these are regularly accompanied by something like “would my character’s association with X be enough to get a roll for this [knowledge check]?”

The DM will sometimes then go with “no, that doesn’t apply to this” or “yeah, give me a X check”, or maybe just “yeah, you know …. about this.”

[Psyren]

Not necessarily. What if on going through the tunnel, the Barb lifted the 600 lbs portcullis. But then the group got split, and now the Wizard is trying to retreat. The portcullis is still 600 lbs and again requiring lifting.

Same DC for the Str 8 Wiz as it was the Str 20 Barb? Are they not each trying to lift 600 lbs in portcullis form, separately?

Agreed, great example.



I think it’s more common than you've encountered.

Our current campaign is this way (I’m not the DM). Checks like these are regularly accompanied by something like “would my character’s association with X be enough to get a roll for this [knowledge check]?”

The DM will sometimes then go with “no, that doesn’t apply to this” or “yeah, give me a X check”, or maybe just “yeah, you know …. about this.”

Ours work this way too.


History checks as presented by the book are the only ones where I think thats an appropriate solution, because theyre technically a check to "recall" information you already read about at one point, not a check to see if you encountered it. Now, I havent actually heard of any groups that play it the former rather than the latter because playing the former fairly requires defining more of a character's life than is really feasible even in a prepublished setting like the Realms, let alone a DM's personal setting.

Arcana could be used to decipher runes without history or having seen these specific ones before.


If you want the untrained, untalented Barbarian to be unable to succeed, set the DC higher than 20. Otherwise, the intent of the system is that it isnt an incongruous result for the Barbarian to occasionally, rarely succeed where the Wizard cannot.

If you want to run your games that way, great. I'm sticking with the 5 possibilities I laid out earlier.

[Keltest]

Arcana could be used to decipher runes without history or having seen these specific ones before. I bring up History because thats the only skill that explicitly, universally is supposed to be dependent on the character making the check to determine how difficult it is. And I have never heard of anybody actually running it that way in practice because its basically just the DM arbitrarily deciding if they know something and then, if they do, rolling to see if they dont get it anyway.


If you want to run your games that way, great. I'm sticking with the 5 possibilities I laid out earlier.

I'm not about to beat you with a stick for playing wrong or anything, but the specific intention of the game is that the "problem" you keep talking about is a feature, not a bug.

[Psyren]

I bring up History because thats the only skill that explicitly, universally is supposed to be dependent on the character making the check to determine how difficult it is. And I have never heard of anybody actually running it that way in practice because its basically just the DM arbitrarily deciding if they know something and then, if they do, rolling to see if they dont get it anyway.



I'm not about to beat you with a stick for playing wrong or anything, but the specific intention of the game is that the "problem" you keep talking about is a feature, not a bug.

Do you have a rules quote supporting either of these positions?

[Keltest]

Do you have a rules quote supporting either of these positions?

PHB 177-178

History. Your Intelligence (History) check measures your ability to recall lore about historical events, legendary people, ancient kingdoms, past disputes, recent wars, and lost civilizations.

Emphasis mine. You are recalling information you already learned, not rolling to see if you ever learned something. This is thus inherently based on information your character already theoretically had access to, which is thus individualized.

As for the second one, its the entire premise of the bounded accuracy system.

[Sorinth]

I agree that each task should have a set DC (that some characters can bypass entirely - e.g. the high priest of Moander recognizing Moander's* symbol) - but I also agree that two approaches that look to be the same task at a surface glance ultimately might be two completely different ones.


For the portcullis example specifically, I don't necessarily agree with RSP's approach - I think it's the same portcullis whether the barbarian is trying to lift it or the wizard. I have made it a static object in my world; the barbarian doesn't have a different connection to it by being a barbarian.

I think many would say that lifting the portcullis might look the same at a surface glance but are ultimately two completely different ones. Namely for one character it's trying to exceed your standard lift by 1 pound and for the other it's trying to exceed your standard lift by 301 pounds. That sounds like two different tasks no?

[Psyren]

I think many would say that lifting the portcullis might look the same at a surface glance but are ultimately two completely different ones. Namely for one character it's trying to exceed your standard lift by 1 pound and for the other it's trying to exceed your standard lift by 301 pounds. That sounds like two different tasks no?

I'm not saying that reading is invalid but I wouldn't use it. RSP would. Neither of us is wrong.

For me, the difference between the king and the portcullis is that the king's relationship to the characters is subjective, while the portcullis' relationship to them is objective. That's my take.



Emphasis mine. You are recalling information you already learned, not rolling to see if you ever learned something. This is thus inherently based on information your character already theoretically had access to, which is thus individualized.

For the record, I don't use History to see if you ever learned something either, but I don't see what that has to do with the arcane runes example. Determining their meaning does not mean you have to recall having seen them before. I can figure out most rebuses even if I've never seen them either. I'd be inclined to let a player use both their Arcana or their History proficiency (if they had either) without too much pushback.


As for the second one, its the entire premise of the bounded accuracy system.

There being no quote to support this as the One True Playstyle is exactly the sticking point in these threads. In fact, the book itself specifically allows you to not ask people to roll on everything, for any reason or no reason at all - DMG 236.

[Elves]

You want to double or triple PC modifiers and you think that will solve bloat? :smallconfused:
I said 5e overcorrected for the bloat in prior editions. D20 plus or minus 10, with DCs capped at 30, is far too small. There’s too little difference between the worst and the best, it’s too random, and we see this constantly with the absurdities the skill system produces. (It’s also reflected in the fact that 5e-world doesn’t really need adventurers and heroes to exist: a bunch of low-level guards can take care of most monsters, especially if they have a couple magic weapons between them.)

The minimum you want for a max deviation is equal to the value of the die (+20), not half the die. It’s kind of shocking that 5e devs didn’t understand this, but they were recognizing a past problem and banking too hard the other way.

Another problem with having the modifier ceiling be so low is that it’s in fundamental opposition to the system. The d20 system was designed as an arithmetic system where good things are expressed as pluses and bad things as minuses. But because of its commitment to low scaling, 5e has to go out of its way to avoid pluses and minuses whenever possible! It’s a system at war with itself.

[RSP]

PHB 177-178

History. Your Intelligence (History) check measures your ability to recall lore about historical events, legendary people, ancient kingdoms, past disputes, recent wars, and lost civilizations.

Emphasis mine. You are recalling information you already learned, not rolling to see if you ever learned something. This is thus inherently based on information your character already theoretically had access to, which is thus individualized.

As for the second one, its the entire premise of the bounded accuracy system.

“Recall” is used with all the knowledges, not just History. This, to me, is a reason why factoring in the character is so important to these checks: you can’t recall something you’ve never seen, heard about, or otherwise experienced.

It’s why I like the example of a god’s holy symbol. If you don’t know about the god, and never saw the holy symbol before, there’s zero chance you’ll recall what god represents the symbol.

Yet if you worship the god and wear the symbol on your person daily, I don’t know how you fail to recall that god/symbol.

If you want to draw conclusions about an unknown god, based on the viewing of their symbol, cool; but that’s not the same task as trying to recall what you’ve previously learned.

And there’s no reason to enforce a rule that says the DC is the same for each task: “recalling what you’ve previously learned about a god” vs “drawing inferences about a holy symbol”.

[OldTrees1]

I said 5e [B]overcorrected[B] for the bloat in prior editions. D20 plus or minus 10, with DCs capped at 30, is far too small. There’s too little difference between the worst and the best, it’s too random, and we see this constantly with the absurdities the skill system produces. (It’s also reflected in the fact that 5e-world doesn’t really need adventurers and heroes to exist: a bunch of low-level guards can take care of most monsters, especially if they have a couple magic weapons between them.)

The minimum you want for a max deviation is equal to the value of the die (+20), not half the die. It’s kind of shocking that 5e devs didn’t understand this, but they were recognizing a past problem and banking too hard the other way.

I think the devs were too focused on opposed checks (or more accurately on Attacks, Saves, Spell DCs, and Armor). For opposed checks it is less obvious what the minimum max deviation is. (In contrast to solo checks where +20 as the minimum max deviation is quite intuitive).

[PhantomSoul]

PHB 177-178

History. Your Intelligence (History) check measures your ability to recall lore about historical events, legendary people, ancient kingdoms, past disputes, recent wars, and lost civilizations.

Emphasis mine. You are recalling information you already learned, not rolling to see if you ever learned something. This is thus inherently based on information your character already theoretically had access to, which is thus individualized.

As for the second one, its the entire premise of the bounded accuracy system.

You can only recall something you've previously known/seen/heard, so it's really both. Hence it being history (did you ever encounter this information?) not just straight-INT (in my experience, more typical for the generic ability to remember).

[Goobahfish]

This whole discussion just sounds like 3e circumstance bonuses under the hood.

Basically, yes... a DC should be a fixed in stone thing. How hard is this task. That makes sense. Fiddling DCs... sounds dumb.

However, in 3e, you had 'circumstance bonuses' which was basically DM makes a logical description to themselves of why this character would be great at task X and this character poor. Hence, you get +5, you get -5, the DC remains fixed but the math is exactly the same as all the DC twiddling.

Except... the math is slightly harder? Because you have to add in some extra numbers to some other big numbers after a roll rather than just fudging the DC by a few points (the player gives you a number, you compare to a fudged DC which you can do before or while they are rolling).

5e kinda made Bound Accuracy which implies all tasks are possible to all people (gamist). Some players here obviously want that not to be true (i.e. impossible tasks). I am one of them, there are certain tasks that should be impossible for some and easy for others. Bounded Accuracy doesn't do this inherently without hidden circumstance bonuses (which kind of contradicts the idea of bounded accuracy).

Huzzahs!

[Segev]

As noted above, RSP and I agree on the king example, because the nature of the characters' connection to the king (or lack thereof in the Bard's case) matters and can change the fundamental nature of the task, which means the task's DC changes too.As do I, I feel the need to chime in to add.


For the portcullis example specifically, I don't necessarily agree with RSP's approach - I think it's the same portcullis whether the barbarian is trying to lift it or the wizard. I have made it a static object in my world; the barbarian doesn't have a different connection to it by being a barbarian. Therefore, I would adjudicate it one of 5 ways:

1) Extremely heavy (impossible for both - functionally a wall. No roll.)
2) Extremely light (automatic for both - functionally a door. No roll.)
3) So heavy that it's impossible for the wizard, but possible for the barbarian (wizard autofails, barbarian needs to roll).
4) Light enough that it's possible for the wizard and trivial for the barbarian (wizard needs to roll, barbarian autosucceeds).
5) Medium such that they both need to roll (see below.)

Of these scenarios, #5 is the one where you run the risk of an incongruous or unsatisfying result - i.e. both of them are rolling, meaning both have a chance of success or failure, meaning there's a nonzero chance that the barbarian fails and the wizard succeeds, even if you try putting a finger on the scales via advantage/disadvantage. Rather than take that risk, the way I would resolve #5 would be by modulating a "failure" from the barbarian to mean "progress combined with setback" (PHB 173) - essentially turning #5 into #4 but with a complication of some kind. The Barbarian "fails" but lifts the portcullis anyway, but maybe they take some damage or they're restrained until they make a second check to throw it off etc. Whereas a fail from the wizard means it doesn't budge.
For me, (3) doesn't exist.

And it comes down to the fact that for (3) to exist, there has to be something fundamentally different about the tasks being performed, as with the "king's retainer" vs "Bard from a foreign land" thing. "being a wizard" isn't relevant to "using raw strength to lift a gate" in a way that can render it impossible for him.

Or, put another way, if the only thing making it "too heavy for the wizard" is the wizard's strength score, then that should be fully represented by the DC of the check.

Now, if the Barbarian has things that are doubling or quadrupling his lift capacity, maybe that might factor in; it's no longer a function of his strength score, alone, then.

[Telok]

I think the devs were too focused on opposed checks (or more accurately on Attacks, Saves, Spell DCs, and Armor). For opposed checks it is less obvious what the minimum max deviation is. (In contrast to solo checks where +20 as the minimum max deviation is quite intuitive).

The thing is that those checks scale over 20 levels by the hit points (totals & damages) scaling, and the inherent d20 randomness is mitigated by the damage rolls that end up. Its that all those rolls combined make a "one fight" dice pool for a very reliable bell curve probability. If you swap the dice pool fight ability for a single d20 roll to determine the outcome of a combat it goes back to being lol-random, just with dead characters instead of stuff like failed knowledge checks.

Hmm... combat is pretty dull anyways most of the time if you aren't casting spells. Wonder if I could come up with a transform to tally & roll a whole fight dice pool at once. Get fights over with faster so we can get back to the plot and not the stupid filler **** dictated by the inflexible resource drain paradigm.

[Psyren]

And it comes down to the fact that for (3) to exist, there has to be something fundamentally different about the tasks being performed, as with the "king's retainer" vs "Bard from a foreign land" thing. "being a wizard" isn't relevant to "using raw strength to lift a gate" in a way that can render it impossible for him.

Or, put another way, if the only thing making it "too heavy for the wizard" is the wizard's strength score, then that should be fully represented by the DC of the check.

I never said "he can't lift it because he's a wizard." It was shorthand for his strength being low and lacking athletics proficiency. And wizards have plenty of options for lifting heavy things without those two things.

[MadBear]

As do I, I feel the need to chime in to add.


For me, (3) doesn't exist.

And it comes down to the fact that for (3) to exist, there has to be something fundamentally different about the tasks being performed, as with the "king's retainer" vs "Bard from a foreign land" thing. "being a wizard" isn't relevant to "using raw strength to lift a gate" in a way that can render it impossible for him.

Or, put another way, if the only thing making it "too heavy for the wizard" is the wizard's strength score, then that should be fully represented by the DC of the check.

Now, if the Barbarian has things that are doubling or quadrupling his lift capacity, maybe that might factor in; it's no longer a function of his strength score, alone, then.

It seems to me this example is a prime representation of when 3 would exist. A strength 20 character can left 600lbs. So a 500lb gate isn't an issue. Let's really exaggerate it and say it's a strength 8 wizard can left at most 240lbs. (from the x30 rule under weights).

In this very specific instance, I can totally see this not needing a roll for the barbarian (it's 100lbs under his threshhold) and nont allowing it for the wizard (its double what he can push or lift).

[Composer99]

On the general topic, I would disagree with the OP - ability checks should use a d20. But in that case it's key to stick to either or both of:
- having multiple checks to accomplish a task (meaning any task worth calling for a check should be one worth calling for a few), which reduces the difficulty posed by the variance;
- having the uncertainty regarding success or failure to be sufficient as to justify the variance of the d20.

I'm also in disagreement with Elves. If anything, 5e could stand to have a bit of a trim of die roll modifiers!

Mind you, I'm also of a mind that most DCs for ability checks - say, two-thirds or so (or maybe 68.2% :smallwink:) - should be within the range of 10-15. (When in doubt, roll 1d6 + 9 for the DC to give to a player, you might say.) You could even use the standard bell curve distribution as a guideline for how often DCs should appear within higher ranges: 27.2% (or just over a quarter) of the time for DCs between 15 and 20, no more than 1 times in 20 (5%) for a DC between 20 and 25, and vanishingly rarely for anything higher.

Since I've noted that as the kinds of adventures PCs go on become inherently more challenging as they gain levels, to my mind the way to represent this for ability checks without having to escalate DCs is to have one or two (two tops!) keywords/tags that you can add to tasks that add some kind of "you must be this tall to take this ride" criteria to attempt them normally. These criteria could impose a "hard" barrier (e.g. you can't do something if you don't meet the criteria) or a "soft" barrier (e.g. doing something when you don't meet the criteria is a lot harder). One tag might relate to having a proficiency relevant to a task (but doesn't have to), and the other (if you want to have an "epic/mythic challenge" feel) might be, say, a proficiency bonus of +4 or higher.

My go-to example for this is the trap that Haley is working on in OOTS #867 (https://www.giantitp.com/comics/oots0867.html), which to my mind is clearly beyond the reach of the untrained, but may well also be more fitting of an epic-style challenge beyond the means of even a trained but low-level character. But with these keywords/tags to apply, you could slap a DC 20 on it and if our 5ePC!Haley is 20 Dex with expertise in thieves' tools (net bonus of +13 from 9th-12th level, which is IIRC about the level range the protagonists are by that point in the story), she would succeed on any one check as part of a disarm attempt 60% of the time (before factoring in, say, bardic inspiration from 5ePC!Elan as shown in the fiction of the comic). If you decide to make this a best-two-out-of-three "encounter" (as it were), that's just about a two-thirds chance of success overall before bardic inspiration if I've got the probabilities correct.

IMO, a two-thirds chance of success is about on the nose for what is going to acceptably feel challenging to a player (regardless of its statistical probability), so if 5ePC!Haley has invested in thieves' tools and Dexterity enough to reach this point, and gets the boost from 5ePC!Elan, the PC party has earned their likely triumph over the trap- ... er... well, door.

As far as verisimilitude goes, I think we can all think of examples of stuff from real life that are easy enough (more or less) for folks with sufficient training while being very difficult (if not impossible) for folks without, from coding to flying airplanes to various sorts of surgery.


Chiming in on the specific issue of lifting heavy stuff,, to my mind the difficulty of lifting a very heavy object (i.e. one that is nominally beyond 30 times your Strength score) is more realistically (and game...er...ifically?) represented as being proportional to the ratio of that object's weight to your lifting capacity.

That is to say, (1) attempting to lift a very heavy object that is, say, 1.25 times your lifting capacity and attempting to lift such an object that is instead twice your lifting capacity are, in fact, different tasks; (2) it follows logically that attempting to lift the latter is innately more difficult than the former; and (3) if a task's DC is a representation of its innate difficulty, then it also follows logically that the latter task ought to have a higher DC.

Ergo, if you as DM allow both a Strength 8 character and a Strength 20 character to attempt to lift a 700 pound weight, they can face different DCs with absolutely no incongruity. They may be trying to lift the same object, they simply are not trying to attempt the same task.

[OldTrees1]

The thing is that those checks scale over 20 levels by the hit points (totals & damages) scaling, and the inherent d20 randomness is mitigated by the damage rolls that end up. Its that all those rolls combined make a "one fight" dice pool for a very reliable bell curve probability. If you swap the dice pool fight ability for a single d20 roll to determine the outcome of a combat it goes back to being lol-random, just with dead characters instead of stuff like failed knowledge checks.

Hmm... combat is pretty dull anyways most of the time if you aren't casting spells. Wonder if I could come up with a transform to tally & roll a whole fight dice pool at once. Get fights over with faster so we can get back to the plot and not the stupid filler **** dictated by the inflexible resource drain paradigm.

Indeed, these things are true.

[RSP]

For me, (3) doesn't exist.

And it comes down to the fact that for (3) to exist, there has to be something fundamentally different about the tasks being performed, as with the "king's retainer" vs "Bard from a foreign land" thing. "being a wizard" isn't relevant to "using raw strength to lift a gate" in a way that can render it impossible for him.

Or, put another way, if the only thing making it "too heavy for the wizard" is the wizard's strength score, then that should be fully represented by the DC of the check.

Now, if the Barbarian has things that are doubling or quadrupling his lift capacity, maybe that might factor in; it's no longer a function of his strength score, alone, then.

How is it not different to try and lift something that you can lift vs trying to lift something you can’t lift?

If the Str 8 Wizard can lift a 600 lbs portcullis with a DC 15 roll, shouldn’t the Str 20 Barb lift about 1400 lbs with a DC 15 roll?

Why would that be the same for them?

[Keltest]

How is it not different to try and lift something that you can lift vs trying to lift something you can’t lift?

If the Str 8 Wizard can lift a 600 lbs portcullis with a DC 15 roll, shouldn’t the Str 20 Barb lift about 1400 lbs with a DC 15 roll?

Why would that be the same for them?
So, if youre using the lifting capacity rules for the portcullis, which is a fair enough judgement, then you shouldnt be rolling a skill check in the first place. Youre already using one system, you dont need to cross it over with another one.

But if you arent, then determining if they can lift the gate is the whole point of the D20 check.

[Yakk]

So, if youre using the lifting capacity rules for the portcullis, which is a fair enough judgement, then you shouldnt be rolling a skill check in the first place. Youre already using one system, you dont need to cross it over with another one.
Lifting up to your capacity doesn't require a check.

Lifting over your capacity requires a check.

Similarly, running at your speed? No check. Putting in extra effort to go faster? A check.

Casting a spell from a scroll under the level of slots you have? No check. Casting a higher level scroll? Check.

Even if a 3rd level wizard has 20 intelligence and jack of all trades, they still need to roll to cast a fireball from a scroll, while the 8 intelligence no arcana proficiency dunce who is level 5 does it without a check. Even though Jack has a +8 to her check, and Dunce a -1.

Athletics doesn't make you stronger. Athletics trains you in how to push your limits and control yourself.

Someone trained in athletics can figure out the perfect stance to lift that gate if they pull off the check; someone without training who is stronger can lift it without a check.

[MadBear]

So, if youre using the lifting capacity rules for the portcullis, which is a fair enough judgement, then you shouldnt be rolling a skill check in the first place. Youre already using one system, you dont need to cross it over with another one.

But if you arent, then determining if they can lift the gate is the whole point of the D20 check.

So if we were to go with that, I can modifiy the question slightly. Instead of a 600lb gate, it's a 601lb gate.

For the barbarian that's 1lb heavier then their max. For the strength 8 wizard that's 361lbs over their max. Why would the DC for those two be the same? Put another way, shouldn't the barbarian's DC be lower because we're talking about going just slightly higher then their maximum while the wizard is significantly over their maximum.

[Keltest]

So if we were to go with that, I can modifiy the question slightly. Instead of a 600lb gate, it's a 601lb gate.

For the barbarian that's 1lb heavier then their max. For the strength 8 wizard that's 361lbs over their max. Why would the DC for those two be the same? Put another way, shouldn't the barbarian's DC be lower because we're talking about going just slightly higher then their maximum while the wizard is significantly over their maximum.

While it is within the DM's purview to allow just about anything, I dont see anything in the lifting or ability check rules that specifically indicates people should expect to even try to be able to lift above their capacity. Thats sort of the point of the limit in the first place. This is in contrast to the jump rules, which do specifically say its not a hard limit, just what you can do consistently.

Which is to say, as near as I can tell youre already in DM fiat territory. Fiat in whichever direction you prefer.

[Yakk]

So if we were to go with that, I can modifiy the question slightly. Instead of a 600lb gate, it's a 601lb gate.

For the barbarian that's 1lb heavier then their max. For the strength 8 wizard that's 361lbs over their max. Why would the DC for those two be the same? Put another way, shouldn't the barbarian's DC be lower because we're talking about going just slightly higher then their maximum while the wizard is significantly over their maximum.

Sure. Here are some fun DCs:

DC 10: Just over your max lift.
DC 15: About 50% higher than your max lift.
DC 20: About twice your max lift.
DC 25: About three times your max lift.
DC 30: Way more than 3x your max lift.

Strength 20: 600 lbs
Strength 8: 120 lbs

Gate is 650 lbs. Barbarian DC is 10; a bit over max lift. Wizard DC is 30, way beyond what they can lift.

Wizard has expertise athletics, guideance and bull's strength, for 2d20 pick best +11+1d4 mod. Has about a 1/3 chance to lift it.

Barbarian is untrained athletics, so has a 20% chance of failing. Needs to learn technique (trained athletics - becomes automatic), or use a rage (4% failure remains)

[animorte]

We are aware the definition of max (maximum), yes?

[Psyren]

Sure. Here are some fun DCs:

DC 10: Just over your max lift.
DC 15: About 50% higher than your max lift.
DC 20: About twice your max lift.
DC 25: About three times your max lift.
DC 30: Way more than 3x your max lift.

Strength 20: 600 lbs
Strength 8: 120 lbs

Gate is 650 lbs. Barbarian DC is 10; a bit over max lift. Wizard DC is 30, way beyond what they can lift.

Wizard has expertise athletics, guideance and bull's strength, for 2d20 pick best +11+1d4 mod. Has about a 1/3 chance to lift it.

Barbarian is untrained athletics, so has a 20% chance of failing. Needs to learn technique (trained athletics - becomes automatic), or use a rage (4% failure remains)

This is exactly the kind of thing I would not want codified. Because the DM should have the purview to say "yes, that's fine" or "no, you can't lift 3x your maximum even with a level 1 bard and a cantrip assisting you" on a case by case basis without any kind of argument.

[Yakk]

We are aware the definition of max (maximum), yes?
Yes, and we are all aware that in English, you can say things like "push beyond your maximum" and it isn't actually nonsense.

Because it is English, not a programming language or formal logic.

[animorte]

Yes, and we are all aware that in English, you can say things like "push beyond your maximum" and it isn't actually nonsense.

Because it is English, not a programming language or formal logic.

Pushing beyond your maximum and drastically exceeding are entirely different.

[Rukelnikov]

This is exactly the kind of thing I would not want codified. Because the DM should have the purview to say "yes, that's fine" or "no, you can't lift 3x your maximum even with a level 1 bard and a cantrip assisting you" on a case by case basis without any kind of argument.

The DM always has the purview to say "this is fine, this isn't", whether a table like that exists or not.


So, if youre using the lifting capacity rules for the portcullis, which is a fair enough judgement, then you shouldnt be rolling a skill check in the first place. Youre already using one system, you dont need to cross it over with another one.

But if you arent, then determining if they can lift the gate is the whole point of the D20 check.

I agree, if you are using the lifting rules to determine the Barb can lift it, then don't call for a roll to see if he can, at most use the roll to determine how long it takes them.

[Psyren]

The DM always has the purview to say "this is fine, this isn't", whether a table like that exists or not.

Of course they do. But the existence of this table will cause more table arguments and DM's needing to napkin math out their world than it not existing. (Oh crap, how heavy did I make that portcullis again? 550 lbs? Should that have been 650? Carry the three...)

[Rukelnikov]

Of course they do. But the existence of this table will cause more table arguments and DM's needing to napkin math out their world than it not existing. (Oh crap, how heavy did I make that portcullis again? 550 lbs? Should that have been 650? Carry the three...)

But the DMs that will pay attention to the table probably like paying that kind of attention to their world, and those that don't care for that can ignore the table, or use it just a general idea.

[Keltest]

Yes, and we are all aware that in English, you can say things like "push beyond your maximum" and it isn't actually nonsense.

Because it is English, not a programming language or formal logic.

I mean, its either using the word incorrectly or we understand from context that its not meant to be interpreted literally. If you see a sign that says the maximum height to drive under a bridge is 15 feet, nothing you do will get that 15' 6" truck under that bridge safely.

[PhantomSoul]

I mean, its either using the word incorrectly or we understand from context that its not meant to be interpreted literally. If you see a sign that says the maximum height to drive under a bridge is 15 feet, nothing you do will get that 15' 6" truck under that bridge safely.

Or, in a case like this, reasonably an important difference between a maximum carrying capacity (seemingly intended to be what you can bear to lift and carry around for a relatively extended period of time) rather than just your maximum weights that can be lifted, pushed or pulled as a burst (which can all be different; so a single effort with all energy in, and potentially adrenaline to boot). So the "maximum" might not actually be the issue! 5e gives us only one real number and we've got to decide how to use it (actually, that feels like it's about as clear as 5e gets, in hindsight!).

[Psyren]

But the DMs that will pay attention to the table probably like paying that kind of attention to their world, and those that don't care for that can ignore the table, or use it just a general idea.

"Make a table that is likely to cause arguments and hope it gets ignored" is not an approach to design I can get behind.


I mean, its either using the word incorrectly or we understand from context that its not meant to be interpreted literally. If you see a sign that says the maximum height to drive under a bridge is 15 feet, nothing you do will get that 15' 6" truck under that bridge safely.

To be fair, a character is (usually) not a truck. A hole labelled 4' clearance could still potentially be squeezed through by a Goliath. I don't see a problem with "exceed maximum" as shorthand, I just think the degree to which that can be done and even the circumstances under which it can be attempted or not should be fully up to the DM.

[Keltest]

To be fair, a character is (usually) not a truck. A hole labelled 4' clearance could still potentially be squeezed through by a Goliath. I don't see a problem with "exceed maximum" as shorthand, I just think the degree to which that can be done and even the circumstances under which it can be attempted or not should be fully up to the DM.

You are correct (a character can duck, for example), but the point stands, sometimes a hard limit is a hard limit. And in the case of lifting a weight, its given as a hard limit, the same as maximum hit points or movement speed. They allow for exceeding the limit for jumping, so its clearly something theyve considered at least once.

[KorvinStarmast]

what a completely uncharitable and unhelpful statement. As is your title.
Skill Checks shouldn't use a d20

I said 5e overcorrected for the bloat in prior editions. D20 plus or minus 10, with DCs capped at 30, is far too small. they were recognizing a past problem and banking too hard the other way. Interesting point. Their solution met the 'good enough' standard - if you are demanding perfection they certainly fell short.
(Yes, bounded accuracy was apparently devised with combat most in mind)

I think the devs were too focused on opposed checks (or more accurately on Attacks, Saves, Spell DCs, and Armor). For opposed checks it is less obvious what the minimum max deviation is. (In contrast to solo checks where +20 as the minimum max deviation is quite intuitive). Had not thought of it in that light, but it certainly sheds a different light on the numbers issue/DC table noise.

I would disagree with the OP - ability checks should use a d20. But in that case it's key to stick to either or both of:
- having multiple checks to accomplish a task (meaning any task worth calling for a check should be one worth calling for a few), which reduces the difficulty posed by the variance;
- having the uncertainty regarding success or failure to be sufficient as to justify the variance of the d20.


Mind you, I'm also of a mind that most DCs for ability checks - say, two-thirds or so (or maybe 68.2% :smallwink:) - should be within the range of 10-15. Based on the published adventures I have run, that range seems to be very common. A DC 16 check gets to be a bit tougher if the PC rolls with disadvantage, and setting up disadvantage situations requires more thought, or, if it's in a combat scenario, particular skills/abilities of enemies.

Ergo, if you as DM allow both a Strength 8 character and a Strength 20 character to attempt to lift a 700 pound weight, they can face different DCs with absolutely no incongruity. They may be trying to lift the same object, they simply are not trying to attempt the same task.

RSP made a good point that I missed yesterday, in that 'framing the task' is important.
Yes.
Wrestle that waitress.
Wrestle that waitress in a mud filled ring
Wrestle that waitress in a mud filled ring after you have had 5 flagons of ale.
Wrestle that waitress in a mud filled ring after you have had 5 flagons of ale and you don't know she's a brown belt in Judo.
(This occurred one fine evening a few decades ago at a certain bar in California, and a friend of mine provided the entertainment) Needless to say, he did not succeed ... we drove him home even though he was utterly covered in mud.

[Psyren]

You are correct (a character can duck, for example), but the point stands, sometimes a hard limit is a hard limit. And in the case of lifting a weight, its given as a hard limit, the same as maximum hit points or movement speed. They allow for exceeding the limit for jumping, so its clearly something theyve considered at least once.

My take on it is that the lift limit in D&D isn't totally nuanced. The so-called maximum capacity still allows you to move, and it involves lifting something up to 5ft up. So it stands to reason if you're only lifting something up to anywhere between a few inches to three feet, and totally unable to move - such as the mother-lifting-car-to-free-child example - that the amount of weight should be slightly above that "hard limit."

Relating that back to the portcullis example and the "5 scenarios" I mentioned before - maybe the portcullis is just too heavy for the wizard, and even the Barbarian is struggling (it's definitely above his listed "maximum" too.) A "failure" from the Barbarian on his Athletics check, rather than meaning the portcullis stays shut, might mean he only manages a few feet of clearance and is completely immobilized. The rest of the party must now squeeze through the Small gap he's created, which both slows their escape and gives the pursuing lizardfolk advantage with their javelins as they close in. Meaningful consequence for failure achieved, but the Barbarian player still feels their character quite strong as he has still managed to accomplish something no one else in the party could. And maybe if he burns a use of his Rage he gets another roll (with advantage) to lift it higher than he currently has it so he can get through; if he fails that, the party might need to find another way for him to get around, or he might even die heroically having saved his friends.

[Keltest]

My take on it is that the lift limit in D&D isn't totally nuanced. The so-called maximum capacity still allows you to move, and it involves lifting something up to 5ft up. So it stands to reason if you're only lifting something up to anywhere between a few inches to three feet, and totally unable to move - such as the mother-lifting-car-to-free-child example - that the amount of weight should be slightly above that "hard limit."

Point of order: Lifting does not allow you to move, you are inherently immobile while lifting. If you attempt to move while lifting something, you are now carrying it, and use the rules for carrying instead, namely that your effective weight limit is cut in half.

[Rukelnikov]

"Make a table that is likely to cause arguments and hope it gets ignored" is not an approach to design I can get behind.

That's not what I'm saying.

What I do say is, give us lots of examples, for those of us who want them, and those that don't want them can freely ignore page X to page Y.

[Keltest]

That's not what I'm saying.

What I do say is, give us lots of examples, for those of us who want them, and those that don't want them can freely ignore page X to page Y.

"If you dont like it then houserule it" is never going to be a helpful response when talking about game design decisions.

[Psyren]

That's not what I'm saying.

What I do say is, give us lots of examples, for those of us who want them, and those that don't want them can freely ignore page X to page Y.

I would rather they do literally anything else with their development resources.


Point of order: Lifting does not allow you to move, you are inherently immobile while lifting. If you attempt to move while lifting something, you are now carrying it, and use the rules for carrying instead, namely that your effective weight limit is cut in half.

If that's true the rules are very unclear about that. They don't say anywhere that your speed becomes zero while lifting. And even if they did, my point about a lift of less than 5 feet stands.

[Keltest]

If that's true the rules are very unclear about that. They don't say anywhere that your speed becomes zero while lifting. And even if they did, my point about a lift of less than 5 feet stands.

Your speed doesnt become 0 because you can still move (assuming the thing youre lifting is not fixed to a mechanism like the portcullis), its just that by doing so while lifting an object, you are no longer lifting it, you are carrying it.

[Psyren]

Your speed doesnt become 0 because you can still move (assuming the thing youre lifting is not fixed to a mechanism like the portcullis), its just that by doing so while lifting an object, you are no longer lifting it, you are carrying it.

But Carrying and Lifting have different limits (15x and 30x respectively). So if moving turns the latter into the former, what happens if you're already at 30? The closest we get is that if you're in excess of your capacity, your speed drops to 5ft, but even that is only for pushing and dragging, not lifting. And if we are meant to apply that to lifting too, that means you can move 5ft at 30x, so how high can you get while completely unable to move? That I think is what the Athletics check would be for.

[Segev]

I never said "he can't lift it because he's a wizard." It was shorthand for his strength being low and lacking athletics proficiency. And wizards have plenty of options for lifting heavy things without those two things.

Right. But the thing is, "his strength is low and he isn't proficient in athletics" is already represented in the system by having a lower bonus to the d20 roll. This is just as sensible within the rules as saying that a rogue with a javelin cannot hit a knight on horseback because he's not proficient and his strength is not very high, so he's not even allowed to roll with his net -1 to hit against the knight's 18 AC. Or saying that the barbarian with no Wisdom Save proficiency and a low Wisdom cannot even try to make the saving throw against the DC 15 mind control spell, because he's not proficient and his wisdom is too low. You rely on the rule for ability checks that says the DM decides when a roll is called for, but the trouble is, you've already decided a roll is called for when you let the barbarian attempt it at DC 18. Because whether a roll is called for or not, like its DC, is task-dependent, not task-attempter-dependent. The consequences of success or failure are obviously interesting either way. The task is technically possible, but not guaranteed. The wizard can make a DC 18 strength check 10% of the time, so if 18 is the appropriate DC, then the wizard does, in fact, have a chance of succeeding. Condition (3) cannot exist, except as a matter of technicality on the DC and bonuses / penalties. All factors you're citing to disqualify the wizard are already represented, and a DC 18 task is within the realm of possibility for him.

Whereas with the bard from a foreign land vs. a knight who's been in the king's service and his personal friend for years are actually attempting two different tasks, due to plot tokens one has and the other does not. There's no need for "plot tokens" to represent the wizard's low strength and nonproficiency in athletics; the numbers already encompass those.

[Keltest]

But Carrying and Lifting have different limits (15x and 30x respectively). So if moving turns the latter into the former, what happens if you're already at 30? The closest we get is that if you're in excess of your capacity, your speed drops to 5ft, but even that is only for pushing and dragging, not lifting. And if we are meant to apply that to lifting too, that means you can move 5ft at 30x, so how high can you get while completely unable to move? That I think is what the Athletics check would be for.

Presumably the same thing that happens if you were to go from not lifting it to try and carry it, which is to say either the DM says no or you become heavily encumbered.

[Composer99]

"Make a table that is likely to cause arguments and hope it gets ignored" is not an approach to design I can get behind.


The available empirical evidence, such as it is, does not support your position.

The 5e DMG includes tables of explicit DCs for tracking (page 244) and conversation reactions based on NPC attitude (page 245), and on these boards at least there is a just about complete lack of threads specifically created with respect to DMs complaining about their players arguing with them over those tables or players complaining about DMs ignoring those tables. I am a frequent participant at ENWorld, and I can attest to a similar dearth of conflict over the existence of those tables as such. I'm as not familiar with D&D discussions on Reddit, YouTube, or other major sites, so I won't say anything as firm about those - however if disagreement or conflict about those tables were widespread on big-name platforms such as YouTube or Reddit, or if big-name D&D streamers were posting videos articulating disagreements over such tables, I feel comfortable assuming there would be enough spillover that we'd see more discussion of them around here.

The DMG explicitly frames the tracking DCs table about as you would expect ("The Tracking DCs table offers guidelines for setting the DC, or, if you prefer, you can choose a DC based on your assessment of the difficulty."), while not including any such allowance in the Conversation Reactions table ("The creature's current attitude determines the DC required to achieve a specific reaction, as shown in the Conversation Reaction table.") - in that latter case the DM's authority to make use of or ignore the table at their discretion remains implicit following from statements elsewhere in the DMG to that effect.

The disagreements I do find when searching these boards relate to hypothetical addition of further/stronger guidelines to a revised DMG, or to the way that the 1D&D playtest makes these tables player-facing (which I would say is more likely to engender conflict), but not any DM/player conflict over the existence and use of these tables themselves.

Does the 5e DMG need a table of DCs for how much a character can temporarily exceed their lifting capacity? I don't think so - instead it might be helpful to have a paragraph such as "If you're encouraging a more epic feel in your games, consider allowing characters to make ability checks to temporarily lift weights that exceed their normal lifting capacity" with a few sentences suggesting how you might set a DC, as part of a wider discussion of concrete ways the DM can make the game work within the "flavour of fantasy" they're going for in their campaign.

But asserting that the existence of such a table - or other ones besides - would engender conflict? The existing evidence not only doesn't support that assertion, but contradicts it.

[Segev]

The available empirical evidence, such as it is, does not support your position.

The 5e DMG includes tables of explicit DCs for tracking (page 244) and conversation reactions based on NPC attitude (page 245), and on these boards at least there is a just about complete lack of threads specifically created with respect to DMs complaining about their players arguing with them over those tables or players complaining about DMs ignoring those tables. I am a frequent participant at ENWorld, and I can attest to a similar dearth of conflict over the existence of those tables as such. I'm as not familiar with D&D discussions on Reddit, YouTube, or other major sites, so I won't say anything as firm about those - however if disagreement or conflict about those tables were widespread on big-name platforms such as YouTube or Reddit, or if big-name D&D streamers were posting videos articulating disagreements over such tables, I feel comfortable assuming there would be enough spillover that we'd see more discussion of them around here.

The DMG explicitly frames the tracking DCs table about as you would expect ("The Tracking DCs table offers guidelines for setting the DC, or, if you prefer, you can choose a DC based on your assessment of the difficulty."), while not including any such allowance in the Conversation Reactions table ("The creature's current attitude determines the DC required to achieve a specific reaction, as shown in the Conversation Reaction table.") - in that latter case the DM's authority to make use of or ignore the table at their discretion remains implicit following from statements elsewhere in the DMG to that effect.

The disagreements I do find when searching these boards relate to hypothetical addition of further/stronger guidelines to a revised DMG, or to the way that the 1D&D playtest makes these tables player-facing (which I would say is more likely to engender conflict), but not any DM/player conflict over the existence and use of these tables themselves.

Does the 5e DMG need a table of DCs for how much a character can temporarily exceed their lifting capacity? I don't think so - instead it might be helpful to have a paragraph such as "If you're encouraging a more epic feel in your games, consider allowing characters to make ability checks to temporarily lift weights that exceed their normal lifting capacity" with a few sentences suggesting how you might set a DC, as part of a wider discussion of concrete ways the DM can make the game work within the "flavour of fantasy" they're going for in their campaign.

But asserting that the existence of such a table - or other ones besides - would engender conflict? The existing evidence not only doesn't support that assertion, but contradicts it.

Well said! Thank you for all of this.


As for lifting->carrying, I would argue that, if you're lifting more than you can carry, you just can't move. Moving your feet to enable motion may destabilize the delicate balance and/or overburden the one foot left holding things up, causing you to fall over. Maybe, if the DM desires and feels the consequences are interesting either way, he might set a DC for a Strength (Athletics) check or even a Dexterity (Acrobatics) check to keep your balance while moving with more than you can carry.

[Psyren]

Right. But the thing is, "his strength is low and he isn't proficient in athletics" is already represented in the system by having a lower bonus to the d20 roll.

Sure, but given bounded accuracy and the lowest possible starting ability score for a PC using Point Buy or Standard Array being 8, you can still achieve incongruously high results on that roll, especially with a level 1 bard with a cantrip amping you up and *numerically* erasing any deficiency you might have. I'm not saying that should never be allowed, but I think the DM should get to decide when that makes sense and when it doesn't, which they get to do by choosing when to call for a roll or not.



The 5e DMG includes tables of explicit DCs for tracking (page 244) and conversation reactions based on NPC attitude (page 245), and on these boards at least there is a just about complete lack of threads specifically created with respect to DMs complaining about their players arguing with them over those tables or players complaining about DMs ignoring those tables.

That's because there are barely any threads about these tables at all. It wouldn't surprise me if most playgroups didn't know they were there. I view them not as contentious, but as a waste of space. Rather than add a bunch more of them for a bunch more skill uses, I would rather use the space for something more impactful.

The edition that had tables all over the place for a variety of skills was 3.5, and arguments/exploits about those skills happened all the time.

[Segev]

Sure, but given bounded accuracy and the lowest possible starting ability score for a PC using Point Buy or Standard Array being 8, you can still achieve incongruously high results on that roll, especially with a level 1 bard with a cantrip amping you up and *numerically* erasing any deficiency you might have. I'm not saying that should never be allowed, but I think the DM should get to decide when that makes sense and when it doesn't, which they get to do by choosing when to call for a roll or not.

I get that, but if the 90-lb. wet noodle of a creature is only able to do it with a bard's (magical?) encouragement and/or a divine blessing, surely the burly strongman can also get these things to increase his own capacity.

I do not find it incongruous that the weakling wizard strongarms open a gate that a barbarian can barely lift without assistance if the wizard also can barely do it and needs magical aid to even have that chance. A DC 20 is "hard," by the rules. It isn't "nigh impossible." A -1 to the roll makes it impossible, though. But merely having some encouragement or a bit of divine guidance apparently can let them have at least a chance to succeed.

Your own position on setting DCs is that you should set them based on the probability of various kinds of training and talent being able to do the task at hand. You clearly have in mind a level of training and talent that you see as a bare minimum to succeed. I would personally suggest that you set the DC such that a 20 on the die is required before you can hit the DC with that level of training and skill. If magic or other buffs are added to replace training and skill, so what? Those can be given to somebody with the right training and skill to make him even more likely to succeed.

The error that leads to situation 3, to my mind, is setting the DC too low. The barbarian is not strong enough that he should have that much greater a chance of lifting the gate than the wizard, if the DC is that low.


That's because there are barely any threads about these tables at all. It wouldn't surprise me if most playgroups didn't know they were there. I view them not as contentious, but as a waste of space. Rather than add a bunch more of them for a bunch more skill uses, I would rather use the space for something more impactful.

The edition that had tables all over the place for a variety of skills was 3.5, and arguments/exploits about those skills happened all the time.

I don't find them to be wastes of space. I do think they need to be highlighted better so that DMs - such as myself - don't need to be told they're there in order to hunt them down and find them. I found the ones I know of to be VERY useful when DMing, because they helped me establish a baseline of what the game expects "easy" and "hard" to be for the realms they cover, and I can extrapolate and interpolate from there.

[Telok]

"If you dont like it then houserule it" is never going to be a helpful response when talking about game design decisions.

Corollary: "there's no rules or guidelines for that, you have to house rule it" is never going to be a helpful response when talking about game design decisions unless you want to design a Lasers & Feelings one page & one number, type of game.

[Keltest]

Corollary: "there's no rules or guidelines for that, you have to house rule it" is never going to be a helpful response when talking about game design decisions unless you want to design a Lasers & Feelings one page & one number, type of game.

A ruling is not a houserule though. Its a ruling. Thats just the DM doing their job as intended.

[Psyren]

I get that, but if the 90-lb. wet noodle of a creature is only able to do it with a bard's (magical?) encouragement and/or a divine blessing, surely the burly strongman can also get these things to increase his own capacity.

I think there is room to say "the burly strongman can use those things to do something that's possible for him more easily, while the wet noodle cannot use those things to make the impossible possible." I'm totally okay with agreeing to disagree on that.



The error that leads to situation 3, to my mind, is setting the DC too low. The barbarian is not strong enough that he should have that much greater a chance of lifting the gate than the wizard, if the DC is that low.

The problem I have is that with bounded accuracy, there is no DC that is totally out of reach of a commoner with a level 1 bard nearby. It's extremely unlikely at DC 30, but I don't want "extremely unlikely", I want truly impossible for them. There is no/shouldn't be any bardic song that will allow a seal to climb a vertical tree, especially not at level 1 - that doesn't mean that tree has to be DC 30.

Sure I can break that and say DC 35 or DC 45, but on top of being unsupported in the system, that doesn't feel particularly good for the Barbarian player either.

[Segev]

The problem I have is that with bounded accuracy, there is no DC that is totally out of reach of a commoner with a level 1 bard nearby. It's extremely unlikely at DC 30, but I don't want "extremely unlikely", I want truly impossible for them. There is no/shouldn't be any bardic song that will allow a seal to climb a vertical tree, especially not at level 1 - that doesn't mean that tree has to be DC 30.

Sure I can break that and say DC 35 or DC 45, but on top of being unsupported in the system, that doesn't feel particularly good for the Barbarian player either.

The problem I have here is that if you bring in a bard and a cleric to both bolster the rock-common commoner, it's not ridiculous that the commoner with literal divine guidance AND a morale boost from the bard could possibly pull off such a thing.

If you really, truly, cannot stand magic making the commoner able to do that even on a natural 20, there are things other than skill checks that can do what you want better. Rather than a skill check, make it require a total strength score. If you want to threshold it further, maybe make it so that it requires a total strength score to even attempt it. This actually gets into the space you're looking for, I think, and makes room for the wizard and bard and cleric to work together to lift it.

To elaborate with an example, if you set the minimum strength score required to even attempt to lift it to 20, the barbarian almost certainly has that strength score, and can make the attempt. The wizard and his Strength 10 bard buddy can't even manage that, together, but if their cleric friend steps in as well, the three of them will certainly meet the minimum strength threshold. Now the highest-bonus member of the team can make the ability check, probably with advantage from the Help action as well as the bonuses from bardic inspiration and guidance, since the casters are there.

So, I suppose really I just want a bit more codification of the task so that it's clear what the wizard or hte commoner(s) need(s) to do it. And for it not to be something that says, "No matter what you do, you can't succeed, unless you do something that ignores your physical strength entirely." (Obviously, wizards have other options than physical strength, especially if the task is really about bypassing the portcullis rather than specifically lifting it.)

[Sindeloke]

I'm as not familiar with D&D discussions on Reddit, YouTube, or other major sites, so I won't say anything as firm about those

I'm on D&D reddit, and have 6000x more exposure to D&D tiktok than I would like because one of my players incessantly spams them to our Discord, and I can confirm for you that it's not there either.

[RSP]

Right. But the thing is, "his strength is low and he isn't proficient in athletics" is already represented in the system by having a lower bonus to the d20 roll.

RAW, there is no Athletics applied to lifting: Athletics applies to Jumping, Climbing and Swimming.

To your point, you’re ignoring what the Str stat means, and only wanting to embrace the modifier it gives. Having a Str 20 means something in the game. If you throw that out (that is, choose not to consider what we know about Str 20, like they can lift 600 lbs), then you’re making a choice that stats don’t matter.

You’re essentially allowing a d20 roll to effectively increase the characters strength score by over 200%.

You can fully make that choice, just realize that’s what you’re doing.



This is just as sensible within the rules as saying that a rogue with a javelin cannot hit a knight on horseback because he's not proficient and his strength is not very high, so he's not even allowed to roll with his net -1 to hit against the knight's 18 AC. Or saying that the barbarian with no Wisdom Save proficiency and a low Wisdom cannot even try to make the saving throw against the DC 15 mind control spell, because he's not proficient and his wisdom is too low. You rely on the rule for ability checks that says the DM decides when a roll is called for, but the trouble is, you've already decided a roll is called for when you let the barbarian attempt it at DC 18. Because whether a roll is called for or not, like its DC, is task-dependent, not task-attempter-dependent. The consequences of success or failure are obviously interesting either way. The task is technically possible, but not guaranteed. The wizard can make a DC 18 strength check 10% of the time, so if 18 is the appropriate DC, then the wizard does, in fact, have a chance of succeeding. Condition (3) cannot exist, except as a matter of technicality on the DC and bonuses / penalties. All factors you're citing to disqualify the wizard are already represented, and a DC 18 task is within the realm of possibility for him.

First off, no, your “just as sensible” isn’t just as sensible.

Now, lifting almost 300% of your max lift isn’t “very difficult” it’s impossible.

Likewise lifting something less than your max lift shouldn’t be difficult at all.

You’re ignoring what those DC’s mean. Why is it a very difficult task for the Str 20 Barb to lift 1 lbs more than their max, and the same level of difficulty for a Str 8 character to lift 361 lbs more?

Shouldn’t it be easier to lift 1 lbs more than 361 lbs more?

[Segev]

Now, lifting almost 300% of your max lift isn’t “very difficult” it’s impossible.

Likewise lifting something less than your max lift shouldn’t be difficult at all.

You’re ignoring what those DC’s mean. Why is it a very difficult task for the Str 20 Barb to lift 1 lbs more than their max, and the same level of difficulty for a Str 8 character to lift 361 lbs more?

Shouldn’t it be easier to lift 1 lbs more than 361 lbs more?

This parses to me like you're saying, "There shouldn't be a roll at all. Just check maximum lift." Is that what you're saying?

[OldTrees1]

This parses to me like you're saying, "There shouldn't be a roll at all. Just check maximum lift." Is that what you're saying?

RSP's position seems to be:

RSP sets the DC of the check based on how many times heavier the lift is compared to their Str Score's max lift. HOWEVER RSP then has the character use their differing Str modifiers without accounting for that difference too.

So for example if character A's max lift was LA and character B's max lift was LB, then the DC for A to lift 110% of LA would be the same as the DC for B to lift 110% of LB. This forgets the Strength modifiers for A and B but is otherwise quite a reasonable way of handling the DC.

The steelman version of RSP's position is to then call for a 1d20 check instead of a Str check. This removes the different Str modifiers since they were already accounted for in the Str Score based DC.

It probably results from Str Score's contribution to max lift vs Str Score's contribution to Str checks being incongruous. So RSP fixes it with subjective DCs.

[Psyren]

The problem I have here is that if you bring in a bard and a cleric to both bolster the rock-common commoner, it's not ridiculous that the commoner with literal divine guidance AND a morale boost from the bard could possibly pull off such a thing.

I disagree, I do find that ridiculous. A cantrip and a level 1 class feature should not overcome any and all physical limitations.

(Moreover, you don't need the cleric even in 5e. You definitely won't need them in 1DnD.)


If you really, truly, cannot stand magic making the commoner able to do that even on a natural 20, there are things other than skill checks that can do what you want better. Rather than a skill check, make it require a total strength score. If you want to threshold it further, maybe make it so that it requires a total strength score to even attempt it. This actually gets into the space you're looking for, I think, and makes room for the wizard and bard and cleric to work together to lift it.

Sure I could do all that, but I'd rather play 5th edition, in which I can choose not to call for a roll if something is impossible.

[Rukelnikov]

"If you dont like it then houserule it" is never going to be a helpful response when talking about game design decisions.

If its not there in the first place you are gonna have to houserule it anyway.

[RSP]

This parses to me like you're saying, "There shouldn't be a roll at all. Just check maximum lift." Is that what you're saying?

For which? Barb lifting within their max lift is no check, for me.
Barb trying to increase their max say to within 700 lbs (roughly 15% increase) would be DC 15, 800 lbs (roughly within ~30% increase) DC 20.

Over ~30% of max is getting close to impossible.

Likewise for Str 8 Wiz: 15% increase would be DC 15. ~30% increase DC 19 (so need a nat 20). More than that, not possible.

And yes, this makes it harder, less likely to pass a check the weaker you are. However, as I said previously, that makes sense to me as the stronger, fitter someone is, the more likely they can do stuff in excess of their norm.

For instance, if I’m regularly lifting and my body is in shape, I can do more that I’m not regularly doing without injury. If I’m not, and I try to go outside what I’m regularly trying, there’s a good chance I’ll not succeed, and get injured in the process.

I’m not saying 5e needs to be reflective of real life, I’m just saying it makes sense that stronger, fit people are more capable of pushing their limits of strength.

[PhoenixPhyre]

For which? Barb lifting within their max lift is no check, for me.
Barb trying to increase their max say to within 700 lbs (roughly 15% increase) would be DC 15, 800 lbs (roughly within ~30% increase) DC 20.

Over ~30% of max is getting close to impossible.

Likewise for Str 8 Wiz: 15% increase would be DC 15. ~30% increase DC 19 (so need a nat 20). More than that, not possible.

And yes, this makes it harder, less likely to pass a check the weaker you are. However, as I said previously, that makes sense to me as the stronger, fitter someone is, the more likely they can do stuff in excess of their norm.

For instance, if I’m regularly lifting and my body is in shape, I can do more that I’m not regularly doing without injury. If I’m not, and I try to go outside what I’m regularly trying, there’s a good chance I’ll not succeed, and get injured in the process.

I’m not saying 5e needs to be reflective of real life, I’m just saying it makes sense that stronger, fit people are more capable of pushing their limits of strength.

Personally, I agree with this (at least for this scenario).

[KorvinStarmast]

I disagree, I do find that ridiculous. A cantrip and a level 1 class feature should not overcome any and all physical limitations. I don't find it ridiculous at all. With a double dose of magic (some bardic and some clerical) a commoner can do some amazing things that they can't otherwise do. It's Magic!

[RSP]

I don't find it ridiculous at all. With a double dose of magic (some bardic and some clerical) a commoner can do some amazing things that they can't otherwise do. It's Magic!

RAW, Bardic Inspiration isn’t magic though. Certainly you can make it magic at your table, if that’s your preference:

“You can inspire others through stirring words or music.”

(Non-magical) Bardic Inspiration, in my view, cannot increase someone’s effective Str. It can help them push their boundaries though.

The flip side is, BI works in an AMF.

[KorvinStarmast]

RAW, Bardic Inspiration isn’t magic though. Certainly you can make it magic at your table, if that’s your preference: It has a more powerful effect than a magic spell, (guidance) it has limited uses, and it isn't magic? Your argument fails at the implementation level.
As but one example: with bardic inspiration, which adds to a saving throw, a commoner can save versus an ancient red dragon's fear effect (DC 21, Wis). Using a 1d6, all the commoner has to roll is 15 or higher (depends on what the d6 roll was) and all other commoners fail.
With bless operating from a cleric, that same commoner still hast roll 17 or higher (depending on what the roll was by the d4)
Hell, BI is in this case better than magic. :smallyuk:
Granted, Calm Emotions, the spell, allows for an auto save to that same Dragon's fear effect.

(Non-magical) Bardic Inspiration, in my view, cannot increase someone’s effective Str.
You can rule it that way at your table but that's not RAW.

The flip side is, BI works in an AMF. Indeed.

[RSP]

It has a more powerful effect than a magic spell, (guidance) it has limited uses, and it isn't magic? Your argument fails at the implementation level.

I don’t agree with the premise of “more effective mechanical effect than any spell=it must be magic”.

Certainly you can implement BI being Magic at your table, but it does not pass the SAC “is it magic” test. Nor does it RAW say it’s magic.

[KorvinStarmast]

I don’t agree with the premise of “more effective mechanical effect than any spell=it must be magic”.

Certainly you can implement BI being Magic at your table, but it does not pass the SAC “is it magic” test. Nor does it RAW say it’s magic. Which is utterly irrelevant at the implementation level, since adding a d4 and a d6, in the example given (guidance and BI for a commoner's attempt to do something) offers a commoner a huge boost to what they can do, once.
My calling it "magic" (and it's not what a guy at the gym can do) has apparently offended your semantic sense. Your pedantry is both noted and rejected. Your objection is based on a distinction without a difference.

[RSP]

Which is utterly irrelevant at the implementation level, since adding a d4 and a d6, in the example given (guidance and BI for a commoner's attempt to do something) offers a commoner a huge boost to what they can do, once.

Adding a d4 and a d6 can swing a die roll from failure to passing, whether a Save, skill check or attack.

Making claims that the above being true necessitates how others need to run their table isn’t a thing.

If you want BI to be “magic”, go ahead and play it as magic at your table. You telling us we can’t have it be what the book describes it as (“stirring words or music”), doesn’t hold any value to me.

[Psyren]

I don't find it ridiculous at all. With a double dose of magic (some bardic and some clerical) a commoner can do some amazing things that they can't otherwise do. It's Magic!

1) As mentioned previously, a level 1 human bard can do this solo, no cleric needed. In a single round even!

2) I'm all for "magic lets you do amazing things" - but even Magic has limits, or at least it should. Refusing to consider or account for those limits is a major cause of the caster supremacy that gets constantly trumpeted around here. A commoner should need more to hit any DC 30 check in the game than a level 1 bard and some luck or time.

[Segev]

For which? Barb lifting within their max lift is no check, for me.
Barb trying to increase their max say to within 700 lbs (roughly 15% increase) would be DC 15, 800 lbs (roughly within ~30% increase) DC 20.

Over ~30% of max is getting close to impossible.

Likewise for Str 8 Wiz: 15% increase would be DC 15. ~30% increase DC 19 (so need a nat 20). More than that, not possible.

And yes, this makes it harder, less likely to pass a check the weaker you are. However, as I said previously, that makes sense to me as the stronger, fitter someone is, the more likely they can do stuff in excess of their norm.

For instance, if I’m regularly lifting and my body is in shape, I can do more that I’m not regularly doing without injury. If I’m not, and I try to go outside what I’m regularly trying, there’s a good chance I’ll not succeed, and get injured in the process.

I’m not saying 5e needs to be reflective of real life, I’m just saying it makes sense that stronger, fit people are more capable of pushing their limits of strength.

I think I could get behind this take on it: the roll isn't to "lift this heavy thing," but rather to "see how much you can exceed your normal limits." Just like the jump spell trebles your jumping distance, making the 8 strength wizard leap a prodigious 24 feet, and the 20 strength barbarian leap an amazing 60 feet.

[PhoenixPhyre]

I think I could get behind this take on it: the roll isn't to "lift this heavy thing," but rather to "see how much you can exceed your normal limits." Just like the jump spell trebles your jumping distance, making the 8 strength wizard leap a prodigious 24 feet, and the 20 strength barbarian leap an amazing 60 feet.

I agree. And I'd be totally fine with a note that, say, you can generally make a Strength (Athletics) check to exceed your normal limits on Strength things, with some set of default DCs expressed in proportions: DC x to exceed by a factor of Y. With it set so that 3x <= DC 30 (so you can do the jump spell just with a check). And then you could just change the spell to say "you get +x to your Strength (athletics) check made to exceed your normal jump distance."

[Osuniev]

You know, I've been toying for a while with the idea of replacing the d20 by a Hit Dice in every skill check, and divide every DC by two.
Dividing by two means the average roll has the about the same effect for a d8/d10 dice based class than RAW.

It means, however, that failure of an easy check (was DC10 before, DC5 now) never happens if this is your specialty and you have training from lvl 1. (+3 from attribute, +2 from training)
And it never happens in everything you're trained in by lvl 9.
By level 13, the MINIMUM you can do on a check was your MAXIMUM of lvl 1. So you FEEL yourself becoming better at the things you trained into.

It means the lvl 1 Fighter or Ranger has a 50% chance of succeeding a VERY HARD check (was DC20 before, DC10 now) if it's his specialty. Instead of 25 % now. But the peasant or the Wizard cannot.

It means the Wizard has to use spells for everything DIFFICULT he is not trained in.
It means the Barbarian has a decent chance of doing everything is not trained in, but not reliably, neatly fixing his lack of out of combat options. He can also be INCREDIBLE in the things he's good at from time to time.


It means the Rogue and Bard Expertise is HUGE from lvl 1, providing a lot of reliability, and INCREDIBLE feats at high level. You could reliably feel the difference in play when you know your minimum roll is 15, when that used to be your maximum.

Getting advantage or disadvantage out of combat is still important but not as much. So you can still function with one Exhaustion level, but only at the things your character is supposed to be good at.



To prevent multiclassing from breaking everything, you could just say you roll the Hit Dice of which you have the most. (So you can multiclass into Barbarian but that won't give you their powerful dice unless you're MOSTLY a Barbarian).

You could also model elegant little options for re-trying checks but then you'd SPEND the Hit Dice... And model Fatigue by you having a ZERO on the roll if you're out of Hit Dice, but you can still succeed at the easy things you're trained in.

[OldTrees1]

You know, I've been toying for a while with the idea of replacing the d20 by a Hit Dice in every skill check, and divide every DC by two.
Dividing by two means the average roll has the about the same effect for a d8/d10 dice based class than RAW.


Nice. Reducing the die size has the obvious benefits (I mentioned using 1d8 earlier), however tying it to the HD* is not much more complicated and has nice effects on most classes.

I think Monk is a bit left out, maybe bump them to 1d10 Hit Dice?



* Always I am considering how it works with multiclassing. I would use the more frequent Hit Die if you have multiple. A Barbarian 3 / Rogue 2 would use a 1d12 and a Wizard 4 / Fighter 1 would use a 1d6.

[Yakk]

Now, lifting almost 300% of your max lift isn’t “very difficult” it’s impossible.
So, 5e skill checks is DC is is "nearly impossible".

We have situations where people have lifted nearly impossible weights in extreme circumstances, like lifting a car off someone enough for them to get free.

...

So how about this.

Take your max lift. To lift more in a short burst, take your athletics check, times 10%.

So a 22 athletics is 220% (2.2x) of your max lift.

This is smooth and doesn't have steps, so you only need to know roughly how heavy something is.

It leads to low end superhuman abilities at the high end of D&D bonuses.

[RSP]

So, 5e skill checks is DC is is "nearly impossible".

I don’t know what this means.



So how about this.

Take your max lift. To lift more in a short burst, take your athletics check, times 10%.

So a 22 athletics is 220% (2.2x) of your max lift.

This is smooth and doesn't have steps, so you only need to know roughly how heavy something is.

It leads to low end superhuman abilities at the high end of D&D bonuses.

Not for me. And again, RAW, Athletics doesn’t do anything for lifting.

But if it’s how you want your table to work, give it a shot.

[stoutstien]

Not for me. And again, RAW, Athletics doesn’t do anything for lifting.

But if it’s how you want your table to work, give it a shot.

RaW athletics can be added to strength checks if the DM says it does.

Every table is going have a different point where that threshold is. Some make the shift when lifting evolves from an instant event to an ongoing situation (lifting a gate for a round- holding it long enough for everyone in the jail to escape).

[RSP]

RaW athletics can be added to strength checks if the DM says it does.

Sure, but the same goes for proficiency in Cartography Tools.



Every table is going have a different point where that threshold is. Some make the shift when lifting evolves from an instant event to an ongoing situation (lifting a gate for a round- holding it long enough for everyone in the jail to escape).

The reason, I believe, the skill system is done as it is; to allow different styles at different tables.

[Yakk]

I missed a 30. DC 30 is "nearly impossible".

Lifting more than your standard max lift (whie moving at 1/2 speed) is a difficult thing. D&D advises you to use ability checks for this.

When you do an ability check, using an appropriate skill makes sense. And athletics does cover weightlifting form -- how to lift something heavy and impractical.

Cartographer tools doesn't.

Saying "by raw, cartographer tools has as much to do with lifting as athletics" is an interesting position. Calling it RAW is an interesting reading of the rules; it is treating "Athletics" as a meaningless set of letters whose only mechanical purpose is exactly what is listed in the rulebooks?

[Kane0]

Man i really should have saved all those links to threads with long chats about DCs and sample tables. My search-fu has failed me

Edit: Oh wait here are some
https://forums.giantitp.com/showthread.php?600950
https://forums.giantitp.com/showthread.php?495062
https://forums.giantitp.com/showthread.php?481855
https://forums.giantitp.com/showthread.php?569662

[RSP]

I missed a 30. DC 30 is "nearly impossible".

Lifting more than your standard max lift (whie moving at 1/2 speed) is a difficult thing. D&D advises you to use ability checks for this.

When you do an ability check, using an appropriate skill makes sense. And athletics does cover weightlifting form -- how to lift something heavy and impractical.

Cartographer tools doesn't.

Saying "by raw, cartographer tools has as much to do with lifting as athletics" is an interesting position. Calling it RAW is an interesting reading of the rules; it is treating "Athletics" as a meaningless set of letters whose only mechanical purpose is exactly what is listed in the rulebooks?

Athletics, RAW, has nothing to do with lifting (unless you’re somehow lifting something while jumping, swimming or climbing). From the PHB:

“Athletics. Your Strength (Athletics) check covers difficult situations you encounter while climbing, jumping, or swimming.”

A DM can absolutely houserule Athletics as dealing with lifting. But it absolutely is a houserule, as RAW, it only deals with climbing, swimming and jumping.

I mention it because adding the potential for +12 to Str checks (possible with Expertise in Athletics) is much different than just using the Str mod, which is what the RAW expects.

[PhantomSoul]

Athletics, RAW, has nothing to do with lifting (unless you’re somehow lifting something while jumping, swimming or climbing). From the PHB:

“Athletics. Your Strength (Athletics) check covers difficult situations you encounter while climbing, jumping, or swimming.”

A DM can absolutely houserule Athletics as dealing with lifting. But it absolutely is a houserule, as RAW, it only deals with climbing, swimming and jumping.

I mention it because adding the potential for +12 to Str checks (possible with Expertise in Athletics) is much different than just using the Str mod, which is what the RAW expects.

Well, slightly more -- there are other spots where Athletics is called out (e.g. for Grappling [PHB], breaking through or out of restraints [PHB, XGtE], beating an enemy grapple [PHB], disarming some traps [XGtE], pit fighting [XGtE], money earned from work [XGtE], opening some doors [XGtE], and more notably Shoving [PHB; pushing?], escaping from under the rubble of the Earthquake Spell [PHB; seems quite like lifting/pushing]) despite that not being in their rather limited pseudo-definition. With how scattered and ad hoc information is in the 5e books, pointing at a single just giving example circumstances doesn't really encompass the range of options, even before dealing with the DM having pretty wide control...

...but the book also supports the possibility:

"A Strength check can model any attempt to lift, push, pull, or break something, to force your body through a space, or to otherwise apply brute force to a situation." (PHB 175, emphasis mine)

Adding in that Skills are tacked on when deemed useful, it even seems quite reasonable that it would be deemed a relevant context (and things like the Earthquake Spell seem to suggest that lifting/pushing would be the perfect context to apply Athletics).

But it's 5e, so of course it's really just fodder for more discussions anyway xD

[stoutstien]

Well, slightly more -- there are other spots where Athletics is called out (e.g. for Grappling [PHB], breaking through or out of restraints [PHB, XGtE], beating an enemy grapple [PHB], disarming some traps [XGtE], pit fighting [XGtE], money earned from work [XGtE], opening some doors [XGtE], and more notably Shoving [PHB; pushing?], escaping from under the rubble of the Earthquake Spell [PHB; seems quite like lifting/pushing]) despite that not being in their rather limited pseudo-definition. With how scattered and ad hoc information is in the 5e books, pointing at a single just giving example circumstances doesn't really encompass the range of options, even before dealing with the DM having pretty wide control...

...but the book also supports the possibility:

"A Strength check can model any attempt to lift, push, pull, or break something, to force your body through a space, or to otherwise apply brute force to a situation." (PHB 175, emphasis mine)

Adding in that Skills are tacked on when deemed useful, it even seems quite reasonable that it would be deemed a relevant context (and things like the Earthquake Spell seem to suggest that lifting/pushing would be the perfect context to apply Athletics).

But it's 5e, so of course it's really just fodder for more discussions anyway xD

There's also the fact that just about every single strength check in published campaigns include (Athletics) in the DC.

[Keltest]

As far as it goes, there's already a rule for lifting something directly, with or without carrying it somewhere. A DM can certainly allow a check to push beyond those limits, but there's nothing to suggest that doing so is an expected or regular part of the system.

[Psyren]

Athletics, RAW, has nothing to do with lifting (unless you’re somehow lifting something while jumping, swimming or climbing). From the PHB:

“Athletics. Your Strength (Athletics) check covers difficult situations you encounter while climbing, jumping, or swimming.”

A DM can absolutely houserule Athletics as dealing with lifting. But it absolutely is a houserule, as RAW, it only deals with climbing, swimming and jumping.

I mention it because adding the potential for +12 to Str checks (possible with Expertise in Athletics) is much different than just using the Str mod, which is what the RAW expects.

DMG 239: "Often, players ask whether they can apply a skill proficiency to an ability check. If a player can provide a good justification for why a character's training and aptitude in a skill should apply to the check, go ahead and allow it, rewarding the player's creative thinking."

If they can explain how Athletics can apply to lifting something and the DM agrees, then it applies - that's black-letter RAW, not a "houserule."

[Keltest]

DMG 239: "Often, players ask whether they can apply a skill proficiency to an ability check. If a player can provide a good justification for why a character's training and aptitude in a skill should apply to the check, go ahead and allow it, rewarding the player's creative thinking."

If they can explain how Athletics can apply to lifting something and the DM agrees, then it applies - that's black-letter RAW, not a "houserule."

Thats if its already an ability check.