Bounded Accuracy And How It Applies To Certain Skill Checks

[Schadenfreuda]

Bounded Accuracy is easily one of the best features of Fifth Edition. Doing away with the enormous bonuses that a character could accrue in 3.5e and Pathfinder 1e is a nearly unambiguous improvement as far as fairness and ease of play were concerned. That said, there are some places where it would do with nuance that it presently lacks.

There have been many threads on these forums concerning how to deal with small bonuses and the randomness that comes with them, and how 5e's skill system doesn't have the right setup for simulating the way actual skills work, so I won't go into a discussion of how skills will work when talking about people and training.

Instead, I wish to discuss animals, and how their given stats portray their abilities. Not monsters, mind you, just normal real-world animals that have stats assigned to them. In particular I want to talk about animal sight, smell, and hearing.

A bear for example has a powerful nose, one that can detect scents as much several million times fainter than a human's nose can detect. That translates to.... advantage on Wisdom (Perception) checks with a +3 flat bonus. That translates to about a +8 bonus, but one that still doesn't let it exceed 23 on a natural 20.

Eagles can see minute movements in fields miles away from them, enough that a human with similar acuity would be able to recognise faces a couple of miles away. That translates to.... advantage on Wisdom (Perception) checks that rely on sight. With a +4 bonus.

Elephants have +0 to their hearing, and no advantage. That's exactly the same as a human with 10 wisdom and no proficiency. They don't have their real life Tremorsense either.

A first-level PC with +2 Proficiency and 14 Wisdom (+2) can match the upper range of these animals' senses, even if not as reliably as a creature with advantage could. Perhaps it's not so good that a 1st-level familiar or 2nd-level druid in wildshape should have superpowers of high perception that far outclass anything a human could do at any level, but still, a bear should have a better nose than a level 1 human with advantage.

Basically, there are many places where Bounded Accuracy is not reasonable to apply, because it's based on how humans perform at certain tasks and humans are not representative of the average animal in most respects.

[Starbuck_II]

Jump is the big outlier I think.
Remember, you move but your Str in feet in a jump if I recall.
Meaning squirrels don't jump well, but Elephants do.

[djreynolds]

You're correct, a growling bear is intimidating, I mean animals intimidate each other all the time and steal kills... it should be a skill for animals

I've always had a dilemma with perception and investigation and what is working and what isn't.

I mean IRL you might hear a whistle above and you're looking and others are running for cover... I guess this is translated into experience/training into proficiency.

And wisdom is an odd ability, because in a sense it has to trigger your dexterity to move, to react... like a trap forces a dexterity save but your wisdom ability noticed the trap.

I've always felt there is disconnection between abilities, like they should influence other abilities.

[HappyDaze]

I've always felt there is disconnection between abilities, like they should influence other abilities.

Play The Dark Eye. Skill checks are made against three attributes.

[djreynolds]

Play The Dark Eye. Skill checks are made against three attributes.

Sounds cool.

Does anyone else feel like there is a physical wisdom related to dex and then almost a cognitive aspect related to intelligence.

Same with con and strength, there is a relation

[ad_hoc]

An 18 Passive Perception is very high. That means an average human only has a 10% chance to sneak by a bear unnoticed. I don't like those odds.

And yes, the ability check system does represent skills well. Checks are only meant to actually happen in tense and exciting moments.

There should be no rolls for run of the mill events. If it isn't exciting don't waste time on it, just move on.

Remember that in 5e an ability check is called for by the DM, not the player. The DM determines if a check succeeds, fails, or is in doubt (and interesting enough to spend time on).

[Anymage]

Specific to animals, it'd be trivial to say that they have their specific sensory ability without having to tie checks into it. Like, a human cannot see in the dark, but assuming well-lit conditions a human and a dwarf with identical perception modifiers can see the same sorts of things. A dog may be able to apply their nose or an eagle may be able to apply their distance vision to do things that a human can't, but that only applies when declaring that a check is impossible or unnecessary, and apples-to-apples comparisons can use the normal rolls.

More generally, the d20 resolution mechanic is clearly unrealistic, with a totally flat probability "curve" and only 5% increments of granularity. A properly realistic resolution mechanic would become a massive headache to adjudicate at the table, though. BA was designed with playability in mind over an attempt at hard simulationism, and I'm okay with the tradeoffs made towards that end.

[Makorel]

I mean is this an issue with the skill system or an issue with WoTC's lack of care in making mundane animals? All you would need to do to fix this sort of stuff is tack on some special abilities that fulfill these attributes. Things like Bear Nose - You cannot be hidden from the bear while you are within 300 feet of it and it can smell you. Jumpless - The Elephant is incapable of jumping. The eagle example is particularly egregious just because Eagle Barbarians get actual eagle sight while the eagle in the monster manual doesn't.

[Tanarii]

You're assuming two things:
1) that the human can make impossible checks that would be possible for an animal
2) that the DC is the same for all tasks, assuming it's even possible for both

The second one is pretty normal assumption of an unspoken rule based on the simplification of the system. But it doesn't have to be the case.

But the former is definitely not. The first step, per the DMG, for a DM determining if a check should happen is to decide if it's an automatic success or failure.

There are many situations in which an animal would be able to smell something, but a human wouldn't be able to. Similarly for many animals where they might be able to smell something but couldn't detect it by sight.

To use a different check, there are many animals and situations where a human could automatically succeed in a climb without a check, and the animal would either need to make a check, or just automatically fail. The same could be true for a human vs centaur consideration of climbing something.

[Telok]

There are three times to roll skills. In combat, as a single opposed check, and when failure is more fun than success.

There are three times not to roll skills. When its going to be multiple rolls (one person rolling more than once or multiple people rolling), when you can't think of at least two fun and exciting things to happen on failure, and when trying to compare the ability of two creatures when the outcome isn't essentially random.

5e doesn't do "can the archmage decipher the magic book" or "does the archmage know more about this 2nd level spell than a man-eating ape". It can't do that with any believability because of the d20 roll. And its not meant to do that. You're supposed to use the skills for deciding if the archmage can guess the correct rune to turn off the trap while up to the waist in caustic acid with ghouls chewing on everyone and skeletal archers shooting arrows.

[Luccan]

D&D skills have never been great at mapping the real-world abilities of animals, probably because the designers aren't zoologists. It's completely serviceable, but far from perfect. In 5e's case, though, cutting down the numbers means that average people can hit what the game defines as Hard checks, even if it's only a 5% chance. In this case, based on the game definitions at least, the bear is fine. With a passive 18 scent based Perception, it doesn't even have to try to pick up the majority of smells. The trouble is that, by RAW, Joe can very rarely beat that with active effort (though the bear will do so more often). 3e had the actually fairly elegant solution of the Scent feature, which they probably should have kept.

One solution for 5e is to decide PC races don't have sensitive enough noses to make scent-based perception checks unless within x distance of the offending odor. If you're a PC (or something similar) you can't track by scent unless you take on an appropriate form.

[Quietus]

One solution for 5e is to decide PC races don't have sensitive enough noses to make scent-based perception checks unless within x distance of the offending odor. If you're a PC (or something similar) you can't track by scent unless you take on an appropriate form.

This is exactly the solution. Joe can't beat the bear at rooting out the smell of his quarry, because he doesn't have the appropriate sensory apparatus, and therefore doesn't even get to make that roll. Now, he might be able to follow someone based on other observable facts, intelligence-based information gathering that the bear wouldn't be able to manage. But that's a different beast altogether.

[JackPhoenix]

There are three times not to roll skills. When its going to be multiple rolls (one person rolling more than once or multiple people rolling), when you can't think of at least two fun and exciting things to happen on failure, and when trying to compare the ability of two creatures when the outcome isn't essentially random.

You're missing "When there's no chance of failure" and "when there's no chance of success".

[Sorinth]

This is one of the reasons the #1 rule is the DM decides. Honestly I don't think filling the DMG/MM with tons of unique special features is the way to go. It adds bloat for almost no gain.

It's much better to just leave it up to the DM, and the DM can and should set different DCs and determine auto-success/failure differently between PCs and animals/monsters. So yeah the PC cleric might have a higher perception mod then an eagle, but if both are trying to spot a mouse a mile away the DC isn't going to be the same. Remember the DC is based on how difficult a task is for that one entity. So for a human, spotting that mouse is impossible and is an auto-fail, for an eagle maybe it's easy and so only has a DC of 10.

[djreynolds]

Perhaps with skill checks and bounded accuracy you cannot have too many static bonuses.

But obviously scent is powerful, but animals hunt each other all the time.

Because we cannot account for upwind and downwind (I mean we could), simply saying you have advantage with perception checks if you have the scent ability is enough of a bonus

[JackPhoenix]

It's much better to just leave it up to the DM, and the DM can and should set different DCs and determine auto-success/failure differently between PCs and animals/monsters. So yeah the PC cleric might have a higher perception mod then an eagle, but if both are trying to spot a mouse a mile away the DC isn't going to be the same. Remember the DC is based on how difficult a task is for that one entity. So for a human, spotting that mouse is impossible and is an auto-fail, for an eagle maybe it's easy and so only has a DC of 10.

That's only sort-of true. Yes, if the task isn't possible for you, you don't get to roll at all, but if the task *IS* possible, the DC is the same for everyone who can attempt it.

[Sorinth]

That's only sort-of true. Yes, if the task isn't possible for you, you don't get to roll at all, but if the task *IS* possible, the DC is the same for everyone who can attempt it.

Disagree, whether somethings is easy, moderate, hard, etc... can be different and therefore can create different DCs for whoever attempts it. Where does it say in the rules that the DC is the same for everyone that attempting a task?

For example if instead of a mouse a mile away it was only 200ft away, now the human has a chance to spot it but's it's nonsensical to use the same DC for the eagle and the human. In fact how does the DM even determine the DC, it's a difficult task for a human, so DC 20, but very easy for the eagle DC 5, so does the eagle now have to beat a 20, even though if the mouse was even further away the DC would be easier because a human would auto-fail? That's nonsensical.

[Sorinth]

The only real problem with the skill check system is that the d20 provides so much variance compared to training/ability.

For example a Str 18 player should beat the Str 10 character in an arm wrestling contest 99 times out of 100. But with the current system the strength 10 character will still win a decent amount of the time.

[Man_Over_Game]

I mean is this an issue with the skill system or an issue with WoTC's lack of care in making mundane animals?

No it's definitely an issue with the skill system. From level 1-20, you're expected to increase your chances of success on a skill by roughly 6 points.

So a level 1 might have a 50% chance of landing a 15 DC. At level 20, that's a 80% chance of success. A commoner would have about a 25% chance.

Or, in other words, you'll fail about 1/5 times, while four commoners doing the same would result in one of them succeeding.

From your grand 1-20 adventures, you have grown from being 4x as proficient as a Commoner, to being 8x times as proficient as a Commoner. Or just "twice as good" as you were at level 1.

There's a lot of reasons to scrap the skill system and rebuild it from the ground up, and "animals aren't realistic" is just a tiny byproduct of some much larger problems.

[Sorinth]

No it's definitely an issue with the skill system. From level 1-20, you're expected to increase your chances of success on a skill by roughly 6 points.

So a level 1 might have a 50% chance of landing a 15 DC. At level 20, that's a 80% chance of success. A commoner would have about a 25% chance.

Or, in other words, you'll fail about 1/5 times, while four commoners doing the same would result in one of them succeeding.

From your grand 1-20 adventures, you have grown from being 4x as proficient as a Commoner, to being 8x times as proficient as a Commoner. Or just "twice as good" as you were at level 1.

There's a lot of reasons to scrap the skill system and rebuild it from the ground up, and "animals aren't realistic" is just a tiny byproduct of some much larger problems.

For most skills checks that's not necessarily wrong. For example a perception based check it would make sense that a group of 4 people on lookout would have as good if not a better chance of spotting a someone stealthily approaching then one level 20 player because it's not like your vision gets better when going from level 1-20. Experience has taught that level 20 guy to pay attention to the subtle signs that indicate someone is there and so it's easier to find them. But a group of people should still be better.

[Man_Over_Game]

For most skills checks that's not necessarily wrong. For example a perception based check it would make sense that a group of 4 people on lookout would have as good if not a better chance of spotting a someone stealthily approaching then one level 20 player because it's not like your vision gets better when going from level 1-20. Experience has taught that level 20 guy to pay attention to the subtle signs that indicate someone is there and so it's easier to find them. But a group of people should still be better.

That'd be fine if we're just comparing actions of common people vs. adventurers doing common things.

But we're talking about breaking out of manacles, throwing boulders, hiding from dragons, lying to mindflayers.

Put another way, without any additional investment, a Wizard will learn to stop time, summon enough meteors to destroy a city, and become slightly more effective at reading books (history).

[Sorinth]

That'd be fine if we're just comparing actions of common people vs. adventurers doing common things.

But we're talking about breaking out of manacles, throwing boulders, hiding from dragons, lying to mindflayers.

Put another way, without any additional investment, a Wizard will learn to stop time, summon enough meteors to destroy a city, and become slightly more effective at reading books (history).

To me it makes sense that a guy who is manacled can't simply bend/break metal with ease, he's still human so yeah he might bust through but it shouldn't be easy/routine. The wizard might channel powerful magics, but no I don't think he should pass every history check and basically know everything that was ever written.

The system is far from perfect, but I only really have issue with opposed roles where the luck of the d20 has way too much influence. Beyond that so long as the DM enforces common sense rules regarding retries and what failure means then it works pretty well. A key part of the fun is that things don't get bogged down, so there's a lot of value in keeping things simple and just having the DM arbitrate the edge cases.

[Chronos]

Another example of the problem, that doesn't even depend on die rolls: A literal god of strength, with a 30 Str and the equivalent of an Enhance Ability spell on him all the time, can lift less weight than a real-world gorilla.

Or to go back to rolls: In the real world, there are some tasks that are hard enough that some people can almost never succeed at them, but which other people can succeed at almost all the time. 5e can just barely model that, but it takes a 17th-level character with Expertise to do it. But the real world has multiple layers of that effect: The guy who can succeed at that first task 95% of the time has other tasks that he can only succeed at 5% of the time, but which a greater expert can do easily and routinely.


Quoth Sorinth:

Disagree, whether somethings is easy, moderate, hard, etc... can be different and therefore can create different DCs for whoever attempts it. Where does it say in the rules that the DC is the same for everyone that attempting a task?
This basically amounts to "give out a whole bunch of stacking bonuses", like in 3rd edition, except without saying what those bonuses are. Someone saying "Well, the DC should be 5 points lower than me, because I'm really good at it" is saying the same thing as "Being really good at that should mean an additional +5 to the check".

[Pex]

The only real problem with the skill check system is that the d20 provides so much variance compared to training/ability.

For example a Str 18 player should beat the Str 10 character in an arm wrestling contest 99 times out of 100. But with the current system the strength 10 character will still win a decent amount of the time.

Someone might say you shouldn't have had a roll. Not every ability (skill) check needs a roll. It's an autosuccess/fail. The ST 18 would always defeat the ST 10 character so there's no problem. Trouble is there is a problem. Some DMs cannot accept there's never a chance of failure for anything. There always must be a roll otherwise there's no challenge. Other DMs accept there are trivial things that don't require a roll, but they disagree on what makes something trivial. What one DM says no roll is needed another DM says a roll is required. Neither is wrong. They just disagree. Add in a third DM who agrees there must be a roll but disagrees with the second DM on what the DC is. He's not wrong either. He just disagrees. It's not as bad if it's an opposed roll such as this arm wrestling contest. Then it's only a disagreement on whether a roll is required or not. Still, that lack of consistency is the problem. For some people that lack of consistency is the greatest thing since sliced bread.

[Sorinth]

Another example of the problem, that doesn't even depend on die rolls: A literal god of strength, with a 30 Str and the equivalent of an Enhance Ability spell on him all the time, can lift less weight than a real-world gorilla.

Or to go back to rolls: In the real world, there are some tasks that are hard enough that some people can almost never succeed at them, but which other people can succeed at almost all the time. 5e can just barely model that, but it takes a 17th-level character with Expertise to do it. But the real world has multiple layers of that effect: The guy who can succeed at that first task 95% of the time has other tasks that he can only succeed at 5% of the time, but which a greater expert can do easily and routinely.


This basically amounts to "give out a whole bunch of stacking bonuses", like in 3rd edition, except without saying what those bonuses are. Someone saying "Well, the DC should be 5 points lower than me, because I'm really good at it" is saying the same thing as "Being really good at that should mean an additional +5 to the check".

I do see your point and to a certain extent you are correct, but I probably wasn't clear but the DC is set based on the generic version of that entity. So the DC for a Human Rogue with expertise in the skill is still determined by how hard it would be for a Human commoner to do that task. The DC for that human commoner is probably the same for a dwarf commoner which is why most checks would have the same DC even if one PC is much better at it then the other. However when we are talking about something completetly different making the check like an Eagle or a Bear or a Dragon, then yeah the DC might be different because of inherent qualities of that creature.

[Segev]

While it won't help with things like elephants having no special abilities related to senses, one thing you can do is simply allow for anything with "Keen [sense]" to automatically detect things that you'd either make others roll for or wouldn't let them detect at all, if it's based on that sense and it's hard to detect without that sense being special.

5e has stripped most of the myriad bonuses, penalties, and special rules associated with size, preferring by and large to make them specific enhancements to specific creatures where applicable. However, you could easily apply bonuses or penalties to jump distances based on size category.

Though if squirrels were statted up, I imagine they'd get a special feature the way grungs do: Tree-strider: The squirrel has a running long-jump of 15 feet, and needs only 5 feet of running start to get it. 10 feet with no running start.

[Man_Over_Game]

To me it makes sense that a guy who is manacled can't simply bend/break metal with ease, he's still human so yeah he might bust through but it shouldn't be easy/routine. The wizard might channel powerful magics, but no I don't think he should pass every history check and basically know everything that was ever written.

I'm not saying that he should. My issue isn't that "Level 20's aren't gods with skills", but that "Level 20's aren't much better than a level 1".

As a level 20, you have massive amounts of HP, can cast spells that destroy entire cities, can be a whirlwind of blades as you cut down almost 10 people in an instant. And in all of that time, you're barely better at running, or jumping, or throwing, or reading, or identifying monsters, or seeing hidden enemies, etc.

There is no scaling between level 1-20 when it comes to skills. Over 20 levels, you get basically 15x your HP, deal about 10x your level 1 damage, and got a flat 30% higher chance to succeed on anything that isn't related to killing stuff. See the disparity?

[Sorinth]

I'm not saying that he should. My issue isn't that "Level 20's aren't gods with skills", but that "Level 20's aren't much better than a level 1".

As a level 20, you have massive amounts of HP, can cast spells that destroy entire cities, can be a whirlwind of blades as you cut down almost 10 people in an instant. And in all of that time, you're barely better at running, or jumping, or throwing, or reading, or identifying monsters, or seeing hidden enemies, etc.

There is no scaling between level 1-20 when it comes to skills. Over 20 levels, you get basically 15x your HP, deal about 10x your level 1 damage, and got a flat 30% higher chance to succeed on anything that isn't related to killing stuff. See the disparity?

How is the level 20 supposed to be much better at skills then a level 1 guy if at the end of the day they should both struggle trying to escape a pair of manacles?

And if we ignore magic, bounded accuracy means a 20th level fighter can and will get taken down by a large group of Kobolds. So the disparity is maybe not as bad as it first appears. And it's also worth noting that you went from level 1 to level 20 by spending a few years going around killing things, so it seems normal that you are now much better at killing things then you are at say crafting armor which you did on and off during those years.

[Man_Over_Game]

How is the level 20 supposed to be much better at skills then a level 1 guy if at the end of the day they should both struggle trying to escape a pair of manacles?

And if we ignore magic, bounded accuracy means a 20th level fighter can and will get taken down by a large group of Kobolds. So the disparity is maybe not as bad as it first appears. And it's also worth noting that you went from level 1 to level 20 by spending a few years going around killing things, so it seems normal that you are now much better at killing things then you are at say crafting armor which you did on and off during those years.

But you're justifying realism. In a game of Dungeons, Dragons, and Wizards. People do not play DnD to play a commoner.

More importantly, skills are something that carries weight in your classes. You pick the Rogue because he deals damage, is mobile, and is good with skills. You pick Half-Elves as a race because they carry skills.

If those choices hold no weight, then every instance that provides a skill is something that you should be reimbursed in. If nobody bothers investing into skills, and Rogues invest into skills, give Rogues something else.

Having skills be mundane, so that magicians have a reference point to see how fantastical they are, is basically saying that some characters are designed to be sidekicks while others are allowed to do heroics. More competition means more diversity. For once, I'd like to see a Rogue outshine a Wizard in something that isn't spending a Bonus Action to Hide.

[Sorinth]

But you're justifying realism. In a game of Dungeons, Dragons, and Wizards. People do not play DnD to play a commoner.

This comes down to what type of game you want to play. Plenty of people play and enjoy low-magic settings for example. Personally I do prefer a degree of realism, where the PCs aren't just Mary Sue's.


More importantly, skills are something that carries weight in your classes. You pick the Rogue because he deals damage, is mobile, and is good with skills. You pick Half-Elves as a race because they carry skills.

If those choices hold no weight, then every instance that provides a skill is something that you should be reimbursed in. If nobody bothers investing into skills, and Rogues invest into skills, give Rogues something else.

Having skills be mundane, so that magicians have a reference point to see how fantastical they are, is basically saying that some characters are designed to be sidekicks while others are allowed to do heroics. More competition means more diversity. For once, I'd like to see a Rogue outshine a Wizard in something that isn't spending a Bonus Action to Hide.

Having between a +2 and a +12 does carry weight, it's just doesn't make you a god at skills. As for higher tier magic making the other classes sidekicks, that has nothing to do with bounded accuracy or skills, just look at 3/3.5 edition.

[47Ace]

This comes down to what type of game you want to play. Plenty of people play and enjoy low-magic settings for example. Personally I do prefer a degree of realism, where the PCs aren't just Mary Sue's.


They just play at low levels don't tell people who want to play a high fantasy game where wizards can cast a wish a day and cause an earthquake with a general level of magic above what is common in myths that their barbarians and fighters cannot preform feats of strength similar to the ones done by Heracules, Samson, and various Norse hero's. There is nothing that stops you from just playing at lower levels. Sorry if that comes off a bit rude but I get frustrated reading arguments that go "I like this level of fantasy your higher level is wrong." "But look at what spell-casters can currently due I just want something similar." "But that's a problem with spell casters doing more then I want them to." When both levels of fantasy could be contained in the same rule set at different level ranges without problem if WOTC had just included some simple rules for stopping advancement at lower levels and said if if you don't want higher fantasy don't play higher levels.

[Man_Over_Game]

Having between a +2 and a +12 does carry weight, it's just doesn't make you a god at skills. As for higher tier magic making the other classes sidekicks, that has nothing to do with bounded accuracy or skills, just look at 3/3.5 edition.

Over time, DnD has been transitioning to a concept that non-magical combatants should be near the same level as magical combatants. It happened in 4e, and people seem happy with the fact that that trend continued in 5e.

The problem is, that same concept hasn't progressed in the non-combat elements to the game. People are wanting that to be a solid pillar of the game, and there's really only a few ways of making it work:

Redo skill system
Rely on magic
Make things up as you go


And the books don't really provide much help on that third topic. I tried to get examples of what a distinct DC 25-30 would look like for a skill-user, to determine what kind of "super heroics" a character should be capable of as real-world people, and it basically boiled down to "Move an Immovable Rod by 10 feet". The only examples I could find of someone using mundane talents to do something supernatural was moving a make-believe object. The closest mention of interacting with a mundane object, at a DC 20, was breaking out of manacles, which almost any level 1 character could do in about a minute.

In comparison, a Wizard could afford to cast Invisibility for an hour on multiple creatures, for multiple hours, as early as level 5. Sure, that cost a resource, but that's kinda paltry compared to about 30% of the Rogue's class features. If a Wizard doesn't use Invisibility, he uses Flaming Sphere. If a Rogue doesn't use Stealth, he doesn't use 30% of his class. Even in the instances where they both are used, the Wizard's solutions are often "Yes" ("Yes, you did Dimension Door the party to safety". "Yes, you do read his thoughts and determine where the hostage is located"), where the Rogue's are often "Maybe" ("Roll a DC 20 for me").

We know it's not perfect. We can live with imperfect. My issue is that it's not good enough.

[Segev]

To be fair, if you play a Rogue or a Bard, or any variation of at-least-half-human, you can get between 1 and 4 skills with "Expertise," which doubles proficiency bonuses.

Additionally, there is up to a +4 shift in all skills related to a single ability score due to ASIs over 20 levels. (IIRC, there are 4 ASIs in most classes.) This requires taking no feats (or taking one half-feat with a bonus in the right Ability, if the Ability starts odd), and stacking all ASIs in one stat, though.

A swing from +4 to +16 in a single Expertise skill that starts with +0 from stat and ends with +4 from stat is pretty good progression. But...that is just one skill, and you've focused REALLY heavily on it.

Finally, there are stat-boosting items, however rare they may be.


The issue here is that the complaints about this are essentially complaining that bounded accuracy isn't what they want. And that's fine. But it's like going to Panda Express and complaining that they don't serve pizza.

There are solutions to this. Most of them lie in specialized abilities, or in the rule that a lot of the time, the DM should just have you auto-succeed (or auto-fail). I've seen, in modules, clauses along the lines of, "Characters with proficiency in Thieves Tools can..." and then list a thing they just succeed at, while those without the proficiency can't do it.

This doesn't make for scaling with level, obviously, but it does mean that your skill selections can be a flat-out be a difference between success and failure at a particular approach to a particular problem.

[Man_Over_Game]

There are solutions to this. Most of them lie in specialized abilities, or in the rule that a lot of the time, the DM should just have you auto-succeed (or auto-fail). I've seen, in modules, clauses along the lines of, "Characters with proficiency in Thieves Tools can..." and then list a thing they just succeed at, while those without the proficiency can't do it.

This doesn't make for scaling with level, obviously, but it does mean that your skill selections can be a flat-out be a difference between success and failure at a particular approach to a particular problem.

So...like Cantrips? What kind of investment would a Rogue need with Tinker's Tools to maintain a similar level to Mend? How high of a Performance skill bonus would I need to do ventriloquism as well as Minor Illusion would allow me to do?

And say you do make that grand investment, so that some things are just automatically successful. What's the upper bound? What are the heroics I can do with Tinker's Tools? Xanathar's recommends a minor jury-rigged upgrade to be a DC 25, which I'd assume is something like putting on spikes to a shield....which shouldn't need to be rolled in the first place? (Since you generally only roll in stressful/chaotic situations).

Am I supposed to roll for Performance checks in most situations, or only life-threatening ones? When does my skill at that ability surpass a Cantrip or a level 1 spell? There aren't answers on this in the books or even decent suggestions by the developers. Just speculation on the internet and the advice of "well, don't do that, and you won't be disappointed".

They kept skill DCs down in order to follow the same system as attacking, while making attacks always revolve around beating AC, and then they kept AC fairly static to make sure that weaker creatures can contribute in combat. However, the situations that a skill might be relevant for do not scale in the same way as AC, which is why Wizards get Dimension Door and Seeming. They didn't account for the fact that, while Casters should be able to deal with more and more perilous situations, that they'll still be able to affordably solve the problems they had before with things like Invisibility, Disguise Self, or Knock. Things that overlap and overshadow a skill-monkey specialist.

Is that what skill monkeys inevitably are rewarded with for levels 1-20? Being able to cast level 2 spells as rituals?
Followup question, how often do you expect a level 20 to need to cast a ritual?

[47Ace]

To be fair, if you play a Rogue or a Bard, or any variation of at-least-half-human, you can get between 1 and 4 skills with "Expertise," which doubles proficiency bonuses.

Additionally, there is up to a +4 shift in all skills related to a single ability score due to ASIs over 20 levels. (IIRC, there are 4 ASIs in most classes.) This requires taking no feats (or taking one half-feat with a bonus in the right Ability, if the Ability starts odd), and stacking all ASIs in one stat, though.

A swing from +4 to +16 in a single Expertise skill that starts with +0 from stat and ends with +4 from stat is pretty good progression. But...that is just one skill, and you've focused REALLY heavily on it.

Finally, there are stat-boosting items, however rare they may be.

Yes except stats are defined by your class with no reasonable option to bump stats just for the skills. So yes stat bumps help with the Wizards Religion check and the paladin's deception check but not the clerics religions check or the sneaky Horizon Walker ranger's deception check.



The issue here is that the complaints about this are essentially complaining that bounded accuracy isn't what they want. And that's fine. But it's like going to Panda Express and complaining that they don't serve pizza.

Yes and no, one solution to the problem of allowing competence in skills and not having class extensively define skill competence is to just allow massive other bonuses but, that is not necessarily the only solution. Bringing back take 10 alone would make a huge difference outside of combat.




There are solutions to this. Most of them lie in specialized abilities, or in the rule that a lot of the time, the DM should just have you auto-succeed (or auto-fail). I've seen, in modules, clauses along the lines of, "Characters with proficiency in Thieves Tools can..." and then list a thing they just succeed at, while those without the proficiency can't do it.

This doesn't make for scaling with level, obviously, but it does mean that your skill selections can be a flat-out be a difference between success and failure at a particular approach to a particular problem.

Yes that can work. Just it would be really nice if there was more of that particularly in a form that lived on a character sheet like spells do. Some sort of skill trick like lock picking being an ability in addition to everything classes currently get that was like

Lockpicking
Requirement: Proficiency in thieves tools
Can pick all normal locks as an object interaction no check required as long as you have you thieves tools DC 0
Can pick advanced locks with a DC 10 check
Highly advances or magical reinforced (arcane lock)locks are DC 15
Improvised tools increases DC by 5
(Optional: this can be attempted without this skill trick with a +10 to all DCs)

or

Drawing
Requirement: none
Can make a drawing of a person you can see in a minuet DC 0 This drawing accurately represent the person well enough that they can be identified from the drawing.
If you saw them recently DC 5
Based off of a description DC 10
If you are having to make do with significantly sub-par tools DC +5
(Optional: this can be attempted without this skill trick with a +10 to all DCs)

[Pex]

To be fair, if you play a Rogue or a Bard, or any variation of at-least-half-human, you can get between 1 and 4 skills with "Expertise," which doubles proficiency bonuses.

Additionally, there is up to a +4 shift in all skills related to a single ability score due to ASIs over 20 levels. (IIRC, there are 4 ASIs in most classes.) This requires taking no feats (or taking one half-feat with a bonus in the right Ability, if the Ability starts odd), and stacking all ASIs in one stat, though.

A swing from +4 to +16 in a single Expertise skill that starts with +0 from stat and ends with +4 from stat is pretty good progression. But...that is just one skill, and you've focused REALLY heavily on it.

Finally, there are stat-boosting items, however rare they may be.


The issue here is that the complaints about this are essentially complaining that bounded accuracy isn't what they want. And that's fine. But it's like going to Panda Express and complaining that they don't serve pizza.

There are solutions to this. Most of them lie in specialized abilities, or in the rule that a lot of the time, the DM should just have you auto-succeed (or auto-fail). I've seen, in modules, clauses along the lines of, "Characters with proficiency in Thieves Tools can..." and then list a thing they just succeed at, while those without the proficiency can't do it.

This doesn't make for scaling with level, obviously, but it does mean that your skill selections can be a flat-out be a difference between success and failure at a particular approach to a particular problem.

What of someone who is not a rogue or bard? I can fully accept a rogue or bard to be overall better at skills in general than others to be their Thing, but that does not equate to others must never be good at anything. What of the arm wrestling contest? Should there be a roll between a ST 18 and ST 10 contestants? If no, why is someone who says yes wrong? If yes. why is someone who says no wrong? What about between ST 18 and ST 18? ST 18 and ST 16? How much disparity between the ST scores does it take to be no roll, higher ST autowins? If never, then people will complain ST 10 can defeat ST 18 or even 20.

Suppose there was a defined rule that settles the arm wrestling contest. It may not satisfy everyone. That's people. There are defined rules for other things people don't like and change. Great Weapon Master feat being infamous for one. At least there's a defined rule so there's common ground. The rules define you need to have proficiency in a Tool to use it, which is why someone with Proficiency in Thieves' Tools can . . . while those who don't can't. 5E is not lacking in rules on how things work, even if someone doesn't like what the rules are. When it comes to skill use there they chose to let everyone do whatever they want. For some people that's hip hip hooray pop the cork. For others that's gnashing teeth of frustration, ergo this thread and others of similar complaint.

The skill rules aren't matching the OP's expectation on what animals can do, though I agree as another mentioned it's loophole wonky about elephants jumping and not specifically on how 5E handles skills. It's not enough that an animal has a +# to something. There's nothing to define what that +# can do. The OP limited it to animals, but PCs have the same problem which is why the thread expanded. The DM has to make it up and not every DM will agree on the solution.

[Sorinth]

They just play at low levels don't tell people who want to play a high fantasy game where wizards can cast a wish a day and cause an earthquake with a general level of magic above what is common in myths that their barbarians and fighters cannot preform feats of strength similar to the ones done by Heracules, Samson, and various Norse hero's. There is nothing that stops you from just playing at lower levels. Sorry if that comes off a bit rude but I get frustrated reading arguments that go "I like this level of fantasy your higher level is wrong." "But look at what spell-casters can currently due I just want something similar." "But that's a problem with spell casters doing more then I want them to." When both levels of fantasy could be contained in the same rule set at different level ranges without problem if WOTC had just included some simple rules for stopping advancement at lower levels and said if if you don't want higher fantasy don't play higher levels.

First off I never said anything close to "your higher level is wrong", there is no right/wrong between low/high magic games just like there's no right/wrong between combat focused vs role play focused games.

And for the record a Barbarian with expertise in Atheletics from say a feat at level 20 has a +19 bonus to his atheletics check so it's not like they can't perform insane feats of strength.

It's also quite easy to play a low magic setting without resorting to limit level advancement. Plenty of tables ban the Wish spell because it's too disruptive, are you claiming they are wrong and should just limit level advancement so they don't get 9th level spells?

[Man_Over_Game]

First off I never said anything close to "your higher level is wrong", there is no right/wrong between low/high magic games just like there's no right/wrong between combat focused vs role play focused games.

And for the record a Barbarian with expertise in Atheletics from say a feat at level 20 has a +19 bonus to his atheletics check so it's not like they can't perform insane feats of strength.

It's also quite easy to play a low magic setting without resorting to limit level advancement. Plenty of tables ban the Wish spell because it's too disruptive, are you claiming they are wrong and should just limit level advancement so they don't get 9th level spells?

I think the issue is that any scenario with 9th level spells isn't a low-magic situation. Wish is just excessively powerful in a powerful-magic world.

Something like the Lord of the Rings knock-off of 5e is a better example of low-magic. Decent rules for mundane travel, martials have more versatility and do more out-of-combat, casters don't do much at lower levels (where even starting a fire is a big deal). And it works, because the universe is low-magic.

But half of 5e is tuned towards low magic, and it calls itself to be low-magic (like how they suggest you handle magic items), but the reality is that about 80% of your character content revolves around using magic in some way, about 50% of that revolves around spells, and the best generic feats are the ones that help you cast or resist magic. It's all over the place, constantly being thrown towards "low-magic" or "high-magic" at any given moment.

It'd take a big overhaul to make 5e a "low-magic" game, I think. Slower spell progression (progresses from level 1-6 instead of 1-9, maybe?), and slightly more spell slots than the current progression (to compensate for the power loss) would do it. Uses the old Rifts design of "Soldiers are there to kill, mages are there to do everything else". But that's just a guess I pulled out of the air.

[Segev]

So...like Cantrips? What kind of investment would a Rogue need with Tinker's Tools to maintain a similar level to Mend? How high of a Performance skill bonus would I need to do ventriloquism as well as Minor Illusion would allow me to do?

And say you do make that grand investment, so that some things are just automatically successful. What's the upper bound? What are the heroics I can do with Tinker's Tools? Xanathar's recommends a minor jury-rigged upgrade to be a DC 25, which I'd assume is something like putting on spikes to a shield....which shouldn't need to be rolled in the first place? (Since you generally only roll in stressful/chaotic situations).Given what you quoted from my post, I am not sure what you're replying to. This seems like a non-sequitor to me.

I'm certainly not going to disagree with you if you're saying 5e's skill system failed to close the loop on DCs when they failed to give examples of what qualifies as a "hard" roll.


Am I supposed to roll for Performance checks in most situations, or only life-threatening ones?You're supposed to roll if the result is in doubt or will depend on a variable quality of it. If you're just performing for your meal, you can probably do that just fine (though not as easily or for as fine a meal as someone with the Entertainer background).


When does my skill at that ability surpass a Cantrip or a level 1 spell?[.quote]Unless you have a specific Cantrip or Spell in mind to which to compare it, I can't tell you that. Arguably, if there's a comparable DC for a save offered by the spell, you could say your skill with the ability surpasses it when 8+proficiency+stat mod for the skill is greater than 8+proficiency+ stat mod for the spell.

[QUOTE=Man_Over_Game;24404137]There aren't answers on this in the books or even decent suggestions by the developers. Just speculation on the internet and the advice of "well, don't do that, and you won't be disappointed".I don't disagree that they could use more examples, but I fear you're not expressing yourself well, here, as I'm not sure I'm agreeing with what you're actually saying or not.


They kept skill DCs down in order to follow the same system as attacking, while making attacks always revolve around beating AC, and then they kept AC fairly static to make sure that weaker creatures can contribute in combat. However, the situations that a skill might be relevant for do not scale in the same way as AC, which is why Wizards get Dimension Door and Seeming. They didn't account for the fact that, while Casters should be able to deal with more and more perilous situations, that they'll still be able to affordably solve the problems they had before with things like Invisibility, Disguise Self, or Knock. Things that overlap and overshadow a skill-monkey specialist.Eh... yes and no. Invisibility doesn't actually let you auto-succeed on stealth, unless vision is the only way to sense you. Seeming replaces the DC of a d20+Charisma(Disguise Kit) check with 8+proficiency+Spellcasting Stat Mod DC; this is a case where you really can judge whether you're better with that Disguise kit than the wizard is with that spell.

Honestly, as a DM, if you had both? I'd follow Xanathar's advice and give you Advantage (so a +5 to the DC).


Is that what skill monkeys inevitably are rewarded with for levels 1-20? Being able to cast level 2 spells as rituals?
Followup question, how often do you expect a level 20 to need to cast a ritual?Level 20s probably cast rituals as much as level 7s. (Basing this on the fact that I don't think there are rituals above level 4 spells.) Wizards might cast a bit fewer if they took a ritual as a signature spell.

The classes that are dedicated skill monkeys are rewarded with more breadth as they level. More skill proficiencies and more Expertise, generally. I am answering your question, here, though, not making a judgment on whether that's "enough."


What of someone who is not a rogue or bard? I can fully accept a rogue or bard to be overall better at skills in general than others to be their Thing, but that does not equate to others must never be good at anything. What of the arm wrestling contest? Should there be a roll between a ST 18 and ST 10 contestants? If no, why is someone who says yes wrong? If yes. why is someone who says no wrong? What about between ST 18 and ST 18? ST 18 and ST 16? How much disparity between the ST scores does it take to be no roll, higher ST autowins? If never, then people will complain ST 10 can defeat ST 18 or even 20.

Suppose there was a defined rule that settles the arm wrestling contest. It may not satisfy everyone. That's people. There are defined rules for other things people don't like and change. Great Weapon Master feat being infamous for one. At least there's a defined rule so there's common ground. The rules define you need to have proficiency in a Tool to use it, which is why someone with Proficiency in Thieves' Tools can . . . while those who don't can't. 5E is not lacking in rules on how things work, even if someone doesn't like what the rules are. When it comes to skill use there they chose to let everyone do whatever they want. For some people that's hip hip hooray pop the cork. For others that's gnashing teeth of frustration, ergo this thread and others of similar complaint.I'm also unsure what your point is, here. You seem to be switching back and forth between "it's fine" and "it's crap."

I'll say this for the arm-wrestling match: you can always model it with a series of points representing shifting arms to one side or the other. Maybe if the winner is 0-5 higher than the other guy, they're deadlocked. For every 5 he beats the other guy by, he pushes one "point" to his advantage. To represent the way it's harder to push back up if you're down, the points-to-advantage add directly to the roll of the one who has them. If you reach +5, you press the other guy's hand to the table and win.


The skill rules aren't matching the OP's expectation on what animals can do, though I agree as another mentioned it's loophole wonky about elephants jumping and not specifically on how 5E handles skills. It's not enough that an animal has a +# to something. There's nothing to define what that +# can do. The OP limited it to animals, but PCs have the same problem which is why the thread expanded. The DM has to make it up and not every DM will agree on the solution.
Oh, this is a definite issue. But it's not a flaw in the skill system, but in the fact that they never closed the loop on it. It'd be easy to fix: a few examples for guidelines on what is expected to be "hard" and "easy" for each skill.

[Sorinth]

Over time, DnD has been transitioning to a concept that non-magical combatants should be near the same level as magical combatants. It happened in 4e, and people seem happy with the fact that that trend continued in 5e.

The problem is, that same concept hasn't progressed in the non-combat elements to the game. People are wanting that to be a solid pillar of the game, and there's really only a few ways of making it work:

Redo skill system
Rely on magic
Make things up as you go


And the books don't really provide much help on that third topic. I tried to get examples of what a distinct DC 25-30 would look like for a skill-user, to determine what kind of "super heroics" a character should be capable of as real-world people, and it basically boiled down to "Move an Immovable Rod by 10 feet". The only examples I could find of someone using mundane talents to do something supernatural was moving a make-believe object. The closest mention of interacting with a mundane object, at a DC 20, was breaking out of manacles, which almost any level 1 character could do in about a minute.

In comparison, a Wizard could afford to cast Invisibility for an hour on multiple creatures, for multiple hours, as early as level 5. Sure, that cost a resource, but that's kinda paltry compared to about 30% of the Rogue's class features. If a Wizard doesn't use Invisibility, he uses Flaming Sphere. If a Rogue doesn't use Stealth, he doesn't use 30% of his class. Even in the instances where they both are used, the Wizard's solutions are often "Yes" ("Yes, you did Dimension Door the party to safety". "Yes, you do read his thoughts and determine where the hostage is located"), where the Rogue's are often "Maybe" ("Roll a DC 20 for me").

We know it's not perfect. We can live with imperfect. My issue is that it's not good enough.

I never played 4e but wasn't that edition somewhat of a bust in large part because in order to balance things everything became generic/boring? I would also point out that in 5e they have not come anywhere close to balancing magic users vs non magic users.

I fail to see how bounded accuracy is the problem with your invisibility example. No matter how good of a bonus you give the rogue to his stealth check he will basically always be outperformed by the wizard turning himself invisible.

You keep bringing up magic as your proof that bounded accuracy/skill system is problematic yet it seems like your real problem is that magic is too powerful and removes the need for anything else. A high level fighter "suffers" just as much from bounded accuracy as the skill check system does, a high level fighter will still be taken out by a horde of weak monsters due to bounded accuracy much like a group of commoners can be more likely to succeed a skill check then a single high level character. But a wizard with a few fireballs will make short work of them.

[ad_hoc]

Bringing back take 10 alone would make a huge difference outside of combat.


There is no need for this at all.

Players declare what their characters are doing.

The DM determines whether they do it. Most of the time they just succeed in whatever they are doing.

Rolls should only be made if the outcome is in doubt and the result will be interesting.

There is no need for 'Take 10' in 5e because in all situations where it would apply the character should just succeed.

If we want to model something on going like a look out then we have a passive score to compare against.

[Segev]

If we want to model something on going like a look out then we have a passive score to compare against.

And, to be fair, the passive score is basically taking 10.

[Sorinth]

I think the issue is that any scenario with 9th level spells isn't a low-magic situation. Wish is just excessively powerful in a powerful-magic world.

Something like the Lord of the Rings knock-off of 5e is a better example of low-magic. Decent rules for mundane travel, martials have more versatility and do more out-of-combat, casters don't do much at lower levels (where even starting a fire is a big deal). And it works, because the universe is low-magic.

But half of 5e is tuned towards low magic, and it calls itself to be low-magic (like how they suggest you handle magic items), but the reality is that about 80% of your character content revolves around using magic in some way, about 50% of that revolves around spells, and the best generic feats are the ones that help you cast or resist magic. It's all over the place, constantly being thrown towards "low-magic" or "high-magic" at any given moment.

It'd take a big overhaul to make 5e a "low-magic" game, I think. Slower spell progression (progresses from level 1-6 instead of 1-9, maybe?), and slightly more spell slots than the current progression (to compensate for the power loss) would do it. Uses the old Rifts design of "Soldiers are there to kill, mages are there to do everything else". But that's just a guess I pulled out of the air.

The simple answer is to ban uber powerful spells like Wish. Honestly I was under the impression most tables didn't allow Wish because it's simply too disruptive.

[Man_Over_Game]

Sorry, Segev. I do realize some of what I said was unclear.

On the first part, I was referring to how skills can occasionally be automatic successes, but that is a similar solution to picking up something like Mend or Message.

One example I did refer to, when comparing skills vs. spells, was Mend vs. Tinker's Tools. The expected and recommended uses of Tinker's Tools is to repair, not create. However, Mend requires no roll, no materials, and reverts the object to its original condition. Tinker's Tools is technically unable to improve an item past its original state until you succeed a 25 DC.

So you could take a cantrip that has a 100% chance of success, no cost, and repairs the item as if it was brand-new.
Or you could take a skill that has a variable chance of success, the cost of materials, and you have a chance to either improve the item or destroy it further.

Which is a better option?

One other issue is that spells that replace skills have a sort of...reversed boolean logic about them. When you Hide, you're not just confirming the Hide, but also may also be worried about mistakes that cause detection. Something like Disguise Self or Invisibility are almost implied to work until something gives a reason for it not to. People do not roll Investigation Checks on people walking by. If you hear a noise down a hallway and don't see anything, you ignore it and keep going. Because the magic is either Works or Doesn't Work, this works differently than skills which have a degree of success. A small sound from a hidden Rogue turns into a small glance at the Rogue, which turns into a big problem.

We could fix that with educating DMs to make fewer nullifying rolls against skill-users, or allowing skills to do slightly more fantastical things (like allowing them to hide behind a suit of armor in a castle hallway without making a noise, normally something like a DC 20 probably) without requiring massive DCs.

Of course, if we make things easier, it'd mean that those who put no investment at all into a skill (a Barbarian trying to hide behind that suit of armor) would have an easier time, too. So it's a complicated issue.


I never played 4e but wasn't that edition somewhat of a bust in large part because in order to balance things everything became generic/boring? I would also point out that in 5e they have not come anywhere close to balancing magic users vs non magic users.

I fail to see how bounded accuracy is the problem with your invisibility example. No matter how good of a bonus you give the rogue to his stealth check he will basically always be outperformed by the wizard turning himself invisible.

You keep bringing up magic as your proof that bounded accuracy/skill system is problematic yet it seems like your real problem is that magic is too powerful and removes the need for anything else. A high level fighter "suffers" just as much from bounded accuracy as the skill check system does, a high level fighter will still be taken out by a horde of weak monsters due to bounded accuracy much like a group of commoners can be more likely to succeed a skill check then a single high level character. But a wizard with a few fireballs will make short work of them.

Bounded Accuracy effectively means that things scale slowly, so that weaker creatures can perform against stronger ones. The problem is, there isn't any reason for non-combat elements to follow this same example, since Bounded Accuracy was all based around AC.

Using the level progression of magic as an example as to how Bounded Accuracy isn't applied to non-combat elements of the game, skills should have followed the exact same suit. What we have now means that a useless creature can be almost as good as skills as a highly proficient creature, which is a problem if your class focuses around using and gaining skills, or if there are versatile and powerful casters in the party (since their investments quickly outscale yours).

That was the intent I was going with. Magic is a problem, but so is the assumption that irrelevant creatures should use skills nearly as well as powerful ones. Skills need to get on the same level as magic, but Fighters should be the best at what they're limited to doing (which is Fighting). The only people complaining about Wizards being able to teleport are the skillmonkeys that try to do the same.

As far as 4e goes, the "boringness" factor came from the lack of support of non-combat elements, which caused the game to be nothing more than battle chess as you jump from encounter-to-encounter. Play a game of chess for 6 hours each weekend and you'll get bored of it, too. Ironically, this is a similar reason why most of us don't do Adventure League games.

[Sorinth]

There is no need for this at all.

Players declare what their characters are doing.

The DM determines whether they do it. Most of the time they just succeed in whatever they are doing.

Rolls should only be made if the outcome is in doubt and the result will be interesting.

There is no need for 'Take 10' in 5e because in all situations where it would apply the character should just succeed.

If we want to model something on going like a look out then we have a passive score to compare against.

Agreed, though when I DM I'll ask for a roll and use the result to determine how quickly they succeed at the check. So for example if the Rogue is picking a lock, they still make a roll, and if they beat the DC they unlock it quickly, if they roll poorly I'll tell them they are struggling and it will take time but eventually they believe they can get through. They can either give up and move along or take the 5-10 minutes of time.

[Segev]

Sorry, Segev. I do realize some of what I said was unclear.

On the first part, I was referring to how skills can occasionally be automatic successes, but that is a similar solution to picking up something like Mend or Message.

One example I did refer to, when comparing skills vs. spells, was Mend vs. Tinker's Tools. The expected and recommended uses of Tinker's Tools is to repair, not create. However, Mend requires no roll, no materials, and reverts the object to its original condition. Tinker's Tools is technically unable to improve an item past its original state until you succeed a 25 DC.

So you could take a cantrip that has a 100% chance of success, no cost, and repairs the item as if it was brand-new.
Or you could take a skill that has a variable chance of success, the cost of materials, and you have a chance to either improve the item or destroy it further.

Which is a better option?Okay, that is clearer. Thanks!

Let's examine the relative value of Cantrips vs. skill/tool proficiencies. I'll use feats. Magic Initiate gives you two cantrips and a 1st level spell (usable 1/day). Prodigy (from Xanathar's Guide) is a human-blooded-only feat that gives half an ASI, a skill proficiency, and Expertise in any one skill in which you're proficient.

I'm going to make some arbitrary subdivisions here, so if you disagree with my valuations, please feel free to make an argument against them. But for now, I'm going to say the 1st level spell usable 1/day and the half-an-ASI are roughly equivalent.

That means that the skill/tool proficiency + Expertise is roughly equal to two Cantrips. Cantrips can come from a variety of sources, including race, class, and (of course) feats. Expertise is...only Bards and Rogues get it. And Bards have to go to level 3, first. I think I'm underselling it when I say Expertise is worth roughly 2x a skill/tool proficiency. That, however, would mean a skill/tool proficiency is about 2/3 of a cantrip in value.

If you take mending, you're fixing damaged items more-or-less for free.
In practice in D&D, if you have Tinker Tools and the proficiency, you're probably doing more or less the same on day-to-day wear and tear, if the DM is even tracking that. It will cost you something, possibly, to repair an item that is damaged enough to care about fixing it being a plot point.
Tinker Tools, per your statement about what you can do with a 25+ on the roll, can fix things better than they originally were, which mending never can do. Under most circumstances, you should probably just succeed if you're proficient with them, but when you do have to roll, mending is probably superior. Certainly more reliable.

If there are any instances where you need to jury-rig or alter things, tinker tool proficiency will let you do it; mending will not.

I'd rate the two as about equally useful (though I confess that mending seems like more fun to me), while tinker tool proficiency would seem to be about 2/3 in build-resource-value what mending is in terms of investment.



One other issue is that spells that replace skills have a sort of...reversed boolean logic about them. When you Hide, you're not just confirming the Hide, but also may also be worried about mistakes that cause detection. Something like Disguise Self or Invisibility are almost implied to work until something gives a reason for it not to. People do not roll Investigation Checks on people walking by. If you hear a noise down a hallway and don't see anything, you ignore it and keep going. Because the magic is either Works or Doesn't Work, this works differently than skills which have a degree of success. A small sound from a hidden Rogue turns into a small glance at the Rogue, which turns into a big problem.

We could fix that with educating DMs to make fewer nullifying rolls against skill-users, or allowing skills to do slightly more fantastical things (like allowing them to hide behind a suit of armor in a castle hallway without making a noise, normally something like a DC 20 probably) without requiring massive DCs.

Of course, if we make things easier, it'd mean that those who put no investment at all into a skill (a Barbarian trying to hide behind that suit of armor) would have an easier time, too. So it's a complicated issue.
Well, systems don't work well if the DMs don't know how to use them, no.

I will say that, for invisibility, not only does it use up somebody's Concentration, but it's a very limited resource, whereas you can Hide all day. Disguise self and using a Disguise Kit both are assumed to work unless somebody is trying to see through it for some reason, in which case they just have different means of setting the DC. With invisibility, too, it is NOT assumed that you're undetected. If you don't take the Hide action, people know you're there unless they have no means save sight to detect you. (Of course, this only matters if they care that they noticed "someone there," and is not assumed to make you stand out for being invisible.) You still need to roll Stealth to sneak past guards, even when Invisible. You just automatically make the minimum requirements to Hide at all.

[47Ace]

First off I never said anything close to "your higher level is wrong", there is no right/wrong between low/high magic games just like there's no right/wrong between combat focused vs role play focused games.

And for the record a Barbarian with expertise in Atheletics from say a feat at level 20 has a +19 bonus to his atheletics check so it's not like they can't perform insane feats of strength.

It's also quite easy to play a low magic setting without resorting to limit level advancement. Plenty of tables ban the Wish spell because it's too disruptive, are you claiming they are wrong and should just limit level advancement so they don't get 9th level spells?

I agree that there is no right/wrong between high and low magic games. That was kinda the point I was making that people can enjoy high magic games with mighty feats of strength by heroes so awesome that they can stand shoulder to shoulder with magic users capable of moving the earth itself, summoning swarms of meteors and calling great beings of fire to fight at there side. While others (or even the same people) lower magic games where abilities are closer to reality. I do question if you can play low magic without limiting the high level magic available but, maybe you can. My point was that I feel that both styles of play would be better served by providing a game structure that allowed you to stop level advancement at 10 well still having some form of progress (feats, boons, something else) so stopping level advancement at level 10 (or 6 or 16) would feel like a choice. Then the designers would have more of a free hand to give non magic users abilities comparable to the level of fantasy that high level spell caster live in.

With regards to the barbarian yes with at level 20 with the right class, race(Human, Half-elf, or Half-Orc), and feat (Prodigy) you can break out of manacle every time but that is a lot of investment to be able to do something a druid could probably due at level 2 by turning into a Brown bear or a Wizard could do at level 7 with Polymorphing into a large beast.

I am sorry again that I probably came off as a little rude that is my fault. And however you play D&D is right as long as you are having fun. If banning wish works for you then great. If running a roleplay focused 5e game that feels meaningful and consistent is something you can do then I'm impressed you are (or have) a way better DM then I am and if your roleplay focused game is not meaningful and consistent and its still fun then all the power too you.

I still feel that the 5e designers could have done a better job in providing a skill system that would help DMs build encounters that alow non-magical solutions that can be consistently done other then hitting things with swords. Or just a skill system that didn't make wizards as good (roughly) as clerics at religion checks.

[MaxWilson]

Bounded Accuracy is easily one of the best features of Fifth Edition. Doing away with the enormous bonuses that a character could accrue in 3.5e and Pathfinder 1e is a nearly unambiguous improvement as far as fairness and ease of play were concerned. That said, there are some places where it would do with nuance that it presently lacks.

There have been many threads on these forums concerning how to deal with small bonuses and the randomness that comes with them, and how 5e's skill system doesn't have the right setup for simulating the way actual skills work, so I won't go into a discussion of how skills will work when talking about people and training.

Instead, I wish to discuss animals, and how their given stats portray their abilities. Not monsters, mind you, just normal real-world animals that have stats assigned to them. In particular I want to talk about animal sight, smell, and hearing.

A bear for example has a powerful nose, one that can detect scents as much several million times fainter than a human's nose can detect. That translates to.... advantage on Wisdom (Perception) checks with a +3 flat bonus. That translates to about a +8 bonus, but one that still doesn't let it exceed 23 on a natural 20.

Eagles can see minute movements in fields miles away from them, enough that a human with similar acuity would be able to recognise faces a couple of miles away. That translates to.... advantage on Wisdom (Perception) checks that rely on sight. With a +4 bonus.

Elephants have +0 to their hearing, and no advantage. That's exactly the same as a human with 10 wisdom and no proficiency. They don't have their real life Tremorsense either.

A first-level PC with +2 Proficiency and 14 Wisdom (+2) can match the upper range of these animals' senses, even if not as reliably as a creature with advantage could. Perhaps it's not so good that a 1st-level familiar or 2nd-level druid in wildshape should have superpowers of high perception that far outclass anything a human could do at any level, but still, a bear should have a better nose than a level 1 human with advantage.

Basically, there are many places where Bounded Accuracy is not reasonable to apply, because it's based on how humans perform at certain tasks and humans are not representative of the average animal in most respects.

That's not a bounded accuracy problem per se, it's a missing-special-ability problem.

If hypothetically 5E Eagles didn't have flight but did have +4 to Athletics, it would be wrong to argue that they need a higher Athletics bonus to let them fly. They need flight. Similarly, the problem with 5E bloodhounds and bears isn't that they have only +3 and advantage to ability checks for smelling--it's that they don't have any abilities that make them capable of following scent trails and perceiving scents from far away. All they have is a numeric bonus in a vacuum, which isn't enough.

So yes, I agree that there's a problem, but the solution isn't bigger bonuses.


I mean is this an issue with the skill system or an issue with WoTC's lack of care in making mundane animals? All you would need to do to fix this sort of stuff is tack on some special abilities that fulfill these attributes. Things like Bear Nose - You cannot be hidden from the bear while you are within 300 feet of it and it can smell you. Jumpless - The Elephant is incapable of jumping. The eagle example is particularly egregious just because Eagle Barbarians get actual eagle sight while the eagle in the monster manual doesn't.

Yeah, this. And fantasy monsters too. I'm sure elephants aren't the only things that shouldn't be able to climb trees and jump from branch to branch. Chuuls, Purple Worms, Remorhazi, Winter Wolves, Black Puddings, and Hydras come to mind. If you say to yourself, "Hey wait, I disagree. I think Hydras probably could climb trees," well, that disagreement is precisely the issue with the MM: they're not described in enough detail for us all to have the same idea of what the monster's author intended them to be like.

5E's MM is waaaaay too focused on HP/attacks/damage and not enough on behavior/ecology/interesting traits.

[Sorinth]

I agree that there is no right/wrong between high and low magic games. That was kinda the point I was making that people can enjoy high magic games with mighty feats of strength by heroes so awesome that they can stand shoulder to shoulder with magic users capable of moving the earth itself, summoning swarms of meteors and calling great beings of fire to fight at there side. While others (or even the same people) lower magic games where abilities are closer to reality. I do question if you can play low magic without limiting the high level magic available but, maybe you can. My point was that I feel that both styles of play would be better served by providing a game structure that allowed you to stop level advancement at 10 well still having some form of progress (feats, boons, something else) so stopping level advancement at level 10 (or 6 or 16) would feel like a choice. Then the designers would have more of a free hand to give non magic users abilities comparable to the level of fantasy that high level spell caster live in.

With regards to the barbarian yes with at level 20 with the right class, race(Human, Half-elf, or Half-Orc), and feat (Prodigy) you can break out of manacle every time but that is a lot of investment to be able to do something a druid could probably due at level 2 by turning into a Brown bear or a Wizard could do at level 7 with Polymorphing into a large beast.

I am sorry again that I probably came off as a little rude that is my fault. And however you play D&D is right as long as you are having fun. If banning wish works for you then great. If running a roleplay focused 5e game that feels meaningful and consistent is something you can do then I'm impressed you are (or have) a way better DM then I am and if your roleplay focused game is not meaningful and consistent and its still fun then all the power too you.

I still feel that the 5e designers could have done a better job in providing a skill system that would help DMs build encounters that alow non-magical solutions that can be consistently done other then hitting things with swords. Or just a skill system that didn't make wizards as good (roughly) as clerics at religion checks.

The problem is in a high magic world where players are throwing around the Wish spell, skills should pale in comparison to that type of magic. If taking profiency in Stealth was better then turning invisible then turning invisible would seem like a pretty crappy spell. It seems logical that in a low magic settings skills matter a lot more then they do in a high magic setting where you use magic for everything.

I think the only way to make skills seem powerful/useful is to make magic worse or make it costly in some way.

[Sorinth]

That's not a bounded accuracy problem per se, it's a missing-special-ability problem.

If hypothetically 5E Eagles didn't have flight but did have +4 to Athletics, it would be wrong to argue that they need a higher Athletics bonus to let them fly. They need flight. Similarly, the problem with 5E bloodhounds and bears isn't that they have only +3 and advantage to ability checks for smelling--it's that they don't have any abilities that make them capable of following scent trails and perceiving scents from far away. All they have is a numeric bonus in a vacuum, which isn't enough.

So yes, I agree that there's a problem, but the solution isn't bigger bonuses.



Yeah, this. And fantasy monsters too. I'm sure elephants aren't the only things that shouldn't be able to climb trees and jump from branch to branch. Chuuls, Purple Worms, Remorhazi, Winter Wolves, Black Puddings, and Hydras come to mind. If you say to yourself, "Hey wait, I disagree. I think Hydras probably could climb trees," well, that disagreement is precisely the issue with the MM: they're not described in enough detail for us all to have the same idea of what the monster's author intended them to be like.

5E's MM is waaaaay too focused on HP/attacks/damage and not enough on behavior/ecology/interesting traits.

The problem is if you want to cover every possible case then the entries for each monster would be pages full of stuff. It's much better to simply leave it up to the DM to decide if something is possible.

But yes adding behaviour, tactics, sociology would've been a welcome addition.

[MaxWilson]

The problem is if you want to cover every possible case then the entries for each monster would be pages full of stuff. It's much better to simply leave it up to the DM to decide if something is possible.

But yes adding behaviour, tactics, sociology would've been a welcome addition.

You can't cover every possible case of every thing, but it wouldn't have been bad to at least write the special abilities that they did write correctly instead of lazily (bears with Keen Scent: should have gotten more than just advantage on perception checks involving smell), and exceptions to the PHB rules (call out which monsters are intended to be exceptions to the general rules on climbing/jumping), and to have had maybe a paragraph or two in the stat block about social organization, behavior patterns, or interesting traits.

Is it really more important for us to know that Abominable Yetis "have advantage on Wisdom (Perception) checks that rely on smell" and that they are "brutal rampagers" who howl loudly (like I wouldn't have guessed?) than to have some interesting facts unique to yetis that a DM could use to spark a player's imagination and/or foreshadow Yeti presence? E.g.

Fun fact #1: yeti hair melts when warmed to slightly above human body temperature.
Fun fact #2: in all but the coldest climates, yetis burrow into the snow and hibernate through the summer.

Maybe it's unfair for me to pick on yetis because they do at least have a tiny bit of info on social organization ("hunt in solitude or in small family groups", cannibalistic) and some out-of-combat info ("can smell living flesh from miles away") that isn't well-integrated with the stat block (not mentioned in Keen Smell).

Nobody is asking for exhaustive lists, just less of what is useless and cliched and more of what is useful before initiative is rolled. (I know, I'm not going to get it.)

[Sorinth]

You can't cover every possible case of every thing, but it wouldn't have been bad to at least write the special abilities that they did write correctly instead of lazily (bears with Keen Scent: should have gotten more than just advantage on perception checks involving smell), and exceptions to the PHB rules (call out which monsters are intended to be exceptions to the general rules on climbing/jumping), and to have had maybe a paragraph or two in the stat block about social organization, behavior patterns, or interesting traits.

Is it really more important for us to know that Abominable Yetis "have advantage on Wisdom (Perception) checks that rely on smell" and that they are "brutal rampagers" who howl loudly (like I wouldn't have guessed?) than to have some interesting facts unique to yetis that a DM could use to spark a player's imagination and/or foreshadow Yeti presence? E.g.

Fun fact #1: yeti hair melts when warmed to slightly above human body temperature.
Fun fact #2: in all but the coldest climates, yetis burrow into the snow and hibernate through the summer.

Maybe it's unfair for me to pick on yetis because they do at least have a tiny bit of info on social organization ("hunt in solitude or in small family groups", cannibalistic) and some out-of-combat info ("can smell living flesh from miles away") that isn't well-integrated with the stat block (not mentioned in Keen Smell).

Nobody is asking for exhaustive lists, just less of what is useless and cliched and more of what is useful before initiative is rolled. (I know, I'm not going to get it.)

You don't want an exhaustive list but your first examples included things like can Hydra's climb trees.

Personally I would much rather the books be filled with your fun facts and out of combat info then put in a bunch of stuff that is just common sense.

[Pex]

I'm also unsure what your point is, here. You seem to be switching back and forth between "it's fine" and "it's crap."

I'll say this for the arm-wrestling match: you can always model it with a series of points representing shifting arms to one side or the other. Maybe if the winner is 0-5 higher than the other guy, they're deadlocked. For every 5 he beats the other guy by, he pushes one "point" to his advantage. To represent the way it's harder to push back up if you're down, the points-to-advantage add directly to the roll of the one who has them. If you reach +5, you press the other guy's hand to the table and win.


I'm not saying it's both good and bad, only pointing out some people think it's good while others think it's bad.


Oh, this is a definite issue. But it's not a flaw in the skill system, but in the fact that they never closed the loop on it. It'd be easy to fix: a few examples for guidelines on what is expected to be "hard" and "easy" for each skill.

I totally agree there should have been benchmarks, but past discussions of this matter has shown people who love the system as is find this idea abhorrent.