Common issues with DC based skill systems

[Zuras]

d&d 5e is a particularly poor application of the skill system. at least based on my experience with baldurs gate 3, which is my main source of knowledge on how 5e works.
you don't gain skill ranks. you can never be a specialist. I can be a senior professor, but ultimately I have nothing more than skill proficiency, and a college freshman is going to have my same modifier. which in turn makes the handling of how difficult a task should be a complete mess-up.

That’s not a flaw of the system so much as a consequence of not making the skill system entirely separate from the class & level system. The system isn’t equipped to distinguish between a professor and a freshmen who don’t have class levels.

[Hrugner]

this is already covered by a dc-based system. become good enough at a certain skill, and you have automatic success. apply measures to make the task easier (like the climbing equipment example), and you can make it an automatic success for anyone. if the task is hard enough, it can't be done.
a dc-based system includes a chance to succeed or fail when a task is just hard enough that you could maybe succeed.
it also includes a way to handle opposite use of skills (like stealth against perception) which your proposal would lack.

The roll based portions of the system are useless, so we're cutting them out. The abstraction to a number is also useless, so we're cutting that out too. This makes skills reliable enough to be planned around, and removes the ability to overwhelm difficulty with raw numbers. If that's basically the same thing to you, we'll just have to agree to disagree I guess.

It works just as well with contested rolls. What you want to know with stealth vs. perception is at what range a person is perceived. You give people a "plain sight with no distractions" detection radius, and decrease the size of that radius based on various advantages like cover, darkness, camouflage, weather, noise, or what have you. The skilled sneaker reduces the radius without the purchased advantages, and the skilled perceiver buys off the benefit of those advantages. This cuts out the slow crawl of multiple checks on approach as you know the encounter distance, and lets you include less stealthy members in your team if a very advantageous approach is available.

[King of Nowhere]

The roll based portions of the system are useless, so we're cutting them out. The abstraction to a number is also useless, so we're cutting that out too. This makes skills reliable enough to be planned around, and removes the ability to overwhelm difficulty with raw numbers. If that's basically the same thing to you, we'll just have to agree to disagree I guess.

I agree to disagree all right, but you got me curious: why would you think rolling is useless?
and why would you remove the random factor when it's sensible? As I mentioned, I rarely call for a skill check, especially at high level; however, there are situations where a skill check may go either way, even for someone super skilled. performing particularly tricky acrobatics, improving a magic ritual that you've never seen before, opening an incredibly complex lock in less than 5 minutes, remembering details on an unimportant wizard who took a minor part in a conflic a century ago... I label those tasks as "stuff that a regular skilled human could not achieve, and someone supernaturally skilled just might", and slap on them a DC between 40 and 50 - the party specialist can try those, and success is not guaranteed. unless I misunderstood, you would basically remove the randomness from such events, turning them into either sure successes, or outright impossible tasks.

This cuts out the slow crawl of multiple checks on approach as you know the encounter distance, and lets you include less stealthy members in your team if a very advantageous approach is available.

You can do that with the dc system too. just decide that the approach is advantageous enough that there's no need to roll - or that the stealthy scout is leading the group and avoiding the guards, covering for the less stealthy members.
maybe you played with some peculiar gm that asked for 10 consecutive rolls as you gradually progress closer to the guard on the minimap, and that's why you dislike the dc system and want to change it; but our little exchange got me thinking, it's not really the system that makes the difference, but how the players and gm choose to interpret it.

Something to note is that 5E and other D&D systems’ skill systems are for ADVENTURING.
They are NOT meant to simulate teaching a college course, or activities like that.

pretty much any party wizard with whom I played was basically a college professors. often enough, other party members also had a background in research and academy. in my last campaign, we rolled knowledge and spellcraft more often than we rolled stealth or detect traps. tampering with magic rituals or finding weaknesses in magic defences were very common activities.
Still, just because the system is not meant to simulate something, it doesn't mean you can't do it.

[JNAProductions]

Sure-but it’s a round peg into a chicken dinner.
You CAN do it, but it’s not designed for it-so expect some issues, if you use it without adjustments.

[TaiLiu]

I think Powered by the Apocalypse games (e.g. Monster of the Week) have basically solved problems a to c. At the very least, they're usually better than d20 systems (e.g. D&D 5e).

1. Forgetting that requiring a series of multiple successes dramatically worsens the odds for any task.
   
   The 2d6 system is a bell curve just from the start, and the odds of an average result increases the more you roll. Technically, the latter is also true for d20 systems, but the curve is much wider and slower to form.
   
   
2. Instinctively scaling up challenges so there is still some chance of failure, without any in-universe logic.
   
   The DC is always the same in PBTA games. You roll a 2d6 plus your modifiers. A 10+ is a success, a 7-9 is a partial success, and 6 under is a failure. You never have to worry about setting the DC too high or low—the system takes care of that for you.
   
   
3. Relatedly, accidentally gating all non-specialist characters out of participating by setting the DCs to be a reasonable challenge for the party expert.
   
   The 2d6 system is brilliant. The most likely result, without modifiers, is a 7—a partial success. In contrast, take a d20 system like 5e. Let's say you need to roll for an easy task (DC 10). Without modifiers, your odds of success and failure are the same. On a d20, the odds are both frightening wide and flat. There's no bell curve of competence to save you.

[NichG]

i have no idea what exactly stamina would be in your system, but while it makes sense for a jump or tumble, why would a knowledge or spot check require stamina? and by "spending a success to succeed at a thing" you basically remove the different difficultes: sure, you need 5 ranks to even try this difficult task, but once you do have those 5 ranks you only need one success, so it's almost trivial. also, there's no good mechanic for an opposed check.

This is intentional. The roll is to determine what things cost and what side-effects you can introduce or avoid, not 'do you succeed?'. In general the question of whether you could or couldn't succeed is answered by your static skill level, not by the roll.

So something like a 'Knowledge check' wouldn't happen to answer the question 'do you know something?'. But it could happen for something like 'I want to use my Knowledge to build a bridge'. You roll to find out how much that will cost you, what side-effects will occur and which ones you can buy off, how long it will take, etc. In that sense, the expert will simply never fail unless something is wrong - they're exhausted, they're under time serious pressure, etc. Because even if they roll zero successes, they just spend another 3 Stamina and succeed.

[Hrugner]

I agree to disagree all right, but you got me curious: why would you think rolling is useless?
and why would you remove the random factor when it's sensible? As I mentioned, I rarely call for a skill check, especially at high level; however, there are situations where a skill check may go either way, even for someone super skilled. performing particularly tricky acrobatics, improving a magic ritual that you've never seen before, opening an incredibly complex lock in less than 5 minutes, remembering details on an unimportant wizard who took a minor part in a conflic a century ago... I label those tasks as "stuff that a regular skilled human could not achieve, and someone supernaturally skilled just might", and slap on them a DC between 40 and 50 - the party specialist can try those, and success is not guaranteed. unless I misunderstood, you would basically remove the randomness from such events, turning them into either sure successes, or outright impossible tasks.


You can do that with the dc system too. just decide that the approach is advantageous enough that there's no need to roll - or that the stealthy scout is leading the group and avoiding the guards, covering for the less stealthy members.
maybe you played with some peculiar gm that asked for 10 consecutive rolls as you gradually progress closer to the guard on the minimap, and that's why you dislike the dc system and want to change it; but our little exchange got me thinking, it's not really the system that makes the difference, but how the players and gm choose to interpret it.

The random element on skill checks makes skills something you can't rely on or plan around. It also introduces problems with multiple people making the same check hoping for a high roll, and reduces the viability of making group checks where any one person failing makes the whole roll fail. Similarly, the longer you spend doing something the less likely you are to succeed if you are making checks every turn you take the action, which is the opposite of the typical relationship between time spent and difficulty. We can make the random die roll work by working around the problems introduced by rolling, modifying the pace of rolling or who is permitted to roll for what, or junk rolling and come up with something that isn't random and never introduce those problems in the first place.

Is the random factor ever sensible? Tricky acrobatic feats are accomplished through practice, and the 2000 sydney olympics showed us what a 5cm difference on a vaulting horse can do to that practice. The lock is a similar issue, if you're familiar with the lock, it's no issue, if you aren't you can't luck through it, and in the end, it could be easier to disassemble it. Recalling trivia doesn't feel random, and at a certain point you're recalling something so minor that your own certainty of what you've read is probably stronger than the certainty of the person who recorded the trivia to begin with.

This makes our random chance look less skill based and more literally random. Did you read the book where the wizard's name was recorded, have you practiced an acrobatic routine similar to this one, have you opened a lock like this one. We can use the die to emulate this, even reusing the roll if the player needs to accomplish the same feat again in the future, or we can discard the die and instead record what benefits the character no longer needs to accomplish certain tasks. From my perspective, we're doing a lot of work to keep the dice, and there's no real benefit to the game in doing so.

[Vahnavoi]

What problems do you regularly encounter with DC based skill/resolution systems?

1) too many small modifiers to add on either side, leading to lot of counting for what is often just a pass/fail check.

2) tables of static target numbers that try to be too complete. (As in, try to tell me about very specific things instead of giving general examples.) BONUS: the game is somehow both this and the above (because precalculating your mods into the tables was too much to ask).

3) being flaky about what type of things ought to be added to the target number versus the roll needed to pass it. EXTRA BONUS: the same thing alternates between where it goes (f.ex. a bad tool can either make your roll worse or the target number higher). EXTRA MEGA BONUS: the same thing gets counted twice due to even the game designer forgetting what goes where (f.ex. a bad tool makes your roll worse AND the target number higher!)

4) Calculation loops; players may love unbounded exploding dice, but I don't.

5) Pretending your target numbers don't just work out to a percentile system with 5% increments. (Could've just given me percentiles.)

6) Ripping off target numbers from a system that DID just work out to a percentile system with 5% increments, then using dice that don't work in 5% increments. (2d6 based systems using 5 and 10 as target numbers, I'm looking at YOU.)

7) purposelessly hiding target numbers from players, making it hard for players to count their odds when they're supposed to.

8) purposelessly revealing target numbers to players when they are not supposed to know their odds. EXTRA MEGA LUCK BONUS: does both this and 7) at the time.

9) using target numbers at all yet insisting on using some stupidly complicated random number generator, because you don't want players to "game the system" by counting their odds. (This only means the game master also has difficulty counting the odds, which is a bother for scenario design.)

10) Somewhat related to 6) and 9): using target numbers that are easy to remember (supposedly) but don't signify useful breakpoints in probability given the random number generator of the game. See also: the other thread about success chances.

11) Being ignorant of or actively hostile to any math above first grade of elementary school in setting or manipulating target numbers. (Because probability is cool but multiplication isn't?!?)

I’ve seen many people complain that they can make RPGs devolve into a game of mother-may-I, and while I clearly see how this could happen, it’s not something I’ve ever encountered myself. Even in systems where you’re literally making everything up as you go along, like FATE, the table always reached a consensus that a given task difficulty was fair and the aspects a player applies to it made narrative sense.

That's because people who use "mother-may-I" as a derogatory are wrong. "Mother-may-I" is an actual game with actual rules, not something a game "devolves" into; there's a pretty good chance your experiences with FATE (etc.) are equivalent to Mother-may-I played correctly.

The problems I *have* encountered, and have to periodically guard myself against as a GM have actually been:

a) forgetting that requiring a series of multiple successes dramatically worsens the odds for any task

b) Instinctively scaling up challenges so there is still some chance of failure, without any in-universe logic. I have to guard myself against making every lock in town suddenly masterwork difficulty after the rogue gets Reliable Talent.

These two are real problems but don't have a lot a lot do with target numbers. A) is about using dice at all while neglecting cumulative probability, b) is dynamic difficulty done badly. I see the first in virtually all games with dice, while the latter is often injected into games by a game master in scenario design phase even if nothing in system rules tells you to do it.

c) Relatedly, accidentally gating all non-specialist characters out of participating by setting the DCs to be a reasonable challenge for the party expert (especially in games with constant scaling/no bounded accuracy). This is fine for tasks like developing a cure for the plague virus, but terrible for many social and exploration tasks where other PCs should reasonably be able to participate.

Now this is a calibration issue that does directly have to do with target numbers. The two main offenders I've noticed are:

1) static target number too high for common tasks; what I'd call a simple calibration error. Caused by a game designer thinking something is either more difficult than it'd reasonably be or underestimating how often the roll needs to be made. Fixed in post by knowing better than the game designer and giving a more appropriate target number, and in general, not asking rolls for trivial things.

2) target number derived from unbounded opposed roll. This creates an arms race effect where eventually only people who can compete are those who heavily invested. Caused by the game making it too easy to get a high bonus to a check. Fixed in post by directly bounding target numbers or capping investments.

[Satinavian]

That’s not a flaw of the system so much as a consequence of not making the skill system entirely separate from the class & level system. The system isn’t equipped to distinguish between a professor and a freshmen who don’t have class levels.

It's a barebone bad excuse of a skill system designed by people who believe that combat and possibly magic are the only important parts of he game.

It's a big reason to not play D&D5.

[Telok]

If this thread is going to be nothing but a rehash of how to houserule/interpret the current D&D version of everything noncombat to make it not lol-random then could we mark it as such and move it to the right subforum?

[Zuras]

If this thread is going to be nothing but a rehash of how to houserule/interpret the current D&D version of everything noncombat to make it not lol-random then could we mark it as such and move it to the right subforum?

I actually, started the thread in response to complaints about Pathfinder 2e’s skill system, but I did ask for specific examples, which is going to naturally lead to more D&D 5e examples than anything else.

Also, it’s not like the various procedural ideas (tracking multiple successes, failing forward, success at a cost, the three clue method for mysteries) aren’t reasonably system neutral. I’m not sure how such a discussion is anything other than general. I play 5e without any codified house rules myself, but I use ideas from GURPS, FATE, Dungeon World and half a dozen other games in handling skill-based challenges.

[Tanarii]

That's because people who use "mother-may-I" as a derogatory are wrong. "Mother-may-I" is an actual game with actual rules, not something a game "devolves" into; there's a pretty good chance your experiences with FATE (etc.) are equivalent to Mother-may-I played correctly.

I've been seeing this terminology used on these forums for years without ever realizing Mother-may-I is an actual game. And knowing that an looking up the rules, you're entirely correct. What's being derogatorily called Mother-May-I is nothing like the game.

I can't see playing Fate or any other RPG like it though. Because as I understand it, playing an RPG like Mother-may-I would be the GM instructing the players on what their characters are to do, and them either asking GM-May-I then executing it, or forgetting to ask and ... I dunno, going back to town?

Edit: What's being disparaged in terms of a skill system would probably be more effectively disparaged by calling it after the game "Twenty questions" instead. Which I have occasionally seen it called as well.

[Zuras]

I've been seeing this terminology used on these forums for years without ever realizing Mother-may-I is an actual game. And knowing that an looking up the rules, you're entirely correct. What's being derogatorily called Mother-May-I is nothing like the game.

I can't see playing Fate or any other RPG like it though. Because as I understand it, playing an RPG like Mother-may-I would be the GM instructing the players on what their characters are to do, and them either asking GM-May-I then executing it, or forgetting to ask and ... I dunno, going back to town?

Edit: What's being disparaged in terms of a skill system would probably be more effectively disparaged by calling it after the game "Twenty questions" instead. Which I have occasionally seen it called as well.

Fundamentally the game is supposed to be a conversation between the players and GM, with periodic consultations to an RNG when you want to avoid predetermined outcomes. Calling it Mother-May-I is just one way of stating you’re having a bad conversation.

It can be a big problem, but in my experience it’s just a sub-category of problems caused when the players and GM aren’t on the same page.

[Tanarii]

Calling it Mother-May-I is just one way of stating you’re having a bad conversation.

That's not the context I see it used in.

I see it used as "I don't know what value the GM is going to assign to my probability of success before I tell her what the action is."

Meanwhile the game Mother-May-I appears to be "do as you're told, but only if you remember to ask nicely first, otherwise get punished". (Reference: https://www.toyassociation.org//geni...x#.ZDOzoy9lAgo)

[Vahnavoi]

I've been seeing this terminology used on these forums for years without ever realizing Mother-may-I is an actual game. And knowing that an looking up the rules, you're entirely correct. What's being derogatorily called Mother-May-I is nothing like the game.

And now you now better. I'll explain what the spirit of the derogatory is along with more things about the actual Mother-may-I game below.

I can't see playing Fate or any other RPG like it though. Because as I understand it, playing an RPG like Mother-may-I would be the GM instructing the players on what their characters are to do, and them either asking GM-May-I then executing it, or forgetting to ask and ... I dunno, going back to town?

Consider what the people who want to know every target number beforehand actually want. They want a game designer write in a book instructing exactly what their character can do and how, which they then have to ask their game master to execute, and if they fail to follow the instructions or communicate their intent to their game master, either nothing happens or they fail and suffer the consequences.

In short: everybody who is following rules set by someone else is already playing Mother-may-I in the sense you are using it in this paragraph. Those who complain about it just have this weird insistence that a person who usually isn't even present (writer of a gamebook) would or should be a better "mother" than the person actually holding their game (the local game master).

There is, however, slightly more to this, see end of this post.

Edit: What's being disparaged in terms of a skill system would probably be more effectively disparaged by calling it after the game "Twenty questions" instead. Which I have occasionally seen it called as well.

Which is even sillier. I would guess that, along with the complaint, you have seen someone say "I don't want to guess what my game master is thinking". The entire point of "Twenty questions" is that with twenty yes-or-no questions it's possible to logically deduce what's in anyone's mind, regardless of what it is. The game's interesting enough that some roleplaying games already include it as a subgame - see, for example, D&D's divination spells. The corollary being that Twenty questions would make an excellent backbone for information-acquiring skills in a skill system.

Furthermore, since any consistent scenario is open to logical inquiry, a savvy player can use the format of Twenty questions to inquire and deduce information in any game that allows them to ask questions. So, people who try to use name of the game as a disparaging comment usually just mean they think their game master isn't giving them enough information, they hate having to ask questions or they hate having to reason about the answers. The first part can be valid, the latter two usually just means the complainer is a negative nelly. If the problem is that a player isn't allowed to ask questions to deduce information, then the comparison to Twenty questions is invalid to begin with.

That's not the context I see it used in.

I see it used as "I don't know what value the GM is going to assign to my probability of success before I tell her what the action is."

Meanwhile the game Mother-May-I appears to be "do as you're told, but only if you remember to ask nicely first, otherwise get punished". (Reference: https://www.toyassociation.org//geni...x#.ZDOzoy9lAgo)

That ruleset is a bit odd because it is different in one important respect to one classic version of the game - this is not unusual, children's games have numerous common variants under the same name. However, that rule difference explains the disconnect. Let me quote Wikipedia on the same topic and you'll see:

"One player plays the "mother", "father" or "captain". The other players are the "children" or "crewmembers". To begin the game, the mother or father stands at one end of a room and turns around facing away, while all the children line up at the other end. The children take turns asking "Mother/Father, may I ____?" and makes a movement suggestion..."

Underlines for emphasis.

In the other common variant, the players have to negotiate their move with the "mother", not simply do as they're told. Which is, in a nutshell, the FATE paradigm of negotiating your action and fair arbitration to it.

But it doesn't stop there. This negotiation paradigm is something that can be used to play any turn-based game, and is in fact a natural way to play any turn-bases game with a referee when a player is uncertain of or unfamiliar with the rules. (Player: "hey, referee, may I do this?" Referee: *checks rules* "You may / you may not".) This applies regardless of how hardcoded the rules are - to bring us back to main topic, whether target numbers are taken from a preplanned table or decided on the spot makes little difference to a player who doesn't know what the target number is.

People who use "Mother-may-I" as derogatory hates this negotiation step. They want the information to already be there on a cheatsheet so they can spy it on their own leisure. They might want this because skipping the negotiation step makes a game faster. This may or may not be fair; as noted above with "twenty questions", it's possible a game master is genuinely giving too little information. But quite often, these possibly valid criticisms blend together with other pet peeves that have little to do with Mother-may-I. Often, the game is brought up a derogatory simply because it's a children's game, and some people are actively hostile to any notion that their super-serious-game-for-adults-that-is-totally-not-playing-pretend-with-friends is somehow equivalent to a children's game

Meanwhile, the actual Mother-may-I game exist to subtly teach kids, in a fun way, basic things like identifying who is in charge, patiently waiting for your turn, making reasonable suggestions and deciding whether the person in charge is giving reasonable commands. You know, things that are good manners in almost any tabletop game. I would go so far to suggest that, given time, any playgroup seeking to play roleplaying games together ought to play few rounds of Mother-May-I first. People who do badly in it or refuse to do so in principle aren't going to be good company.

[Captain Cap]

In short: everybody who is following rules set by someone else is already playing Mother-may-I in the sense you are using it in this paragraph. Those who complain about it just have this weird insistence that a person who usually isn't even present (writer of a gamebook) would or should be a better "mother" than the person actually holding their game (the local game master).

It's about expectations: if I know beforehand (as it is already laid out in an accessible ruleset) what it takes to be a good climber in the game, what I can expect a good climber to do, how much of a good climber a character can be, then I can create a good climber character whose skill is gonna reflect in the game, that is, I can be sure the role I intend to play is gonna match with how it will actually play out, at the benefit of immersion.
If my super-detective can't be expected to reliably analyze a crime scene, then I'm not playing a super-detective, I'm playing a guy who's pretending to be a super-detective.

Of course, you can always (and should anyway) talk with your DM beforehand to lay out such expectations, but having an already established and agreed upon base for consistency makes everything easier and the expectations more reliable.

[Vahnavoi]

@Captain Cap: there's no point in building expectations on a ruleset without acceptance of a "mother" equivalent; an "already established and agreed upon base" doesn't spawn out of nothing, it requires people to accept authority of some source.

That's why using Mother-may-I as derogatory misses it's mark; if you aren't playing it with your local game master, you are playing it with writer of a game book, a panel of other players, or some other entity that fills the role of the "mother" and confirms "yes, these are the expectations you may use".

This doesn't mean what you said is wrong, it just means we don't ever have to mention Mother-may-I to discuss your actual points, namely, merits of having extensive foreknowledge to build your expectations on.

There, I'd tell you to take a peek at the difficulty thread, where one point under discussion is how much predictive planning is necessary for a functional game to begin with. Personally, I find many aspirations of the sort you refer to as "creating a good climber" to be build on sand; trying to codify (and then stick to!) enough expectations so that the idea in the player's head will be matched in actual play is frequently a harder problem than negotiating such a character at a table. (Which is why I, when I have any truly exotic or esoteric character concept in mind, play freeform to realize those concepts and just tell other people in plain English how they are supposed to work.)

To tie this back to the main topic, when it comes to games that use target numbers, the games where it's easiest to build a character so that it works as a player thinks it would, tend to be those with simplest and clearest system math; the target numbers may be floating (read: decided on the spot by a game master) but they are bounded in a way that makes counting a range of odds easy in one's head or at least with minimal use of a pocket calculator. For the same reason, these games often require the least amount of expectations to play; understanding what you're supposed to do takes 5 minutes of back-and-forth with a game master, even if you've never read the game rules yourself.

[Captain Cap]

@Captain Cap: there's no point in building expectations on a ruleset without acceptance of a "mother" equivalent; an "already established and agreed upon base" doesn't spawn out of nothing, it requires people to accept authority of some source.

But they get to accept or not the authority after already knowing what the answers to "may I" are gonna be, at least relatively to that "mother". I imagine that's the difference prompting the use of the expression.

I find many aspirations of the sort you refer to as "creating a good climber" to be build on sand

It depends on the ground you're building it on, though. In 5e, yes, it would be mostly sand, while other systems with more guidance could provide a firmer ground.
On the other, if instead I want to create a "good fighter", 5e is as as good as concrete.

trying to codify (and then stick to!) enough expectations so that the idea in the player's head will be matched in actual play is frequently a harder problem than negotiating such a character at a table.

That's right, it is harder, that's why such approach would require a stronger leg to stand on than case by case rulings.

To tie this back to the main topic, when it comes to games that use target numbers, the games where it's easiest to build a character so that it works as a player thinks it would, tend to be those with simplest and clearest system math; the target numbers may be floating (read: decided on the spot by a game master) but they are bounded in a way that makes counting a range of odds easy in one's head or at least with minimal use of a pocket calculator.

That makes sense. The less the moving parts, the less a GM has to adjudicate each time.

[kyoryu]

I feel like there's three aspects of "Mother May I" that are useful to tease apart.

1. When difficulties are given
2. How difficulties are given
3. What difficulties are given

The first two, to me, are mostly a matter of perception. The info is coming from the GM. Whether the GM says "DC 15" when you ask the difficulty, or whether they mark down "solid stone wall" doesn't change that. So there's every bit as much of permission either way. Of course, the first one is rather exhausting for the GM, as they must write down, in some way, the difficulty (or the factors leading to the difficulty) for every single object in ever single scene.

The third one is, perhaps, the interesting one. I can see two possible negative GM behaviors that could lead to it being an issue.

The first one is just the GM increasing difficulties for things they don't want the players to do. This is just bad GMing. Note that for the most part, it can still be accomplished even if the information is given up front.

And, honestly, it's just bad GMing, and so the "don't play with bad people" advice comes to mind.

The second is more interesting, I think. And perhaps relevant in a 3.x-focused forum. This is the idea that there's no standard, and so the GM sets up "fair" difficulties based on what they think the difficulty should be. But, those are also based on the skills that they see. So if the best climber has a climb of 5, they'd make the difficulty 15 to get a 50/50 chance. But if the best climber had a 10 skill, the difficulty would be 20.

This completely negates any skill investment. And it's a problem in any game that really allows you to spend arbitrary points to increase some skills.

I don't see this in Fate, because Fate doesn't allow for arbitrary skill buys. You get your "pyramid" and that's it, so as a GM I know what skill calibration is.

The other solution is to make sure there are well communicated standards for what ratings should be used - this can be by the actual in-world description of the thing (the infamous oak tree). I think "target challenges by level" is also reasonable in this case, provided the GM also understands that this should be when tackling level-appropriate challenges.

I find that second case more interesting because it can be a bad result when everybody is actually operating in good faith. But I think that a reasonable calibration of expected difficulties is, generally, enough to preserve the value of skill investment.

It's about expectations: if I know beforehand (as it is already laid out in an accessible ruleset) what it takes to be a good climber in the game, what I can expect a good climber to do, how much of a good climber a character can be, then I can create a good climber character whose skill is gonna reflect in the game, that is, I can be sure the role I intend to play is gonna match with how it will actually play out, at the benefit of immersion.
If my super-detective can't be expected to reliably analyze a crime scene, then I'm not playing a super-detective, I'm playing a guy who's pretending to be a super-detective.

This gets into the calibration issue I mentioned. In Fate, the climber is going to be "take a high Athletics and an appropriate stunt, maybe toss in an aspect" and you're done. If there's any lack of clarity, talk about your expectations in Session Zero so everyone is on the same page.

Yes, there's a conversation there, but in my experience (again, with the right game support), I find that to be more useful than trying to encode every possible specialty or skill in a set of game rules.

[PhoenixPhyre]

It depends on the ground you're building it on, though. In 5e, yes, it would be mostly sand, while other systems with more guidance could provide a firmer ground.
On the other, if instead I want to create a "good fighter", 5e is as as good as concrete.

Except...yeah. Either I'm fundamentally misunderstanding what people mean by "building a good climber", or people are expecting way more than the system promises. 5e is built around the idea that you can't, generally, go "off the d20" and assure success. On anything. So if you want to build someone who can always climb anything...well...you can't do it via the ability check system[1]. Because the system says "no, that's not something playable (at least not that way)." That's asking why your pitchfork doesn't work well as a soup spoon. So asking for absolute skill (being able to hit some arbitrary benchmark for "can make climbing checks X% of the time), that's not something the system even allows for[2].

But if what you want is to be good at climbing, relative to other ability checks, the way forward is trivial. High Strength and/or expertise in Athletics. Take thief rogue to level 3 or 10+ for extra bonuses (being able to climb at normal speed, not half speed at level 3, getting to always take a minimum of 10 on every proficient check at level 10+). Or...learn spiderclimb.

So if you truly want to build a guy who is the best at climbing, two possible builds are:
* Human Barbarian X, max STR, take Skill Expert: Athletics at level 1. Now while raging you have (at level 1): +7 and advantage on checks made to climb. At level 20, you have +19 and advantage on checks made to climb. So if the DC is below 21, you succeed. And can routinely (50% of the time, not even counting rage's advantage) succeed on DC 30 checks, the hardest ones the game allows for (and which are supposed to be extremely rare).
* Any race thief rogue X, don't dump STR (say start at +2), take expertise in Athletics early on. Numbers are slightly lower, but you're faster at climbing past level 3 AND post level 10 your floor is tons higher. At level 20, assuming you don't actually increase your strength at all, you're looking at a floor of a 24 on the check. And you're above 20 minimum from level 10 on. And that's true even if your STR is +0.

So yeah. Even without magic, it's fairly trivial to make a character that can effortlessly succeed on any reasonable climb check at higher levels. In fact, all you have to do is take the obvious options. No splat diving, no obscure items, no obscure races. Choose things that obviously make you better at climbing and go.

[1] spells can give some options here. And you can push beyond DC 20 fairly easily (as shown in the examples).
[2] not to mention that climbing is a horrible example, because the rules for that are actually very clear. Climbing checks are the exception, not the rule. Anyone can climb most surfaces, no check needed, just half speed. Checks are only for exceptionally smooth/hard to climb surfaces. Compare that to 3e where your average fighter couldn't climb the easy wall at the climbing gym most of the time and only specialists could climb the hard wall. In 5e, since there are handholds, the stated default in the PHB is "yeah, you move at half speed. Climb away."

[GloatingSwine]

I've been seeing this terminology used on these forums for years without ever realizing Mother-may-I is an actual game. And knowing that an looking up the rules, you're entirely correct. What's being derogatorily called Mother-May-I is nothing like the game.

What, language shifts and subcultures develop their own meanings for phrases? Say it ain't so!

When people on this forum uses the phrase "mother-may-I" what they're generally referring to is nothing to do with the game of the same name but a lack of alignment in expectations of what will be possible outside the bounds of the printed rules of an RPG system, leading the players to have to guess at what the GM is and is not going to allow them to attempt.

To take the "being a good climber" example, the system can tell you what a good climber is, it can't tell you what you're allowed to try and climb. If you and the GM have sufficiently different expectations for that you may find yourself playing mother-may-I.

[Tanarii]

That ruleset is a bit odd because it is different in one important respect to one classic version of the game - this is not unusual, children's games have numerous common variants under the same name. However, that rule difference explains the disconnect. Let me quote Wikipedia on the same topic and you'll see:

"One player plays the "mother", "father" or "captain". The other players are the "children" or "crewmembers". To begin the game, the mother or father stands at one end of a room and turns around facing away, while all the children line up at the other end. The children take turns asking "Mother/Father, may I ____?" and makes a movement suggestion..."

Underlines for emphasis.

Okay, that makes the use of the phrase far more understandable.

As in the player states they want to undertake activity X, and then depend on the DM saying any of:
Yes
Yes if you can roll DC N
No, but ...

I agree with your later statement that this doesn't really address the real question of knowing the answer before determining if you will actually undertake the activity.

There are a ton of modern systems that have a rule like PbtA or MYZ engines:you do the thing you roll and if you roll you do the thing.

In the case of committing to the action and rolling being the same thing, knowing your rough probabilities of results in advance (which I'd say you usually do with both PbtA and MYZ) it makes decision making less scary. So to speak.

[Quertus]

@Captain Cap: there's no point in building expectations on a ruleset without acceptance of a "mother" equivalent; an "already established and agreed upon base" doesn't spawn out of nothing, it requires people to accept authority of some source.

That's why using Mother-may-I as derogatory misses it's mark; if you aren't playing it with your local game master, you are playing it with writer of a game book, a panel of other players, or some other entity that fills the role of the "mother" and confirms "yes, these are the expectations you may use".

This doesn't mean what you said is wrong, it just means we don't ever have to mention Mother-may-I to discuss your actual points, namely, merits of having extensive foreknowledge to build your expectations on.

There, I'd tell you to take a peek at the difficulty thread, where one point under discussion is how much predictive planning is necessary for a functional game to begin with. Personally, I find many aspirations of the sort you refer to as "creating a good climber" to be build on sand; trying to codify (and then stick to!) enough expectations so that the idea in the player's head will be matched in actual play is frequently a harder problem than negotiating such a character at a table. (Which is why I, when I have any truly exotic or esoteric character concept in mind, play freeform to realize those concepts and just tell other people in plain English how they are supposed to work.)

I mean, in chess, you sit down to the game knowing the rules, and can take your actions knowing that they will work; in Mother May I, the GM makes up the rules when you ask if you can do something. That’s not sounding like misusing the concept to me to use it to contrast a game where you don’t know the outcome of an action until you ask the GM to one where you do. Football would be a very different game if each umpire could make up how many points each touchdown / field goal / touchback / etc were worth - or even what ways there were to score points in the first place, how many downs to make what yardage, etc. Arguably, “Mother May I” has too many rules, and it’s not actually strong enough to describe the problem at hand.

It’s also related to the concept of public vs hidden information. I mean, I like Discovery, I like exploring the unknown. But, when dealing with the known, I expect to be able to make reasonable predictions about what the character can do, rather than be surprised when the overweight programmer GM declares that climbing a tree or standing on one foot are “hard”, while hacking someone’s phone is “easy”.

Put another way, there’s things that should be left to the passion and vision of an individual, and things that should be resolved in committee, tested and revised over time. “Physics”, how reality works, is in the latter category, at least where humans are concerned.

So, to the topic, common issues with DC-based skill systems, here’s the problems I remember seeing:

1) insufficient defined content / DCs / modifiers to be able to play through common “known” content without asking the GM “Mother May I” to determine DC.

2) insufficient range of DC / modifier to accurately model differences between characters. If the system has both “brain surgeon” and “middle school dropout” or “illiterate Barbarian” and “trained her mind to alter the fabric of reality” as viable characters, then there should be no comparison, and auto-success vs auto-failure should happen, not just in contrasts between them, but between any of them and Average Joe.

3) limited to producing pass/fail results.

3b) or other options poorly or inconsistently implemented

3c) often by designers who fail to realize that this is a good place to allow players to define their characters (by the costs they’re willing to pay, by the costs they pay up front vs gamble, etc).

4) emphasis on rolling / chance.

5) incoherent gatekeeping (“you’ve trained your mind to alter the fabric of reality, but can’t have more than a basic grasp of mathematics” or “in the same amount of time it took to learn to bend reality, your brother can’t have learned to balance on one foot”).

6) lack of realism / verisimilitude // produces an incoherent reality if taken to its logical conclusion.

7) no ability to handle alternate resolutions. For example, the best skill checks in 3e modules IME are the multi-DC Gather Information checks… yet they make no mention of what a Historian with Knowledge: History, or a rumor monger with Knowledge: Local might already know (not that I’m a fan of Knowledge skills).

8) bad pricing on skills.

9) fail at Character Creation.

10) poor choice of skills for skill list - skills that are too general (“magic”, “roll combat”), or too specific and incomplete (“none of these 200 skills cover what my character concept can do”), or even both in the same system.

[Telok]

8) bad pricing on skills.

9) fail at Character Creation.

Yeah, that's an amusing set. Like I've got a D&D 5e character who I chose skills/profs for at first level based on background and character stuff. Because the difference in play of +0 vs +2 or +3 vs +5 is super minor. Turns out at 15th level I now consider having to roll any skill/prof a failure. Character has +11 deceive, which las literally never come up in game, and -1 to +6 otherwise. Since all checks rolled are vs 15+ (GM is fairly good at not requiring rolls for silly trivial stuff but things like "know your own religion basics" but all "know/decipher magic" and "find info" type stuff is a check unless your character has cast the spell) any action calling for a roll is assumed to fail unless magic gets added or replaces it. The GM has ended up encouraging us to dogpile checks and get magic items/spells to skip checks just to compensate for our tendency to fail anything outside of hurting/murdering stuff no matter what our +number on the character sheets are.

Compare with several dice pool systems or Classic Traveller where high (not even max) ability stat or basic training put you over the default median targets. Even in games where the reasoning behind what to roll for is identical to D&D 5e having both max ability plus maximum skill means characters do 100% expect to out score a no bonus & no training character in challenging and difficult circumstances.

Then the whole pricing thing, yeah. Traveller is harsh, four years to learn the basics of, say, xenobiology. But that gets you from no roll or 2d6-2 vs 8+ to 2d6+1 vs 8+ and a possible bonus for high int/edu. So from a 17% chance to a 58%+ chance on the average "DC" 8 check. Although to be fair Traveller does give you 3 to 4 weeks of downtime for every hyperspace jump and you can do this concurrent with adventuring. Most other games make it easier to pick up new skills at the basic "and now I can do X" level. D&D has had it pretty weird over the years but we're at a level now of recommendations to take the "happy go lucky magic lute player" or "nimble stabby law breaker" classes for a warrior to get significantly better at swimming & climbing or for the religious clasess to get better at knowing religious stuff. Granted, those will get the character from (usually about & assuming dediacring character advancement to it at 12th-16th level) +5 to +10 or +10 to +15, which is significant but hella costly given the game's character advancement structure.

[Vahnavoi]

It depends on the ground you're building it on, though. In 5e, yes, it would be mostly sand, while other systems with more guidance could provide a firmer ground.

I don't play 5th edition Dungeons & Dragons and wasn't thinking of it when I made my statement. If I was thinking of an edition of D&D at all, it would've instead been the 3rd, and other games more complex than 5th edition.

At this point, you're likely confused, because such complex games are often what people who complain about 5th say they want to play. To clear this confusion, I suggest you look at my earlier multi-point list of flaws I've encountered in target numbers.

Have you done that? Good. Simply put, systems that try to be really comprehensive often have more of these flaws than simpler systems. The extra rules serve to obfuscate what would be a good idea, they are better at building false expectations than actually being predictive. Which, in practice, means that a character build around expectation of, say, being a "good climber", might be anything but.

In such games, conceptualizing a character as a "good climber" before play starts is a fool's game. They are, instead, just a "climber" and how good they are is best answered empirically.

That's right, it is harder, that's why such approach would require a stronger leg to stand on than case by case rulings.

What I wished to communicate is that making your "stronger leg" is a harder problem than making case by case rulings. See also: the "sweet spot for success" thread, especially Telok's posts.

---

I mean, in chess, you sit down to the game knowing the rules, and can take your actions knowing that they will work. In Mother May I, the GM makes up the rules when you ask if you can do something.

Chess is a complete deterministic rules game made to work without a referee, while Mother-may-I and most tabletop roleplaying games are incomplete indeterminist rules games reliant on a referee. In other words, the gap between Chess and most tabletop roleplaying games is already bigger than the gap between most tabletop roleplaying games and Mother-may-I.

It's possible to make a roleplaying game closer to Chess, but you might want to take a minute to consider why most tabletop roleplaying games aren't like that.

That’s not sounding like misusing the concept to me to use it to contrast a game where you don’t know the outcome of an action until you ask the GM to one where you do. Football would be a very different game if each umpire could make up how many points each touchdown / field goal / touchback / etc were worth - or even what ways there were to score points in the first place, how many downs to make what yardage, etc.

Every game of D&D is different because a game master has to select or make a scenario - equivalent to an umpire choosing how many goals there are on the field, where they are and how big the playing field is. Same applies to most other tabletop roleplaying games.

Point being, these comparisons don't really get to the point why target numbers are the place where a player ought to put their foot down. Quite often, they give the appearance of sifting a fly out of your soup but swallowing a camel.

Arguably, “Mother May I” has too many rules, and it’s not actually strong enough to describe the problem at hand.

As noted, the game doesn't ever need to be mentioned to discuss the actual points. It was used in an internet essay once and people latched on to it. It's up there with TV Tropes using mangled proverbs and puns as name for tropes that don't have a thing to do with the proverbs (they have a list, but you can't find it under Non-Indicative Names).

If the issue is that you don't want to petition a game master for common information, it's possible to say just that.

(Yes, GloatingSwine, language shifts and hobbyists come up with special meanings for old words. Sometimes, language shifts for stupid reasons such as hobbyists pointlessly coming up with special meanings for old words. It's normal, but only in the same sense as it's normal for food to spoil if you leave it in your fridge for too long. No-one has to be happy about it nor pretent nothing can be done about it. )

[LibraryOgre]

a) forgetting that requiring a series of multiple successes dramatically worsens the odds for any task, and

Related to this, something I did last night with my Eldest was to have a roll that helped the actual roll, instead of was necessary. His character was telling a story (about how the local powerful family had ritually murdered her dad), and so I gave him two rolls: If she succeeded at her Performance test, she got advantage on her Persuasion test. Failure on her Performance meant a regular roll on Persuasion; a natural 1 on Performance meant disadvantage on Persuasion. He rolled a 1 on Performance... then a 16 and an 18 on Persuasion.

c) Relatedly, accidentally gating all non-specialist characters out of participating by setting the DCs to be a reasonable challenge for the party expert (especially in games with constant scaling/no bounded accuracy). This is fine for tasks like developing a cure for the plague virus, but terrible for many social and exploration tasks where other PCs should reasonably be able to participate.

This is a big problem, though, as you mention, it's more of a problem when you don't have some form of limitation on skill gaps. 5e does bounded accuracy, which does well. Savage Worlds has skill caps... it's rare to have above a d12 in a skill, and some skills have a floor of d4. It keeps expectations of skill levels more in line than 3.x games, where DCs might scale with the player levels, soon outstripping anyone who didn't specialize.

[King of Nowhere]

Relatedly, accidentally gating all non-specialist characters out of participating by setting the DCs to be a reasonable challenge for the party expert (especially in games with constant scaling/no bounded accuracy). This is fine for tasks like developing a cure for the plague virus[emphasis mine], but terrible for many social and exploration tasks where other PCs should reasonably be able to participate.

This is a big problem, though, as you mention, it's more of a problem when you don't have some form of limitation on skill gaps. 5e does bounded accuracy, which does well. Savage Worlds has skill caps... it's rare to have above a d12 in a skill, and some skills have a floor of d4. It keeps expectations of skill levels more in line than 3.x games, where DCs might scale with the player levels, soon outstripping anyone who didn't specialize.

I see a greater problem with games that don't have this skill gap. It is only sensible that developing a cure for the plague can only be attempted by the specialist. A system that caps skill check basically has no task that's reserved for specialists. basically, the party barbarian can walk into a lab, sneeze into a test tube, and develop the plague cure that the scientist has been seeking for months. I hate 5e for that.
Yes, high dc checks could be abused by poor dming, setting every dc at 30+ for anything that the characters ever try to do, but that's not a good reason to remove a feature just because some people misuse them. it would be like forbidding kitchen knives because they can be used to stab people. sky-high skill bonuses and skill dc have their purposes, removing them entirely is bad.
and this kind of premise - remove a mechanic because you don't trust the dm to handle it correctly - is a terrible premise for rpgs. the dm has the power to screw up your character regardless, intentiojally, or by accident. you have to trust your gm in order to play. else, you may as well avoid any combat encounter entirely.

[Pex]

I see a greater problem with games that don't have this skill gap. It is only sensible that developing a cure for the plague can only be attempted by the specialist. A system that caps skill check basically has no task that's reserved for specialists. basically, the party barbarian can walk into a lab, sneeze into a test tube, and develop the plague cure that the scientist has been seeking for months. I hate 5e for that.
Yes, high dc checks could be abused by poor dming, setting every dc at 30+ for anything that the characters ever try to do, but that's not a good reason to remove a feature just because some people misuse them. it would be like forbidding kitchen knives because they can be used to stab people. sky-high skill bonuses and skill dc have their purposes, removing them entirely is bad.
and this kind of premise - remove a mechanic because you don't trust the dm to handle it correctly - is a terrible premise for rpgs. the dm has the power to screw up your character regardless, intentiojally, or by accident. you have to trust your gm in order to play. else, you may as well avoid any combat encounter entirely.

An alternative to high DC numbers is to have a separate set of DCs that only characters trained or the appropriate specialized level can do. For example, your barbarian can make a DC 15 Medicine check to do CPR and bring an unconscious 0 hit point character to conscious at 1d4 hit points, but only the Specialist Healer who invested a character building resource cost into Doctor of Medicine on the character sheet can make a DC 15 Medicine check to cure a disease.

This should be words on paper part of the rules of the game system skill use not DM fiat so that players can make an informed decision on where to allocate their Specialized Skill Use resource expenditures.

[Tanarii]

Off the top of my head I'd rather see the advanced skill be a different scaling check than just an additional use of the same check.

Otherwise you still need to worry about bonuses that push the check off the die vs small bonuses for untrained for normal stuff.

Of course, another approach is to ditch the paradigm of one die roll + bonus compared to TN entirely. Each comes with its benefits and drawbacks of course. Percentage (so the range of PC skill is 0-100%) is one method that's been used extensively. Or bell curves (2d6 or 3d6) with very limited bonuses are another. Dice Pools (/ugh ).

[Quertus]

An alternative to high DC numbers is to have a separate set of DCs that only characters trained or the appropriate specialized level can do. For example, your barbarian can make a DC 15 Medicine check to do CPR and bring an unconscious 0 hit point character to conscious at 1d4 hit points, but only the Specialist Healer who invested a character building resource cost into Doctor of Medicine on the character sheet can make a DC 15 Medicine check to cure a disease.

This should be words on paper part of the rules of the game system skill use not DM fiat so that players can make an informed decision on where to allocate their Specialized Skill Use resource expenditures.

Huh. I like skills and DCs that far exceed the die range, but, from a Simulationist perspective, I can see how what you said makes sense. To put it in 3e parlance, I took the feat that lets me use the Computer skill to easily write code others can't; some people took the Layout feat to let them use the Computer skill to easily create good interfaces that others can't; yet other people the Hardware feat to let them use the Computer skill to easily fix physical problems others can't.

I guess there's more than one way to gatekeep, and to identify your specialists.