The 5th Edition Skills/DC Reference they should have included

[Steerpike]

I stumbled across this file on a D&D Tumblr. It's a very good reference document with various DC suggestions for skills, suggestions for ability score contests, and other clarifications. For example, it suggests that while Deception should usually be opposed by Wisdom, Persuasion and Intimidation should generally use a static DC (with examples of difficulty).

It's not totally exhaustive - it could be longer and more detailed - but this is the sort of reference sheet I think they should have included with the 5th edition PHB, and goes a long way to helping quick DC-assignation.

[sparkletwist]

It's helpful, but I don't really understand and don't particularly like some of their methodology. For example, why is Deception an opposed check but Persuasion and Intimidation have a static DC? The static DC is based on an assessment of the target (a "spineless noble" is DC 10, "street thugs" or "highway thieves" are DC 15, and "a creature larger than you" is DC 25) so there probably could and should be some way to quantify that... or conversely, why can't Deception work the same way? Making some rolls opposed and some unopposed means that essentially some checks are based on 1d20 and some are based on 2d20, which is mathematically ugly.

[Steerpike]

I've sometimes used opposed checks for Persuasion and Intimidation, but I can sort of see the logic here.

For Deception, success is based on how well the target sees through your lies, which is determined essentially by their skill as a lie-detector (observation, judgment, etc) and your skill as a liar (masking your intentions, talking loquaciously, etc). These things are relatively easy to quantify as skills, hence a contest is useful.

In the case of Persuasion and Intimidation, context and circumstance and social status play a much larger role than individual skill when determining "difficulty," and, sometimes, you're dealing with larger groups where individual rolls would become more awkward. So, for example, for "Convince a chamberlain to let your party see the king," the difficulty doesn't lie primarily in the personal characteristics of the chamberlain per se, or even his social skills exactly, but in the situation and social ranks and societal contexts of those involved, which don't track very well to Charisma, level, or proficiency. To put it another way: a more or less charismatic chamberlain would not necessarily be easier to harder to convince of something, because Charisma doesn't equal stubbornness, and because it's the chamberlain's rank and job more than their personal attributes that make the situation difficult. To expand even further, it might make as much sense for a very skilled chamberlain who is good at their job to actually be convinced by the request as it would be for a chamberlain who is bad at their job, so it doesn't make sense to assign difficulty based on the chamberlain's "skill," as a contest would necessitate. Indeed, to extend this logic, it just doesn't really make sense that the more skilled or charismatic (or wise) a person is, the less easily they can be persuaded. Maybe the persuader is making a really good, rational argument, for instance, and an open-minded person should be persuaded. Kind of ditto with intimidation: sometimes a skilled or powerful individual will recognize the "truth" of the intimidator's threats, so it doesn't make sense to "penalize" them for their skill in a situation where they reasonably should back down.

Similarly, trying to "stop an agitated mob in their tracks" either necessitates opposed rolls for each mob-member or essentially just assigning the mob as a whole a check, which is getting close to just assigning them a DC anyway.

There are some examples they give where I would still be tempted to do an opposed roll, like the interrogation I'd be tempted to all a Wisdom saving throw for the prisoner to represent their willpower.

I admit to not seeing the big deal about the greater variability of a contest versus a flat DC vis a vis mathematical elegance, but then again, that's never mattered to me, so it makes sense I wouldn't see that!

[sparkletwist]

In the case of Persuasion and Intimidation, context and circumstance and social status play a much larger role than individual skill when determining "difficulty," and, sometimes, you're dealing with larger groups where individual rolls would become more awkward. So, for example, for "Convince a chamberlain to let your party see the king," the difficulty doesn't lie primarily in the personal characteristics of the chamberlain per se, or even his social skills exactly, but in the situation and social ranks and societal contexts of those involved, which don't track very well to Charisma, level, or proficiency.

This is actually a problem I've had in Fate, too, due to the structure used in Fate of abstracting every social contest as opposed social skills that inflict stress. In the case of Deception (your lies vs. their intuition) or Intimidation (your meanness vs. their willpower) it works pretty well, and it can work for certain persuasive arguments, as well-- your persuasive ability vs. some knowledge skill to represent trying to make an argument to some intellectual, or something like that, but, yeah, honestly, sometimes it does get a bit tricky and break down a bit, so this isn't solely an issue with 5e. The part of me that likes elegant and extensible game design would like to see something other than "just pull an arbitrary DC out of the air that seems to fit the context," but I'll have to admit that I do that from time to time when I'm running Fate, too...

Maybe the persuader is making a really good, rational argument, for instance, and an open-minded person should be persuaded. Kind of ditto with intimidation: sometimes a skilled or powerful individual will recognize the "truth" of the intimidator's threats, so it doesn't make sense to "penalize" them for their skill in a situation where they reasonably should back down.

That's true, but, in cases like this, I often don't even roll at all. The player usually roleplays the speech well, so I just let them have it.

I admit to not seeing the big deal about the greater variability of a contest versus a flat DC vis a vis mathematical elegance, but then again, that's never mattered to me, so it makes sense I wouldn't see that!

The issue is that 1d20 is a flat probability distribution while 2d20 is a curve, which actually changes things. Things are pretty much the same near the middle-- rolling 1d20 vs DC 10 has 55% chance of success while opposed 1d20's has a 52.5% chance of success (assuming ties go to the first 1d20, just like hitting a DC exactly counts as a success) but things change considerably as you start getting to the edges of the curve. 1d20 vs DC 20 has only a 5% chance, while 1d20 vs 1d20+10 has a 13.75% chance; you have no chance of hitting DC 25 on a roll of 1d20 but actually do have a 3.75% chance of beating 1d20+15 with a roll of 1d20. You sort of do have to take this into account when you're the DM setting difficulties or you won't be able to estimate the players' chance of success at a task at all... and if you're not going to do that, why even set a difficulty at all, since it doesn't matter anyway?

[Steerpike]

That's true, but, in cases like this, I often don't even roll at all. The player usually roleplays the speech well, so I just let them have it.

Yeah, in practice for me the social skills DCs are probably the least useful part of the reference sheet.

The issue is that 1d20 is a flat probability distribution while 2d20 is a curve, which actually changes things. Things are pretty much the same near the middle-- rolling 1d20 vs DC 10 has 55% chance of success while opposed 1d20's has a 52.5% chance of success (assuming ties go to the first 1d20, just like hitting a DC exactly counts as a success) but things change considerably as you start getting to the edges of the curve. 1d20 vs DC 20 has only a 5% chance, while 1d20 vs 1d20+10 has a 13.75% chance; you have no chance of hitting DC 25 on a roll of 1d20 but actually do have a 3.75% chance of beating 1d20+15 with a roll of 1d20. You sort of do have to take this into account when you're the DM setting difficulties or you won't be able to estimate the players' chance of success at a task at all... and if you're not going to do that, why even set a difficulty at all, since it doesn't matter anyway?

I guess what I'm saying is, what's wrong with having skills that have somewhat variable difficulties, especially when in practice any DC over 20 is very rare, and few characters have bonuses bigger than about +5 or +6?

EDIT: Like, sure I get that differences exist, but I don't feel like DC 25 rolls or +15 skill contests come up enough to make it too worrisome, personally. Doesn't this logic sort of apply to all contests versus static DCs?

I guess you could argue that it slightly statistically "over-values" the skills that mostly use contests rather than static DCs... but the alternative is either making everything a contest or nothing a contest, right? Or, alternatively, spelling out intricate rules for when to use a contest and when to use a static DC (which feels a bit excessive); or leaving it vague (which is what the PHB does, but the reference sheet corrects).

If you really wanted to maintain strict mathematical equality you could scrap static DCs entirely and just give every obstacle a "bonus" of + or - the amount it's over 10, which could totally work, but means extra rolling and arithmetic.

[sparkletwist]

If you really wanted to maintain strict mathematical equality you could scrap static DCs entirely and just give every obstacle a "bonus" of + or - the amount it's over 10, which could totally work, but means extra rolling and arithmetic.

Or make everything a roll against a DC and do away with opposed rolls entirely, which is kind of the direction D&D goes anyway. There is even some semblance of this approach in the system as written; battle master maneuvers use a save rather than an opposed roll, for example.

(Your way works too, and it's what Asura does, but the other way feels more "traditional D&D")

[Steerpike]

I think both are fine for those craving mathematical uniformity. It's a war in my mind between the drama and uncertainty of contests as opposed to a static rolls, versus the speed and ease of rolling as few dice as possible. In practice I don't mind the loosey-goosey "mixed method" where there are rolled contests when it seems to make sense that two people are pitting their skills against one another, and static rolls when it feels like they're up against an unchanging obstacle. Of all the static numbers I've been tempted to turn into contests, AC is the most enticing, but would also add the most time to combat.

This is kind of why I like 5th edition, even if it's the same reason it probably annoys you: I could imagine making any of these alterations without accidentally introducing game-breaking/balance-destroying difficulties, because the rules are imperfect and deliberately squishy to begin with.

[sparkletwist]

I could imagine making any of these alterations without accidentally introducing game-breaking/balance-destroying difficulties, because the rules are imperfect and deliberately squishy to begin with.

I think it's a false dichotomy to assume that games with a more sound mathematical structure behind them are also always more brittle when it comes to tweaking. They can be, but this doesn't always have to be the case. Something like Fate has a much more cohesive resolution mechanic but it's also pretty amenable to various sorts of tweaking the numbers, mostly because the underlying mechanic of the fate point economy lets the players and the GM "fix" any numerical aberrations in play so everything still comes out in a way that produces a good session and a good story. And, don't get me wrong, you can get a good session and a good story in 5th edition, too, but it generally involves roleplay things that have basically nothing to do with the "squishy" rules in the rulebook. I still don't see the imperfection of the 5th edition rules as anything but a net minus in practice.

[Steerpike]

Yeah, to it's credit, Fate is very good at being hacked, and openly invites it. But I don't think Pathfinder or 4th handle it nearly as well. Obviously there are plenty of roleplaying games apart from D&D, but of the various editions of D&D itself or its close offshoots/retroclones, I feel like 5th comes the closest for me of being robust/detailed enough to be fairly satisfyingly playable out of the box - in a way that OD&D/AD&D and B/X kind of aren't and which I feel 3.X and 4th are pretty good at - while also being very easily hacked and house-ruled - which OD&D through to AD&D 2nd sort of demand, but which 3rd and 4th kinda fail at.

It feels actually a little bit like 2nd edition AD&D to me, which I briefly played when I got started, but without the weirdness of THAC0 or level-caps or dual-classing or speed ratings, and with better class/race/spell options.

[sparkletwist]

Yeah, to it's credit
[box_white]to it's credit [box_white]it's credit[/box_white][/box_white]

:stick:

[Steerpike]

Aw man I never make that mistake either!!