Since we have a couple of threads discussing mechanic preferences I'm wondering is there any appreciable difference between difficulty checks, which I believe were introduced in 3e D&D and carried over into 5e versus ability checks (roll under the ability score) in 2e and earlier versions of D&D.

Not sure when I discovered ability checks, I think they always existed as an alternate rule from the very early days of the game, but once I did I thought they were very elegant and they largely replaced saving throws at our table.

So what say you? Is there any mathematical difference between the methods or is it a tomato/tomatoe thing? Is the advantage of DCs simply that they eliminate the need for bonuses and penalties to the roll or that they are a unified mechanic with attack rolls, etc?

There are various differences, but essentially they boil down to "Roll Dice vs. Your Talent to Decide Shit"

Roll under Ability Score was a general 0e-2e houserule, but some people used D20, others 3D6. It was the D&D answer to the RuneQuest roll under Stat x 5% and amounts to the same math.

2e had proficiencies (first introduced in 1e's Oriental Adventures) where you had a skill that you rolled under a number based on an ability score.

In recent years, Castles & Crusades created the SIEGE engine where you made a combo skill / ability check rolling high vs. TNs and it was much akin to D20's DC checks.

Can effect campaign flavour.

In a game with low skill emphasis, where you want assumed competency by PC's and that PC's are cut above the norm, with some being incredible then ability checks vs skill checks support that style of play. Suits Swords & Sorcery imo. Also suits slightly more generous ability score generation systems (e.g.. reroll all ones, swap any 2, refill if no positive modifier or total modifiers below 0). I see them as slightly more abstract as they ignore the level of the challenge & focus on the talent of the PC.

In a game where you have placed an emphasis on a complex or involved skill system, ability checks make less sense. If you're going to go up to level 20 and try to make skills relevant most of that way then ability checks perhaps make less sense. But a difficulty check you can scale begins to make more sense if you want pc's to face "epic" challenges. Emphasis becomes on having 6 different levels of locks, intimate situations, blah, blah and grading them according to difficulty.

I prefer the former.

I never experienced the ability check, but as I understand it, the PC's odds of overcoming an issue with a roll stayed mostly flat and predictable across the levels, which sounds nice for a GM trying to run a consistent world. For instance, anywhere the party goes, they are likely to face doors that are reliably going to present a certain level of challenge to lock-pick or demolish, with the occasional easy or extra hard one.

The trouble with difficulty checks is they are usually paired up with a skill system where the PC keeps getting better at doing something... across 20 levels. If the GM is not careful, certain types of verisimilitudinous challenges become too trivial, and they might be tempted to gradually increase the DCs in an arms race with the PC's capabilities, ultimately making their world less realistic and the skill increases pointless.

DM: And you find yourself in front of an adamantine door with locks designed by modrons.
Rogue: Damn, another one? Where were all these adamantine doors with locks designed by modrons three levels ago?
DM:Err, this is the richest part of town.
Rogue: Weird, I feel you sort of implied that about the other part of town we were operating in three levels ago. There the richest people had steel doors with locks designed by gnomes.
DM: Anyway, the DC to pick or break through this is X.
Rogue: *looks at character sheet*. You know, doing a little math, my odds of getting through this door will my supposedly ever increasing skill are roughly the same as the steel doors with locks designed by gnomes three levels ago... and the same again as my odds of getting through the reinforced wood doors with locks designed by human master locksmiths three levels before that.
DM: ... You know what, a patrol of fire salamander sentinels summoned from another plane to guard this place attack you while you're contemplating the door. Roll for initiative. Watch out, these are tougher to face than the ogre sentinels hired from the deadly badlands you were facing in the other part of town three levels ago...

I'm curious to hear people's thoughts on this. When I created my game I figured a static target number for skill rolls would be good enough. Now it feels like there needs to be more differentiation of target numbers. My go-to idea is to correlate target numbers with challenge levels and experience points. Roll X or suffer Y consequence. Get Z exp if you manage to reach your objective, whether or not you suffer Y consequence.

That might be too 4th edition-y. I've seen people criticize 4th edition for adventures being little more than skill challenges connecting set piece battles. (If I understand.)

I'm not sure if the criticism has more to do with issues about railroading or reducing exploration to mechanics. It seems one can avoid those pitfalls and still create rules for target numbers. But I'd like to hear people's thought, if that doesn't detract from OP's query.

Well that's the general problem with the semi-conflicting goals of wanting:
* a consistent world
* PCs to have heroic abilities
* PCs to improve steadily with adventuring experience
* challenge and some chance of failure
* a campaign to last a long time with some PCs surviving and gaining a lot of experience

It's hard to satisfy all of those requirements, and having more than a couple of them at once require a fair amount of work.

Among the different methods of figuring out the odd of success there are ones is where skill level is very important and others  is where the character attributes dominate.

For example in d20 I get a +1 per 2 points of attribute above 10. So a 16 strength gets me a plus +4 bonus. My skill can easily be +4 or higher with a d20 character. Since I am adding the attribute bonus and skill bonus to my roll. So overall both are equally important to figuring out my odds of success.

In GURPS in constrast, the skill level is based one's attribute (average 10). The skill level (or lack of skill) Modifies the actual odds from this starting point. For example 4 points in swords skills will allow my skill to be my attribute +1. So if I have a 13 strength my skill level is 14. I have to roll a 14 or less on 3d6.

With D&D roll under means that attribute dominates. With the d20 approach 5e, Skill and Attributes are equally important.

What you need to decide is what important in your mind. This is definitely an area where there is no right answer. Just be consistent once you made your choice.

Quote from: Skarg;955959Well that's the general problem with the semi-conflicting goals of wanting:
* a consistent world
* PCs to have heroic abilities
* PCs to improve steadily with adventuring experience
* challenge and some chance of failure
* a campaign to last a long time with some PCs surviving and gaining a lot of experience

It's hard to satisfy all of those requirements, and having more than a couple of them at once require a fair amount of work.

What I'm trying:

(1) Parts of the world are getting stronger for reasons that make sense in game.

(2) The players can run away from almost anything.

(3) Evaluating difficulty is the easiest thing you can do.

The biggest difference is that roll-under-ability completely nukes the importance of class and level in favor of stats. That's something I don't like, since I like my D&D games to focus on class/level over stats.

A system using DCs can be better (since there's usually some class/level influence in whatever skill system is grafted on), but I'm not a fan of that in D&D, either. When I ran D&D with such systems, I always found myself backing into a DC/modifiers which gave the chance I was looking for. In other words, I'd think "Hmm, the PC should have about a 60% chance of success. What numbers do I need to assign to get to that?"  More trouble than its worth, IMO. These days, I save myself a lot of time by just coming up with the number and calling for the appropriate roll.

Quote from: Voros;955880Since we have a couple of threads discussing mechanic preferences I'm wondering is there any appreciable difference between difficulty checks, which I believe were introduced in 3e D&D and carried over into 5e versus ability checks (roll under the ability score) in 2e and earlier versions of D&D.

Not sure when I discovered ability checks, I think they always existed as an alternate rule from the very early days of the game, but once I did I thought they were very elegant and they largely replaced saving throws at our table.

So what say you? Is there any mathematical difference between the methods or is it a tomato/tomatoe thing? Is the advantage of DCs simply that they eliminate the need for bonuses and penalties to the roll or that they are a unified mechanic with attack rolls, etc?

I am very much of the mindset that things matter only when there's an actual, systematic effect. So if the percent chances tend to be the same using either system, then they are in my mind the same (so like to me, ascending and descending AC are the same thing so long as your chance to hit is the same).

The big issues that I can think of with ability checks are:
*In classic D&D game (oD&D-BECMI), the stat spread is 3-18 (hard cap). If you used a d20, roll under attribute check, that would constrain your chances between 15-90% success rate. What happens if you want something with a chance of success outside that range?
*If you don't use d20, but instead 3d6 or variable d6, a slight change in average stats can make unexpected changes to likelihoods.
*In AD&D, the stat spread is 3-25 (with some of the upper levels being very close to impossible to achieve). What spread method best captures this range? Further, what do you do with 18/xx strengths?

The issue Shipyard mentioned is the big problem I can identify with difficulty checks. Although... this is really just a microcosm of the whole sandbox vs. party-tailored challenge debate, isn't it?

Quote from: Skarg;955959Well that's the general problem with the semi-conflicting goals of wanting:
* a consistent world
* PCs to have heroic abilities
* PCs to improve steadily with adventuring experience
* challenge and some chance of failure
* a campaign to last a long time with some PCs surviving and gaining a lot of experience

It's hard to satisfy all of those requirements, and having more than a couple of them at once require a fair amount of work.

Yep, it clearly is the sandbox vs. party-tailored issue. Y'know, in my own campaigns, this has never really been the big deal that a million overwrought threads has made this out to be. My campaigns have been semi-sandbox, and the players self-select adventures of approximately the challenge level that they can handle (or at least think they can handle). 3rd level characters shouldn't attempt to break into the king of the world's largest nations vault, and 16th level characters won't bother robbing the alms box at the wayfairer's shrine in bumble-fuck-nowhere-town.

To make a mechanic, however, that scales by level only slowly, is a challenge. Of the WotC-era games, I consider 5e to have done it probably the best--yes a 20th level character can succeed on a roll with a target # above what a 1st level character can do, but they can still roll a 1 and do worse than the 1st level character can do with a roll of 20. That's something that I feel 3e lacked. For the TSR era games, I thought using the saving throw table worked best, perhaps modified by an attribute. slow advancement, there can be differences between characters (depending on class and attribute), very few people will have only 5% chance of failure or success. Overall pretty good.

I am really unclear why people are talking about ability-only checks vs. ability+skill checks, since it's not what the OP asked about.

Quote from: Voros;955880So what say you? Is there any mathematical difference between the methods or is it a tomato/tomatoe thing? Is the advantage of DCs simply that they eliminate the need for bonuses and penalties to the roll or that they are a unified mechanic with attack rolls, etc?

Mathematically they're identical. Whether you're modifying the target number or the die roll doesn't really make any difference. Nor does the difference between using difficulty as the target number and ability as the modifier vs. using ability as the target number and difficulty as the modifier.

In terms of utility, however, there are distinctions. But they tend to be mostly coincidental rather than intrinsic.

Knowing what number the die needs to roll for success before you roll, for example, is a nice feature. It means that when the player rolls a die, there can be an immediate and emotional response to seeing the die result. (As opposed to needing to perform mathematical operations on that die result before knowing what happens.) The other advantage is that it typically means that the players can resolve their checks by themselves and then simply report success or failure to the GM. Old school ability checks tend to fall into this category because the die roll needed = your ability score or lower.

Hidden difficulties, on the other hand, are a nice thing to have in a variety of circumstances. These are generally more awkward to implement when you're using difficulty as a modifier against a target number known to the players. This is particularly true if it's a check that multiple players are making at the same time.

I've also found, personally, that it seems to be easier for players to apply modifiers that are written down in front of them. In a game where the difficulty of tasks is highly variable, therefore, I generally prefer to have difficulty as the static target number and for the modifiers to be things written down on the character sheet.

Quote from: Shipyard Locked;955907The trouble with difficulty checks is they are usually paired up with a skill system where the PC keeps getting better at doing something... across 20 levels. If the GM is not careful, certain types of verisimilitudinous challenges become too trivial, and they might be tempted to gradually increase the DCs in an arms race with the PC's capabilities, ultimately making their world less realistic and the skill increases pointless.

See also shitty games like D&D 4E Gamma World where the system bakes the DC inflation into the system. So glad we're just adding numbers for no reason.

This is generally just bad GMing in general, though, and not limited to action checks. It's the same mindset that starts dumping antimagic areas everywhere to nerf the wizard.

Quote from: estar;955963With D&D roll under means that attribute dominates. With the d20 approach 5e, Skill and Attributes are equally important.

What you need to decide is what important in your mind. This is definitely an area where there is no right answer. Just be consistent once you made your choice.

In 5e its allmost situational. There are times when your Prof bonus is way exceeding your stat bonus. And the other way around as some situations wont allow a proficiency bonus but allow stat bonus. While others go the other way. The majority though indeed take into account both so stats and level.
Sometimes even class matters as one or two classes grant skill bonuses. The Bard and Rogue come to mind as both classes grant a doubled proficiency bonus to at least two skills.

Initially I wasnt overly fond of the difficulty number style. But over time have come to like it. It feels a bit more fair. Least it does in 5e.

Thanks Justin my initial question was about the math as I'm math retarded so didn't know if there was any significant statistical difference, although I suspected it was minor.

It doesn't translate exactly due to ability mods bumping up every 2 ability score points instead of every 1. An ability check when your score is 18 is a DC 7 check in the WOTC game. But if your score is 3, that's a DC 14 check. But overall it's pretty similar.

Quote from: Voros;956002Thanks Justin my initial question was about the math as I'm math retarded so didn't know if there was any significant statistical difference, although I suspected it was minor.

When doing an ability check in D&D think of the stats as percentage chance of success. Equating 1 point each to 5%. So someone with a 5 STR has the equivalent of a 20% chance of success. The trick here is wether or not the PC gets any bonuses or negatives to the roll based on circumstance. A +2 to climb a rope turns the 20% into a 30%. If that were a negative then the 20 becomes a 10% chance on a d20 roll.

With DCs and the like it changes the paramiters. But the overall percentages can still be worked out. Say the DC is a 12. That is equivalent to 60% chance of failure. Heres where skill and/or stat bonuses become important as each + will diminish the chance of failure. In 5e a level 5 character or a 16 in a stat is a +3 bonus. So if a check allows that to be added in then the 60% becomes a 45% chance of failure, or 30% if both are able to be added.

So one is a chance to succeed, the other is a chance to fail. Of course some systems approach it from other directions.

Least this is my view on it after puzzling it out.