D&D General How do you set the DC of an ability check

[Thommy H-H]

DC 15 works, like, 90% of the time.

 

[Charlaquin]

When an action is directly targeted at an NPC, I’ll either set a DC based on 8 + the NPC’s proficiency bonus + an appropriate ability modifier if it’s passively resisted, or call for an opposed ability check if it’s actively resisted (I know 5.5 mostly eliminates the latter, but I like opposed checks so I still use them where I think it makes sense). When an action interacts with the environment or a feature of the environment, I set a DC, typically using the DMG’s guideline of 10 for easy, 15 for medium, 20 for hard, 25 for very hard, 30 for almost impossible. I try to set them in a vacuum because if I’m taking the PCs stats into account it’s just a treadmill. Though sometimes I will set more granular DCs based on a sort of platonic ideal PC (one that has proficiency and +3 in the relevant ability at level 1-3, a +4 at level 4-7, a +5 at level 8-18, and a +6 at level 19-20), thinking about what I would want the percent chance for that character to succeed.

Regarding “situational modifiers”, I tend to try to roll factors that are outside the PC’s control into the DC, and award advantage or impose disadvantage based on things the PC actively does. Sometimes the lines can be a bit blurry though, so when in doubt I try to err on the side of rolling the modifiers into the DC, so as not to render superfluous any PC features or resources that can be used or spent to gain advantage. Also, when a roll would have both advantage and disadvantage, I have the player roll “with emphasis” as opposed to them canceling each other out. That is to say, roll twice and take the result farthest from 10 (ties round up if even or down if odd, which slightly favors the player).

 

[FrozenNorth]

I will echo @VashVet above in that is there is not pressure and the characters should have competence in the matter they pass the check.
I set average at 10, 15 for hard, 20 and 25 for increasing difficulty. I also treat the steps as milestones for greater success where appropriate.

This is my approach and I go even further. If a character is trained (or has expertise), I simply auto pass certain checks that untrained characters must make.

You have expertise in thieves’ tools and you are in a podunk village? You don’t need to roll to pick locks (though there is probably very little to steal). Same thing with swimming across a calm lake while trained in Athletics.

 

[Lazybones]

My practices are quite similar to the other posters here. I generally use the suggested 10/15/20/25 scale. I don't artificially inflate DCs to match player abilities, but sometimes it makes sense to do so (i.e., if the party is facing a cult with high-level members, their trap-setters probably have an appropriately higher skill). Usually once player characters start getting double-digit skill ratings I'll just end up calling for fewer skill checks, as others have noted.

There are a few items I keep in mind when determining the challenge of tasks on the fly. One is pace; a task that can be easy given a minute to catch your breath might be difficult in a rushed situation. I ran a dungeon recently where the doors were all stone and counterweighted, so they tended to stick. I ruled they could be opened easily enough outside of combat, but required a DC 10 Athletics check to do as an Action (and even higher if the party was trying to open them quickly enough to surprise something on the other side). Climbing in a hurry can be similar; I might describe a rise of boulders that doesn't require a check if the party is taking their time, but might require a DC 10 Athletics or Acrobatics (if parkour is an option) to succeed if the person is trying to do it as part of their speed in combat.

I also find that my players use the Help action for most of their checks; the only requirements I put here is that the helper be proficient (this is RAW in 5.5e now), and be in a position to lend assistance. I don't generally allow Help with lockpicking checks, for example, but the way the new rules are written they usually gain advantage anyway if they are proficient in both the tool and the relevant skill. Since both of those rules mean that a majority of skill checks will probably be made with Advantage, keep that in mind when considering DCs.

Guidance is another "always on" thing in my groups. Here I rule that it can be used only if the task can be completed in the minute that the spell lasts (so checking a lock for traps yes, searching a room no).

With all of these factors, the chances of success can improve significantly. Even at the low- to mid-levels, a proficient character with Advantage, Guidance, and maybe Bardic Inspiration will have very little chance of failing even a hard task.

The last point I'll mention is to remember that Perception checks are harder (Disadvantage, or +5 to DC) if the character is relying on Darkvision to see. I apply this as well to other sight-based checks like Investigation.

 

[Shiroiken]

The following is based on 5E & 5.5E, because edition matters a ton. The DCs for 3E would overall be higher, and the DCs for 4E should be level based.

I don't call for a roll unless I think there's a decent chance of failure. Yes, a 10 is "easy" in the sense that it's something that anyone unskilled and untalented can still do 55% of the time. I tend to use 10, 13, 15, 18, and 20 for most of my DCs, based on how hard I think something should be (and yes, I plan out the DCs during prep for most checks). Very, very rarely will I call for a DC: 5 or 25, and then only because of the severity for failure (in the case of DC: 5) or something extremely unlikely (in the case of DC: 25).

 

[cbwjm]

If a PC has the relevant skill and there isn't any pressure they normally just succeed.

If it's a situation I'm not prepped for, I use the 50/50 rule (DC 11) if I think a regular person should be able to complete the task at least half the time, otherwise I'll use DC 15.

If it's something truly obscure or incredibly difficult but otherwise possible, DC 20.

 

[Pedantic]

Reluctantly, because this should be one of the primary functions of the rules, to codify what the players can do, so they can make reasonable decisions about their action declarations.

It's a travesty that I, the DM, would "set" anything, when I should be simply describing a situation in line with some kind of level appropriate challenge advice, and the DC should be derived from the specifics of the situation.

 

[Zardnaar]

Generally pick a number. Often 10,15, 20.

If im designing and write one down its risk= reward. I've seeded one with DC 30-35. Fail and bad stuff happens.

PCs will get warned though vs gotcha.

 

[overgeeked]

Before I bounced form refereeing 5E, I used this setup.

All checks are basically saves against consequences rather than a binary succeed/fail. If there's no interesting consequences of failure, they just succeed. Nothing happens, i.e. you don't succeed, is not an interesting outcome.

There are five categories of checks.

Auto success.
Vs DC with advantage.
Vs DC.
Vs DC with disadvantage.
Auto fail.

The DC is always 8+ whatever their total modifier is. You get a +37 to your roll? Great, the DC is 45. You get a +7 to your roll? Great, the DC is 15.

This means the PCs always have 100%, 87.75%, 65%, 42.25%, or 0% chance of success.

 

[bsss]

The DC is always 8+ whatever their total modifier is. You get a +37 to your roll? Great, the DC is 45. You get a +7 to your roll? Great, the DC is 15.

This means the PCs always have 100%, 87.75%, 65%, 42.25%, or 0% chance of success.

Why have modifiers to the roll, at that point?

For d20 games I take the situational medium/hard/very hard difficulty approach, usually with +5 increments between them. "Easy" is no roll.