From Page 392 of the Core Rules
A challenging DC for a skill check is equal to 15 + 1-1/2 × the CR of the encounter or the PCs’ Average Party Level (APL). For an easier check, you might reduce the DC by 5, while increasing the DC by 5 makes for a more difficult check. Changing the DC by 10 or more makes for either a trivial check with little chance of failure or a prohibitively high check with little chance of success, so be cautious when adjusting skill check DCs!
Skill DCs by Level for Starfinder
A Challenging skill DC for any Average Party Level (APL) is…
Subtract 5 for Easy tasks; add 5 for Hard tasks, and add 10 for Very Difficult tasks.
GMs should most often use the Challenging DC. Occasionally use the Easy and Hard DCs, same as variability of challenge in encounters, so you should also vary skill checks. Use the Very Difficult DC when the players try something they know is nearly impossible.
Some skills are more popular than others. If everyone gets to attempt the check, or if the PCs get to choose who makes the check, use a harder DC for Acrobatics, <strong>Climb,</strong> Diplomacy, Disable Device, Intimidate, <strong>Knowledge (Arcana or Nature)</strong>, Perception*, Sense Motive, <strong>Spellcraft*</strong>, Stealth, Survival, and <strong>Use Magic Device</strong>. For Perception and Spellcraft, feel free to red-line them with lots of Hard DCs and even occasional Extremely Difficult DCs, because parties tend to have at least one or more characters with very high Perception or Spellcraft. Naturally, if you’re running a long-term home campaign, you can just use your intuition of the party’s skills instead of that list.
For tasks with easy to calculate DCs in the game already, use the task-based DC instead of the guidelines DC. For instance, monster identification is easy to remember at 10+CR. That’s easy to remember (and notice that for any given level, it tends to stay between the Easy and Moderately Challenging DC).
Breakdown: How Hard is Hard?
So what do Easy, Moderately Challenging, Hard, and Extremely Difficult mean? Well there’s some fairly easy math involved, and a few assumptions. First the assumptions.
An Easy task is one you can almost rely on succeeding. That means you have to better than 80% of the time.
An Extremely Difficult task is one you can almost rely on failing. That means you have to fail about 80% of the time or worse.
A Hard task is anywhere from “fifty fifty” odds to “one in four longshot” odds.
A Moderately Challenging task is more likely than not to succeed (55% or better), but not reliable enough to feel cocky about (not better than 80%).
Breakdown: Hard for Whom?
How did I arrive at Level plus 12? First, I played a lot of Pathfinder, and even more 3rd edition D&D. Plus a bit of 4th edition (which has a similar skill system). Then I took my Pathfinder experience and made a few reasonable assumptions about how players make characters, as far as skills go.
The equation, Level plus 12, is based on these assumptions:
Pathfinder’s character classes were built so that parties with complimentary combat abilities would tend to have complimentary skills, so while some skills are more common (see above), there’s a good chance someone in the party will have a skill at a decent rating.
Unskilled: Someone who puts nothing into a skill will probably have a +2 in it. You can get this much out of Aid Another.
Half Ranks, A character might put half ranks into a skill, and have Level ÷ 2 ranks at any level, plus 3 either from the class skill bonus or a decent attribute. You don’t usually put half ranks into a cross-class skill you don’t have a good stat in. Consider a halfling Rogue’s Climb. She might not have a lot of Strength, but it’s a class skill, and she might only drop half ranks into it to maintain some degree of competence.
A character with full ranks in a skill that’s not especially good will have a total of Level + 3 in that skill, either from the class skill bonus (such as the Cleric taking Spellcraft) or from a good stat for a cross-class skill (such as the cleric taking Perception).
A character with a good skill likely has Level + 9 in it. That’s full ranks, plus a class skill bonus (3) and a good stat (3 or 4) that’s key to their combat ability, so it increases over levels (eventually much more than 3 or 4). This character might also buy gear to boost this skill a bit. Consider a Barbarian’s Climb, with +3 for the class skill bonus, +4 from 18 Strength and +2 from a climber’s kit.
With four or five PCs in the party, there are probably going to be a few optimized skills there. An optimized skill is assumed to be Level x1.5 plus 8, to account for a rising attribute (which could start at 20 at level 1 and reach 36 at level 20), magic items to boost the skill, and mundane gear to boost it. Consider an elf Rogue’s +20 Disable Device at level 8, with a class skill bonus of +3, +7 from 24 Dex (including a belt of dexterity), and masterwork thieves’ tools (+2).
The goal is to have a skill DC where the majority of these PCs have to roll to see if they succeed. It can’t be so high they can’t succeed, even on a 20. It can’t be so low that they can succeed even on a 1.
I plotted all these assumptions out over 20 levels, and selected a target DC that kept as many of these in that sweet spot range as possible. You’ll notice that they diverge very rapidly. At level 3, the Unskilled and Optimized characters are 11 points apart. At level 18, they’re 33 points apart. So naturally the extremes fall off the chart at higher levels. Here’s what it looks like to use the 12+L equation or the table, above.
Chance to succeed at DC 15, 20, 25, and 30 at level 3, 8, 13 and 18, respectively.
This chart shows us what we want to see.
We want to see the “Good” skill do significantly better than the “Full Ranks” (but not good) skill. We want to see the “Unskilled” character have a chance, albeit a poor one, at low levels; and then drop off to no chance at all in the teen levels. We’re happy to see “half ranks” slowly losing ground at higher and higher levels – that’s how it should be.
We are OK with seeing the Optimized character quickly skyrocket up to “don’t even have to roll.” See, if you design skill DCs to keep the optimized characters on their toes, you have to drop all the other non-optimized characters (and all the other skills that the optimized characters have) off the other end of the chart. The thing about optimization is that you can manage to get one or two skills in the “never fail” range, but you can never get a lot of skills in that range.
Table or Equation?
It really doesn’t matter too much if you prefer the simple, round numbers of the table or the brevity of the L+12 (optionally -5, +5, +10) equation. Here’s a breakdown or the Full Ranks character’s chances across all 20 levels.
The equation results in a steady result like this:
Easy = 85%, Moderately Challenging = 60%, Hard = 35%, Extremely Difficult = 10%
The equation and the table match for levels 3, 8, 13, and 18. The table’s simple round numbers only ever deviate from the equation by 10% either way. Frankly, the skill values of your PCs will vary a lot more than 10% off of this chart. So you can be pretty comfortable using it.
Simulating what, exactly?
There are some Pathfinder GMs reading this with steam coming out of their ears. How can you just pull a DC out of nowhere!?
Pathfinder hails from a tradition of RPGs that attempt to simulate the material world with dice. The DC to climb a wall is based entirely off of what the wall is made of, the handholds it might have, if it is located in a corner, if it is slippery, and several other factors.
Sometimes you want to simulate a tense fantasy adventure movie, where the difficulty of getting up the wall is just difficult. It doesn’t matter what the set designer made the wall look like. At this point in the movie, either the hero gets up the wall, or slips and falls. Either way, the action continues from there.
Simulating a material world is fine. But the problem is that it often generates un-fun skill DCs because you decided the wall was brick, and that it was raining; not thinking about how those simple descriptive details completely negate a cool strategy the 3rd level players came up with. All of a sudden, the DC to climb the wall is impossible. On the other hand, an 18th level party might have no trouble escaping a situation you wanted to be stressful just because you described the walls of the oubliette they were thrown into as “rough, jagged natural stone walls.” Oops!
The difference between these approaches is best described as the task or challenge focus: Are DCs task-generated or challenge-generated?
If you’re dead set on task-generated skill DCs, you should still use these guidelines. Say you’re designing a break-in scenario, but the ground floor exterior wall of a building is Climb DC 25, according to task-generated DC rules for Climb. Now you know that a 3rd level party needs Climb DC 15 to be Moderately Challenging. So you make sure to put a tree next to the building, which is DC 15. There you go.
And there will be situations where you don’t know what DC to set for a skill. Just choose L+12, or pick from the table.
Another common complaint about challenge-generated skill DCs is that they create a treadmill. Each level the PCs should get better, but the DCs just go up at the same time. The solution to this is to describe the challenges as harder and harder at higher levels. At 1-5, when an Easy DC is 10, they’re talking to sniveling criminals and a drunken reeve. At 6-10, when an Easy DC is 15, they’re talking to a sheriff and bailiffs. At 11-15, the Easy DC goes to 20, and they’re talking to knights and barons. At level 16-20, with an Easy DC of 25, they’re talking to generals and bishops. If you used task-generated DCs, you’d probably see a similar progression. “Oh we can’t have them intimidate a drunken reeve, that’s too easy for level 17 PCs. Try to intimidate a great general. There we go.”
CR and Skill DCs
You can substitute CR for Level in the equation, above, allowing you to create easier and harder skill scenes. Say you want a CR 9 breaking and entering challenge. That requires DC 21 climbing, lockpicking, and sneaking rolls for a Moderately Difficult challenge.
CR can vary above or below the PCs’ level, so you can create a challenge that’s got Moderately Difficult DCs but has a CR that is 2 above the party’s APL, making it a difficult overall scene.
Now, we’re getting really technical here. Why bother? CR maps to experience points and treasure. If there is a non-trivial risk of loss in the skill scene, then it should be worth experience points and possibly include a treasure reward, just like a combat scene. A risk of loss includes:
Hit points and ability damage, such as from a trap or environmental hazard
Time, in a scenario where time is of the essence
Story loss, such as an important NPC’s life hanging in the balance
A tactical loss, such as making the next combat encounter easier or harder
A combat encounter – failing the skill scene triggers a fight, such as negotiating with highwaymen or sneaking past guards.