Developing skillcheck DCs for an adventure

Question

I'm looking something alongside the lines of:

Calculating and balancing challenge ratings

...but for DCs for skillchecks in an adventure. Traps, chests, climbing, swimming, etc. how hard those checks should be? And how many of each difficulty?

I'd like to have some kind of standardized DCs table specified in relation to player/group levels. So far I was tailoring that to the skill numbers of my players (for example so there is 25% or 50% fail chance most of the time) which didn't feel fair, because it didn't reward player for investing a lot in certain skills nor did it punish players for ignoring investment in important skills either.

So what I'm looking for are tables that tell me:

• What DC is considered very easy, easy, medium, hard, very hard, etc. in relation to player level?
• How many very easy, easy, medium, hard, very hard, etc. checks should average adventure have?

Why do I need such tables? Because I don't want the world to feel like it's tailored exclusively for the players and I don't like tailoring adventures for specific groups. It feels like cheating. Players (as a group) should build characters with "anything can happen" mindset in my sessions while also prioritizing skills they believe will be usefull more often than others. Every skillpoint should matter and skillchecks should be just as emotionally engaging as fights or even more.

Creating DC's of skillchecks based on character/group levels make just as much sense as CR stat for combat. As a DM I should expect higher level thief to be taking on harder traps and locks than lower level thief, just as I expect higher level fighter to take on tougher baddies. I just need a standarized numbers to be able to do that, just as it's done for CR. And while rulebooks do tell me what I can throw at warrior of level X (in my example) and what should feel easy or hard for him those same books don't tell me what kinds of traps or locks I can throw at thief of level X.

P.S. I'm completely fine with answers based on either 3.0 or 3.75 (Pathfinder). It's easy enough to translate that I don't mind.

Answer 1

Dungeon Master's Guide (third edition), p93

The 3e DMG (the one in my hand is the first printing, September 2000) has a table on page 93 "Table 3-21: Difficulty Class Examples" that describes 40+ example tasks with DCs ranging from -10 to 43, an indication of the test, and who could do it. A few examples are:

• DC 15: Make indifferent people feel friendly. Diplomacy (Cha) A 1st-level paladin
• DC 28: Disable a glyph of warding. Disable Device (Int) A high-level rogue (but not anyone of another class)
• DC 43: Track a goblin that passed over hard rocks a week ago, and it snowed yesterday. Wilderness Lore (Wis) A 20th-level ranger who has maxed out his Wilderness Lore skill and has been fighting goblinoids as his favored enemy since 1st level.

It leaves some work for the DM to match up or interpolate for the specific tasks at hand, but the guidance to do so is there.

Answer 2

You can estimate max rolls by adding max skill ranks plus max roll plus other modifiers, like feats or magic. For example, a 4th level character has a maximum skill rank of character level plus 3 (+7), with up to +4 ability modifier, so a DC higher than 20 + 11 (31) is going to be unattainable without other modifiers, like magic or racial bonuses, etc.

Keep in mind, if you put a skill check in the path of the players, be prepared for failure. It's a pretty weak ending if the characters can't enter or find your cool whatever you prepared. In my opinion, you shouldn't even call for skill checks unless failure is interesting.

For example, if access to the portal that leads to the next adventure is gated by a skill check, it's going to derail if that check fails. I'm not saying don't do it, necessarily, but at least be prepared for it.

Another example; if you prepared some juicy information for the players that they are supposed to get from a patron you spent quality time developing, don't gate their running into this patron on a gather information check. Instead, maybe add corroborating patrons on success, or contradictory ones on failure, but leave the interesting guy in the story in any case.

And, probably the most famous example; the climb check. Is it interesting to have players make climb checks, fail them, them pass them, then eventually make it up the side of the tower? If so, by all means, but adjudicating a series of stalls and recoveries is a terrible way to spend time around the table in my opinion.

Answer 3

Note: If all that's needed are guidelines for determining the DCs of Dungeons & Dragons 3.5 static skill checks, the Player's Handbook (2003) provides general guidelines for static DCs on Table 4–3: Difficulty Class Examples (64) and the Dungeon Master's Guide (2003) has more examples on Table 2–5: Difficulty Class Examples (31).

Skill check DCs don't, by default, scale with the PCs' levels…

In a role-playing game system like Dungeons & Dragons, Third Edition and its ilk that tries to emulate reality—that is, in a largely simulationist system—unchanging skill check DCs are the norm. The difficulty of a task is established first and then folks try to perform that task. (Contrast this with determining who's performing the task first and then setting the difficulty based on that performer!)

To a degree, such a system empowers both the DM and the players: both know exactly what creatures can accomplish, and what they can accomplish serve as benchmarks that indicate their proficiency. For example, a level 1 PC has a Climb skill modifier of +3. The player can look up in the Player's Handbook the skill Climb and determine that, if the PC can take 10, then the PC can climb a knotted rope or cavort about in a ship's rigging, but the player must roll if the PC is wants to go up a tree.

Thus the player and the PC know that if that PC takes 10 to monkey around in the rigging and still fails that something untoward's afoot. For example, maybe circumstances (sinister circumstances) have made the rigging more dangerous, or maybe the PC's affected by something (a sinister something) that penalizes his skill? No matter. This failure—when success should've been certain—merits an explanation or, at least, an investigation.

This notion of knowing a PC's capabilities (and, perhaps, by extension possible outcomes of those capabilities) before an action is attempted—this reliability—is often absent in role-playing game systems that use exclusively scaling DCs.¹ If the DC for a Climb skill check, for example, depends on the PC who's making the Climb skill check, then neither player, PC, nor DM actually knows how good a climber that PC is! There's no way to determine any creature's actual proficiency at a task with the universe always increasing a task's difficulty so that it's in line with the performer! Rather than some tasks becoming easier as the creature gains in proficiency, everything stays just as difficult.

To return to the example: If the PC arbitrarily can, even in the most generous of circumstances, always possibly fail to cavort about in the ship's rigging or shinny up a knotted rope, the player and the PC will be reluctant to attempt those tasks because disaster always looms. A chance to roll is a chance to fail.²

This, too, may seem realistic: in real life, there is always the chance of failure, after all, but life's random number generator is far bigger than the d20 this game uses! I mean, if even a master climber always fails 5% of the time when he tries to get up that knotted rope, for instance, that failure is at least embarrassing and at worst deadly. Imagine if you, dear reader, had a 5% chance of failing at everything you attempted, from making breakfast to tying your shoes to driving to work. Were the universe that horrifyingly random, I doubt many of us would make it through the day alive much less unscathed. (Also see this question.)

(To be fair, Dungeons & Dragons, Third Edition et al. do have some elements that scale to the PCs—the most important probably being an encounter's Encounter Level—, but these are specific exceptions rather than general cases.)

…Used sparingly, though, this isn't all bad!

You, however, aren't alone. Other folks do feel as you do: that some tasks should always carry with them a risk of failure and that element of risk should be in line with the potential proficiency of the creature attempting the task.

For example, Crafty Games's d20-inspired FantasyCraft (2009) has a section Using Sliding DCs (370). (Yes, it's a big book, but, to be fair, this is near the end.) That section sets the DC of a level 1 PC's Easy task to 10, Average to 13, Tricky to 16, Hard to 19, and Desperate to 22. For every 2 levels a PC gains those DCs pretty much increase by 1. As FantasyCraft has many of the same benchmarks as Dungeons & Dragons, Third Edition et al., these numbers might serve as a decent scaffold on which to build your for own similar house rules for Dungeons & Dragons, Third Edition.³

To be clear, accustomed as I am to Dungeons & Dragons, Third Edition et al., I've found FantasyCraft's idea of sliding DCs very difficult to conceptualize. That is, it's difficult as a GM and as a player to imagine that the universe thinks that no matter folks' experience levels they should always remain challenged by the same things.⁴ Nonetheless, I've also found the idea playable enough: sliding DCs are the exception, and there's sufficient reliability in the remainder of the system to be able to determine what folks are actually good at.

"How many skill checks should an adventure have?"

I think this may be two different questions.

• How many skill checks should PCs have to make to experience all of an adventure's content? This depends mightily upon the adventure's depth and length. To be clear, though, as this fine answer explains, if success on a skill check is required for the adventure to progress, then either success on that skill check should be automatic or the players should be told that they'll have to adventure elsewhere until their PCs are higher level and return later to continue this adventure.

  Beyond that, it's totally okay to gate some extended content behind certain skill checks. That rewards PCs who've invested in certain skills. For instance, if it takes an Appraise skill check (DC 30) or Knowledge (local) skill check (DC 25) to determine that the little 100-gp gem is actually worth 10,000 gp to the old wizard in Drawkcabnowt—and the PCs won't be shorted gp after the adventure if they don't realize this—, then that's totally okay. Likewise, if the PCs don't succeed on a Search skill check (DC 25) to discover the secret chamber, then they wont fight the ghouls, won't get the XP for that encounter, and won't acquire the ghouls' meager hoard. That, too, is usually okay… y'know, as long as the adventure isn't called, like, Where Are the Ghouls Hiding? or something.

  In short, I recommend preparing necessary content first and then gating some interesting but unnecessary content behind skill checks… and, of course, behind other things, too, like character background, experiences in other adventures, rumors picked up in town, and so on.

• How many skill checks does it take to yield a challenge that's worth XP? Based on Traps (Dungeon Master's Guide 67-76), it seems that if it's possible that, by failing one or two skill checks, the PCs will exhaust some of their resources, then making those skill checks can be considered an encounter. Based on the difficulty of the checks and the resources that'll have to be expended to deal with the consequences of failure, those skill checks can then be assigned a Challenge Rating (and an Encounter Level and subsequently XP). However, that's not the whole story.

  XP is typically earned for encountering the trap, whether that means drunkenly blundering into it or adroitly disarming it after having discovered its cunning trigger. Either way, the XP's the same. What matters is that there's a risk of the PCs depleting some of their resources.

  For instance, PCs that fall into a pit might have to climb out and, sure, those who don't will die, but the PCs already should've gotten XP for the pit. Similarly, if success or failure on a series of Climb skill checks to reach the mountaintop is worth XP to the fighter that does so, that shortchanges the druid who bypasses the Climb skill checks yet nonetheless expends resources to use wild shape to assume eagle form and fly up there.

  To sum up, I recommend assigning a CR to a general problem (like The trap is CR 3) and not assigning a CR to a specific task or an absolute number of skill checks (like It's CR 3 to avoid the trap by succeeding on 1 Search skill check (DC 23) then succeeding on 1 Disable Device skill check (DC 25)). That makes it so the players can solve the problem without worrying that their PCs solved it wrongly because the players failed to read the DM's mind about how the DM thought the problem should have been solved! If you're going to make something worth XP, assign a CR to reaching the mountaintop rather than assigning a CR to climbing the mountain.

---

¹ I've seen such systems sometimes disparaged as quantum bears. That is, a system in which the GM determines a task's outcome based solely on the result of the character attempting the task, ignoring the setting's role. For instance, a poor roll when exploring the cave may see the PC attacked by bears that wouldn't've been present had the player rolled better.
² A chance to roll is a chance to fail was coined by a player in a near-decade-long GURPS, Third Edition fantasy campaign that I GMed. In that 3d6-roll-under system, an 18 is always an automatic disastrous failure. The coining player's PC routinely failed Ride (Horse) skill rolls despite a skill of 15 or less. On several occasions, the PC severely injured his limbs,… and the game mandates a Ride (Horse) skill roll every time a character climbs aboard a horse! Eventually, even when horses were available, the PC walked.
³ In the interest of full disclosure, I have a playtest credit for FantasyCraft, but don't worry: you buying it doesn't give me money.
⁴ As a fan of the video game series Borderlands, though, I get why this happens.

Answer 4

I don't believe there is going to be a straightforward answer for this. The reason being is that no one knows what your characters can and cannot do.

The suggested DC's start at DC10, for example, set as an "average" task. This means that at the very least, they need a 10 to pass. However, at level 1, people have skill proficiencies and attribute modifiers that come into play, that can affect this. A wizard, for example would have potentially +3 to their Intelligence score, and a Barbarian might have -2. So making an Intelligence check for either of these becomes easier, or harder, respectively.

As characters progress, a DC10 check would become even easier - as even the Barbarian might have put more points into their Intelligence, let alone the Wizard, who might reach the point of +5 Intelligence, or beyond. So the best way to handle this would be to estimate the party's chances of succeeding a DC over a range of DC values. With a +5 and DC 10, that's a success on rolling a 5 through 20, aka 75% of possible rolls. On the other hand, a DC 15 is a success of rolling a 10 and above, which is only 50%.

This leads on to how many DC's you want in your game. Again, there is no hard answer for this. It all depends on what kind of adventure you want to run. For the most part, tasks should be passable, but in some cases, there might be tasks that are insurmountable, either because the players simply don't have the abilities or equipment to do anything about it, I.e. climbing a sheer cliff, with no magic, climbing gear or other equipment, would make for a DC of 25 or higher, and likely only increase with each check. However if you made every check that difficult, nothing would ever happen.

3.5 has the "Take 10" and "Take 20" options that can help with some checks, but this requires time. If there is no threats or time constraints, players can take their time on an activity, and they can add 10, or 20 to their "roll" - i.e. statistically, eventually a player will roll a 10 (take 10) or a 20 (take 20), on a roll. Even so, this doesn't mean that any skill check is doable by anybody. In a DC 25 Intelligence check, a player with +4 Intelligence will still not be able to pass.
^Source

Skill checks should be used as a mechanic to involve the players, add a degree of challenge to the game, and can be used as a way to control the flow of the game - but they should be used in this regard very sparingly.

Answer 5

It's not that simple...

Unfortunately, I am afraid that a single set of numbers is not as suitable as the DMG 3.0 would like to advertise it.

If you look at the DCs listed for a variety of skills, you will realize that they scale differently, even when backed by the same ability score and destined for the same class:

• Balance: Maximum DC listed, outside modifiers, for a narrow surface less than 2 inches wide => 20.
• Open Lock: Maximum DC listed, outside modifiers, for an amazing lock => 40.

You may certainly encounter a Balance DC of 30: less than 2 inches wide, severely obstructed and severely slippery. It still pales before the DC 40 of an amazing lock, which can also be improved.

---

And may be unexpected.

As a player, when deciding how to allocate my skill points, I use the DCs published for each skill to guide me; the SRD is very helpful.

For example, I know that Disable Device skill is a hit or miss: a magic trap has a DC of 25+spell level (so between 25 and 34), and failing by more than 4 sets off the trap.

Contrast this with the Ride skill where fighting with a war horse is a mere DC 10, and the hardest listed DC is 20¹ (assuming I do not plan on riding bareback). A high-Dexterity character (+10), riding with a military saddle (+2) and having 5 ranks in Handle Animal (+2), can succeed on that DC 20 with a roll of 6 with no skill point invested. And if they invest 5 points, they'll even succeed on a roll of 1!

I would not expect a DM to change the DCs one-sidedly!

¹ Of course, a DM may certainly impose circumstance penalties, or a player may ask for actions beyond the difficulty expected of listed DCs.

---

Smoothing the DCs is game-changing.

And like any house-rule, it is best agreed ahead of time with the entire group.

You may find it easier to adjudicate DCs on a skill by skill basis instead; using the published DCs as a guideline.

Answer 6

If you want every skill point spent to be worth something, you will have to keep track of each character's skill list.

First, look for skills that can not be used untrained and see which character does have point invested in them. Second, create situation in which some of these are useful. Third, make sure the players know that they can use these skills. You could have an NPC lament how a particular skill set would be useful to do a specific action but they do not have it or there is not one with that skill set around. You can also tell them outright since their characters should have the knowledge to recognize that they can use their skill in certain situation (for example, if there are runes and a character has rank(s) in the appropriate skill such as Decipher Script or Spellcraft then they should not have to ask to make a check, you should be the one to tell them).

In 3.5, skill DC are not directly based on the character's level. You really should read the chapter 4 of the Player's Handbook. The skill DC follow specific rules depending on which skill is used. You do not have to use all of the modifiers that are in there but you should tell your player if you change something that could negatively affect them.

The table 4-3 gives a summary of the difficulty of DC as follow :

• Very easy DC 0
• Easy DC 5
• Medium DC 10
• Hard 15
• Very hard 20
• Heroic (Superhero like) 30
• (Nearly) impossible 40

There is a section in the Dungeon Master's Guide on trap (Chapter 3 page 67). The sample traps have a CR that you can use to determine whether you should or not use the trap against your party by using the CR system.