Skill challenges were the best kept secret of 4e D&D. They were awesome, but they were so poorly explained that it caused a lot of confusion. Not only that, it took the designers two whole errata revisions to get skill DCs right in the first place. Given how 5th edition D&D’s design process worked, the complex math undergirding a skill challenge sub-system was not likely to happen.
So I did the math and wrote an explanation that makes it clear how this works.
Running a Skill Challenge
A skill challenge is used when the PCs are attempting something complicated that requires multiple steps and multiple contributors. Tasks like persuading a noble, fortifying a ruin, winning a court case, searching the wilderness, tracking down a criminal, escaping from prison, crafting a magic item, climbing a mountain, or figuring out a complicated clockwork machine are good skill challenges. The practice of running a skill challenge is simple, and resembles the way a typical DM runs these kinds of exploration or social scenes, with one exception: The DM is secretly keeping score.
As the players take actions to contribute to achieving their goal, the DM does one of three things:
- Add a threat to the PCs’ progress and call for a roll
- Add an opportunity that’s too good to pass up and call for a roll
- Notice a conflict caused by a PC action and call for a roll
At first, the DM will listen to the players describe their characters’ actions, and call for rolls with the skills the players seem to be signaling they want to use, when the players’ actions lead to conflicts. For instance, the Rogue might decide to pick a guard’s pocket to get a key. That’s a conflict – the Rogue wants the key, the guard wants to avoid getting robbed. The DM would call for a Sleight of Hand check, and note the result.
As time passes, the DM might notice that the players are using the same skills over and over; or the same PCs seem to be doing everything. To add variety, the DM can add an opportunity or threat, forcing a particular player to use a particular skill. The Rogue might get the key, but find that the door it goes to has some magical runes on it. Figuring out what they mean would require an Arcana check.
As you run the challenge, you don’t have to make the players take turns, but you need to make sure one player isn’t hogging all the spotlight. Occasionally toss those opportunities and threats at the players who’ve been quiet for a while.
What makes a skill challenge different is that behind the scenes, the DM is keeping score. The scene ends when the PCs have accumulated a certain number of successes toward their goal, or suffered three failures.
As the scene goes on, the DM narrates events based on how well the PCs are doing. If they just scored their second failure, the DM should narrate how things are getting desperate. If they are one success from victory, the DM should narrate how success is within their grasp. Thus, a skill challenge structures and organizes a skill scene and makes it more than a series of unconnected die rolls. It also serves to force the players to use more than just their favorite skills, and limits how long the scene goes on for.
At the end of the scene, the DM narrates the conclusion based on the final score and the original stakes set for the scene.
From the players’ perspective, it’s no different from a typical exploration or social scene, except that it’s much more organized. Here’s an example of how a well-run skill challenge scene works in practice, written for 4th edition D&D.
Behind the Screen Mechanics
When you want to run a scene as a skill challenge, choose from the table how hard it should be and how complex it should be.
An easy skill challenge is one that the PCs will almost surely succeed at as long as they use appropriate skills and don’t do anything crazy. The challenge is more about “how” they will achieve success than “if.” You can use an Easy skill challenge with really dire stakes (the end of the world, everybody dies, etc.) and be fairly assured the PCs will triumph. Statistically, the PCs have around a 95% chance to succeed if they use the most appropriate skills. The escape scene at the start of Out of the Abyss would be a good place for one or more easy skill challenges. An easy skill challenge gives the players a chance to try crazy things. They can go way off the rails, and it won’t throw things off. If the Fighter starts making Arcana checks with a -1 modifier, there’s still a 50% chance they’re going to succeed.
A moderate skill challenge is one that the PCs will probably succeed, but there’s a good chance they’ll fail, too. With straight skill rolls, the PCs have around a 70% chance of success, which is by no means a safe bet. If they have Expertise in the most appropriate skills for the challenge, or have spells prepared that could give them a major advantage, their chance to succeed improves dramatically. But those things are investments or limited resources. Use moderate skill challenges with interesting stakes, where success and failure could both lead in different, but equally interesting story directions.
A hard skill challenge is one that the PCs will probably not succeed at. There’s only around a 40% chance they’ll succeed with straight skill rolls. However, especially at higher levels, they will have more potent Expertise in more skills and more spells that can contribute toward the challenge’s success. Use hard skill challenges with interesting stakes. It’s also best to use hard skill challenges when the PCs try to avoid other challenges. For instance, if the PCs are trying to use skills to avoid a combat encounter, use a hard skill challenge. Their chance to screw up and have to fight the monsters anyway is pretty high – but not nearly as high as just kicking down the door!
The table below tells you how many successes to require before 3 failures, for the challenge to succeed, at each tier.
Difficulty: Select the challenge difficulty based on the actual difficulty for the kinds of things they’ll be doing. Within the skill challenge, improvise or plan for obstacles of that difficulty. If your challenge is Moderate in DC, set up situations that are moderately difficult. If the PCs try something that would be significantly harder (because they’re crazy!) or find a way to make a challenge easier, definitely vary the DC a bit. Remember, Easy is DC 10, Moderate is DC 15, and Hard is DC 20, but if you need to vary the DC, you don’t have to use neat multiples of 5.
Complexity: Choose an option with more total skill checks for a challenge that’s more complex, requiring more steps to resolve it. Choose an option with fewer total skill checks for a challenge that’s simpler – fewer steps.
Skill Challenge Construction
|Tier||Easy Skill Challenge||Moderate Skill Challenge||Hard Skill Challenge||Minimum Spell Level for Free Success|
|Local Heroes||3-4 Easy||5-6 Easy, or
|Heroes of the Realm||6-8 Easy||9-15 Easy
|Masters of the Realm*||3-4 Moderate||5-6 Moderate
|Masters of the World||3-6 Moderate||7-9 Moderate
* For level 11+, don’t even call for rolls for Easy stuff, since the party should be able to succeed even on a natural 1.
Spells and other daily magic abilities are a precious resource in 5th edition. An appropriate spell might give the caster (or someone else) Advantage on their next roll at the DM’s discretion. Or a powerful spell might provide a free success in the challenge. With skill challenges that the PCs can regain spells during (sea voyage, crossing a desert, building a fort, etc.), a spell should only grant another character Advantage on a single roll, never a full success. Use your judgment as a DM, though. Just because a PC used a spell of the right level doesn’t mean it’s any help. And sometimes even a low-level spell is so appropriate that it should provide a free success.
Magic items with a limited daily use, such as wands, should be counted like spells. Magic items with infinite use might provide one free success, or else a bonus for one character, for all their rolls in the challenge. Consumable magic items should usually earn a free success, unless it’s not clear how they’ll be of any use.
For the purposes of XP (if you use it), a skill challenge counts as a combat encounter of the associated difficulty, at the PCs’ level. If you use a Hard skill challenge of a higher tier, it counts as a Deadly encounter for the PCs’ level.
Stakes and Skill Challenges
I’ve written at length about how everything you do should have clear stakes. A failed adventure, failed scene, or failed roll should impose a hazard. Skill challenges formalize the stakes. The stakes of each individual roll directly affect the scene stakes. See the variations below for an alternative skill challenge where there is a hazard for each failure.
Social Scene: Decide what the other side wants that the PCs won’t want to let them have. Decide why the opposition doesn’t want to give the PCs what they want. If the PCs do well, they can get what they want without giving anything up that they’re not willing to. If they do poorly, they will either fail to get what they want, or be forced to give up something. That’s the essence of any negotiation, threat, or con. Social scenes can mix DCs. An ogre might be Easy to Intimidate or Deceive, but Hard to Persuade. A mind flayer’s alien personality might be Hard to get Insight on, but Easy to Persuade.
Hazard Scene: Hazard scenes can do away with the three strikes rule and have a penalty for each failed check. Use them when the PCs encounter a hazard they need to get past, and they have to take risks to do so. Maybe they’re climbing down a cliff, crossing rapids, or navigating a desert. The stakes here aren’t “can you cross the rapids” but “can you cross the rapids without drowning or losing any gear.” Choose a complex skill challenge and decide that instead of “three strikes,” each failed check causes the character to lose a valuable piece of gear or gain a level of exhaustion, for example.
Variable Result: Another neat thing about a straight skill challenge is that when it ends, it ends with either 3 failed checks (total failure), 0 failed checks (total victory), 1 failed check (partial victory), or 2 failed checks (close call). You can use the final “score” for the challenge results. For instance, if the PCs are fortifying a town against a horde of zombies, the number of failed checks could represent the number of families that the zombies slaughter by getting inside of their homes. 3 failed checks means one of the homes is the home of the town priest, the townsfolk consider the fortifications a failure, and they evacuate.
Keeping Score Openly: Sometimes the “score” has an in-game representation. Examples include persuading a majority of a star chamber’s five judges to side with you in a trial, finding all six lost children before night falls and the full moon rises, or destroying at least five of the seven magical crystals before the dark ritual can be completed.
Taking Turns, Variable Order, or Anything Goes: In some cases, everyone has to participate. In those cases, have the players take turns, going around the table (initiative isn’t important). Sometimes you want everyone to work together, but it doesn’t matter if some folks do more than others. In those cases, use a more variable order: have each player nominate who goes next. This works well for investigative challenges. Anything goes is just that — the PCs go in whatever order they want. The danger with “anything goes” is that talkative players take over and wallflowers fade into the background, so be prepared to interject with opportunities and threats to get other PCs to participate.
Avoid the Combat System: There are some things that are very rare in 5th edition D&D, because the combat system makes them hard to handle. One situation is “getting captured.” Another is “retreating from a battle” or “chasing an enemy.” Getting captured usually involves a combat that lasts until the PCs have all been reduced to 0hp. Retreating is hard because of movement speeds and combat rounds. Chasing the enemy is a pain because of movement speeds, special abilities, and spells. Use a skill challenge to model these instead. Spells still help, and characters with higher than normal movement speed might get Advantage on checks. But the scene will be more fun with skill rolls than with combat rounds.
Avoiding an Enemy: When the PCs want to avoid a dangerous enemy through clever use of skills, you don’t want to let the tension out of the game by making it fast or easy. Instead, use a Hard skill challenge, where three strikes means that the PCs wind up fighting the enemy anyway, perhaps with some tactical disadvantage (e.g. surrounded, surprised, prone, or bottlenecked).
Design and Statistical Notes
I started going into the math of 5e skills on Run a Game earlier this year. I expanded to skill challenges using straight binomial distributions, with the assumption (culled from my previous work) that a Local Heroes party will most often have a +4 in any given skill, a Heroes of the Realm character will most likely have a +7, a Masters of the Realm character will most likely have a +9, and a Masters of the World character will most likely have a +10.
I also pulled data from the 5e Pregens and found their skills were about as expected (granted some of their skills have errors, but then so will your players’ character sheets!). Creating randomized “balanced” parties at each odd level, I found Local Heroes had a median +4 and Heroes of the Realm had a median of +6, which is well within sampling error of my estimated +7. The pregens don’t extend into the upper tiers. Frankly, even though 5th edition is the best edition for high level play in my opinion, things get weird after level 11 in D&D. So I used a simple formula: Take the best skill you can get without Expertise (Proficiency, +5 for a 20 attribute) and subtract 1.
I sacrificed a little accuracy for simplicity, but it’s worth noting that some skills are more commonly chosen than others. Even in the pregens WOTC published, Survival, Religion, Animal Handling, Nature, Arcana, and History had the lowest median “party best” values. You would also expect some skills to be higher than typical: Perception, Insight, Athletics, Persuasion and Stealth are likely to be higher than expected, given players seem to find them important (Perception), or they’re on a limited skill list for a class that favors the related attribute (Athletics, Insight), or they’re high stakes skills common to classes with Expertise (Stealth, Persuasion). On the other hand, even if their Charisma is low, most PCs pick up a social skill so that they can participate in interaction scenes – so, even if you have a Bard with a +8 Persuasion, expect the Fighter to take a stab at Intimidation (yuk yuk) even if the Fighter only has a measly +2.
Party size matters! Note that there are 18 skills in 5th edition D&D. With a party of 3 PCs, it’s impossible for the party to be proficient in each skill. With a party of 4, it’s possible, but it’s likely the party will still suck at one or two important skills and several of the lower priority ones. With a party of 5, it’s not too hard. With a party of 6 or more, it’s easy to be proficient – or even optimized – in all the skills. My base assumption (and my random pregen party sample) is based on a “typical” party of 4, and works just fine for a party of 5, so long as it’s not all Lore Bards.
Speak of Lore Bards… The number of PCs in the party who have optimized Expertise will dramatically affect your skill challenge. A party with no characters with Expertise will have a harder time with skill challenges. My assumption was a party of 1 such PC, and my random pregen parties had 0-2. You might want to give out spell scrolls (especially utility spells) and other consumable items as treasure more often for a group without any Expertise characters. At the other extreme, a party with more than one character with Expertise likely has amazing skill values in several common exploration and social skills. If you have a lot of Expertise in the party, consider using harder skill challenges or even just jumping up a tier.