Skill Challenges (or, They Wasted a Perfectly Good Mechanic)

One thing I was excited about in 4th Edition was skill challenges. I think they’re a great idea. I think they’re such a great idea that I can’t help but notice that every DM in every campaign I’ve seen has run one, even before D&D 3.5 was published. Having a set of rules for it would just contribute to making this free-range, non-combat style of encounter a core part of the gameplay.

But then I read it.

Take, for example, the skill challenge “The Negotiation” on page 76 of the Dungeon Master’s Guide. In this skill challenge, the players can contribute to success by making Bluff, Diplomacy, or Insight checks. Once during the challenge, one player can make a History check to gain a success, and any Intimidate check is an automatic failure. It’s a negotiation, so this sounds reasonable.

But then, look at the eight classes in the Player’s Handbook, the only classes published when the DMG came out. Of those classes, two of them have none of the skills needed in the encounter (ranger, warlord) and one only has the once-per challenge skill (fighter). There’s a very good chance a party has one of these classes, so that character is essentially locked out of the challenge. To be useful, the other characters must have chosen the correct skills, so a paladin who trained in Endurance, Heal, and Intimidate is also locked out.

The only chance those characters (and players) have to participate is if they think of a sufficiently good reason to use one of their good skills, which is subject to DM approval and made at the hardest difficulty by rule. Normally, this isn’t so bad, because it just means that the ranger keeps quiet during negotiations. But in the 4th Edition skill challenge system, players roll initiative and make skills checks in order. Bowing out of a skill challenge is not an option. That player is forced to make a check with a good chance of failure, not because it’s interesting or because the rewards are great but because the rules say that they must.

Wizards did address this somewhat in the Dungeon Master’s Guide 2, which is news to a lot of the people reading this because they either don’t own that book or skipped the section on skill challenges (for reference, I’m in three campaigns now, and I’m the only one in any of them who read the DMG2 of his own volition). They didn’t explicitly say that skill challenges don’t work on initiative any more, and they removed that section from the rules in an errata, but their example of play went in initiative order anyway. They now recommend building skill challenges based on your party’s abilities and options rather than a high concept. They also increased the number of skills in a skill challenge, saying this:

…in play, [a skill challenge with only a few usable skills] is boring. The characters who excel at those skills make their checks while everyone else hangs back reluctant to make their own checks because they don’t want to be the reason that the group fails the challenge. If your potential skill challenge has a similar bottleneck—in which only one skill does the heavy lifting for the challenge—you need to rethink the challenge.

It was a very quiet way of saying “every skill challenge example we’ve given you before now was wrong”. I’m all for fixing a broken system, but maybe a system shouldn’t be published if there wasn’t a single thing about it that was right.

So after the errata, we’re left with a system for skills challenges that has these rules:

• A skill challenge has a complexity rating that determines the number of successful skill checks needed to win the challenge and the XP reward for winning.
• The players lose the challenge if they fail three skill checks.

…that’s it. Everything else is up to the DM to design and adjudicate (or it’s already part of the core rules, like rolling a d20 and adding a modifier). The only thing the skill challenge system gives us is “how do I know when I win, and what do I get for it?”, which isn’t the worst possible way to design a mechanic but is the worst possible way to design an ostensibly roleplay-based encounter. It’s not a sufficient framework in a framework-heavy system like 4th Edition.

So I ended up working on my own skill challenge system. It may or may not work for your style, but it’s allowed me to free-range the challenge as it runs (like I DM with Charisma or something) while still giving me enough structure to explain it to players and track things appropriately.

• Don’t determine any skills beforehand. This isn’t really a rule, but it’s different from the normal system so it merits mentioning. Do determine five possible results of the skills challenge: one for if the players fail significantly, one for if they barely fail, one for if they neither fail nor succeed, one of if they succeed barely, and one for if they succeed significantly. Also, determine the number of checks the players need to succeed or fail (I like seven).
• Inform the players that they’re in a skill challenge. It may also be best to put them in some order like initiative just so everybody gets a chance to speak, but that’s up to you. I like putting them in initiative order when time is of the essence in the skill challenge, or where having a character sit around and do nothing contributes to failure.
• When a player wants to act or when it’s their turn, ask them for the skill they’re using and the reason they’re using that skill. Based on the explanation, determine whether they’re rolling against the easy, medium, or hard DC.
• Have the player roll their skill check.
• Based on their result, decide whether the player failed significantly (“red”), failed (“orange”), broke even (“yellow”), succeeded (“green”), or succeeded significantly (“blue”). (I guess we’ll call these colors “bands” of results. I don’t have a good name for them yet.) Note this result.
• Inform the player what happens as a result of their check and move to a different player. Don’t let them know the specific band in which their result fell, though based on your response it should be somewhat obvious.
• When the party gets some number of results in a given band, the skill challenge ends and the players learn the final result.
• No XP is awarded, because XP is for losers. But if you must, you can award XP based on the number of results needed to succeed or fail.

Determining the DC and the result of a check is more of an art than a science, but I try to stick by a few points:

• I reward good ideas over ideas that the character can do easily. If the wizard wants to make an Arcana check to detect a hidden door, I tend to be more harsh than if somebody wants to use Perception.
• I reward varying ideas over using the same skill every turn. It’s a “skill” challenge, not a “my best skill is Diplomacy, so let’s roll five of those” challenge.
• I reward risk over safety. If a player wants to use Religion to ward against a demon, that’s not as interesting as a player who uses Athletics to grapple it.
• I reward expenditure of resources. I’ll often bump a success up to the next color band if the player spent an action point, a daily power, or some number of healing surges to achieve it.
• I reward awesome over mundane. If an idea is exciting, I’ll often use a lower DC simply because I want to see it succeed.

An example: Varon, a half-elf bard, would like to use his Diplomacy to ask random people whether they saw the fleeing criminal. This helps him if he succeeds but doesn’t hurt him if he loses besides wasting time, so I use a moderate DC. If he’s dealing with orcs, I increase the DC to hard because orcs don’t like elves. If the skill check succeeds, I put a mark in the “green” or “blue” band, depending on how the player phrased his questions and how well the check went. If it fails, I put a mark in the “yellow” band.

His ally is Tela, a goliath warden. She wants to block the traffic in the street, fencing the criminal in, using Endurance or Athletics. Since failure could cause damage to her or create confusion that lets the criminal hide but success would hinder him, I use the moderate DC. If the skill check fails, she gets a mark in the “orange” band.

But Near, the party’s tiefling warlock, had found the criminal and, absent other options, attempts to attack him with Arcana. If she succeeds, she deals damage but doesn’t actually harm the criminal, and failure could injure passers-by. I use the moderate DC, maybe moving it to the easy DC if she’s willing to expend an encounter or daily power. If the check fails, I add a mark to the “red” band, while a success is just “yellow”. If Near opts to use a power that can blind or slow the criminal, a successful check may instead be in the “green” band. In either case, any damage dealt might apply to the criminal if the party fights him at the end of the chase.

After the players make a bunch of skill checks and have gone through a bunch of turns, they get some final result based on which color band first reached the threshold I’ve set. If they his the threshold in the “orange” band, the criminal gets away and the players need to go look for him. If “red”, the criminal gets away and the city watch is considering pressing charges against the players as accomplices. If “yellow”, the criminal gets away but the players have a lead to track him down. If “green”, they catch the criminal, and if “blue” they actually ambush him and get a surprise round.

So what does this give me over the normal skill challenge system? Most obviously, it creates five different results rather than two. The players can be rewarded for slaughtering a skill challenge more than just “you open the magical door”, and they can be penalized for crippling failure and bad decisions more than just “you need another way in”. It also means that they have a lot more freedom to try different ideas because they aren’t beholden to specific skills and, at your option, can fail more than three times before ending the challenge. Most helpfully, it works in 3rd Edition and Pathfinder, while the original system only works in 4th Edition due to the lower number of skills. Basing a skill challenge on seven skills out of seventeen is fairly inclusive, but not if it’s seven out of thirty and a lot of characters only have two skills.

The downside is that there’s a lot more on-the-fly adjudication, including the DCs and the results of every check. It makes it a little easier to flow from the beginning of the skill challenge to the final result with each action, but it’s still up to you to tell the players what happened as a result of each skill check. It’s also a little jarring when the players get six “red” results and six “orange” results, but get seven “green” results first and thus succeed. It could mean that they did well were it really counted, but sometimes it’s a bit awkward to justify.

Is this system better than the official system? Yes, because I wrote it and I’m allowed to make incredibly biased statements like that. But I know it’s not for every DM or every group of players. I know that it’s really best for people who like coming up with ideas on the fly, and it’s not for people who tend to stick to a few things at which they’re very good (Like a skill challenge itself! Gasp!). It might be something worth trying during a future session, and you can take it or leave it based on the opinions of your group.