Skill Challenges: Threat or Menace?

This has been bugging me a while, so I’m finally going to rant about it and get it out of my system.

Skill Challenges are what D&D 4e has in place of roleplaying.  And no, I’m not really kidding.  As a method of injecting some pseudo-RP in a tabletop miniatures skirmish game, they make perfect sense: a series of discrete, finite dice rolls so you can get quickly get past the RP and on with the real business of pushing minis around, and to provide some meta-game tension to the “boring” process of thinking of solutions to problems and playing them out.  As an aid to actual RP, they discourage what you want to encourage (creativity, experimentation, thinking as the character), and encourage what you want to discourage (meta-gaming and thinking inside the box).  I think it’s particularly telling that as originally released, even after all the playtesting 4e got, the Skill Challenge numbers were utterly broken and had to be rewritten and released as errata.

The basic mechanism is that you take something that would otherwise be RPed out and replace it with a Skill Challenge of a particular difficulty: you need to score N successes before M failures (N >M in most of the examples I’ve seen) or you fail.  N and M, and the difficulty of the skill rolls (DC in D&D terms) are determined by the difficulty of the Skill Challenge.  The Challenge will then list the Skills that can be applied and their DCs (at least in terms of easy, moderate, and hard), as well as how many successes (or failures) a check is worth towards completing the Skill Challenge (default 1).  Typically the list of allowed Skill checks, or even whether they can undertake a Skill Challenge, won’t be shared with the players, which is where people get fooled into thinking that this is somehow a form of RP instead of a replacement for it.  The process of discovering whether there’s a Skill Challenge and what Skills are allowed is superficially similar to the discussion you might have around the table of plans and approaches to try, but there’s a world of difference under the hood.

What’s wrong with that?

  • The biggest thing is the “Before M failures” rule.  That changes it from an abstraction to a mini-game.  The difference is that an abstraction simplifies things by ignoring irrelevant (hopefully) features to focus on getting reasonable outcomes, mini-games introduce features that have no counterpart in what you’re modeling and no purpose other than to make the mechanics of what you’re doing game-able.  In D&D hit points are an abstraction…they ignore things like exactly how or where you were hit (or even whether you were hit and actually wounded or just battered, bruised and tuckered out) to focus on the potential outcomes after a certain amount of battle: defeated, unharmed, victorious but weakened and less likely to win the next battle.  The fact that they’re a very high-level abstraction has caused a lot of complaint over the years, and prompted countless attempts to fix them or replace them with a different abstraction, but at least they don’t introduce extra new complications in the form of how do you massage your hit-points to use them most efficiently.  The Skill Check failures rule doesn’t represent anything in the game world (if you’re trying to track goblins through the woods, it doesn’t suddenly become impossible because your companion back in town failed to remember some detail about local politics), it’s just an artificial mechanism to introduce tension and limit the number of things the players can try. Arbitrarily limiting the number of things the players can try is bad. If they’re not expending resources or up against a deadline, cutting off their creativity is the last thing you want to be doing in an RPG, even if it makes perfect sense in a board game.
  • The fact that it’s a mini-game encourages/requires meta-gaming.  The order in which you try the tests is crucial, so rather than leading off with the skills that are the most relevant to the task you’re trying to accomplish you have to lead off with the skills you’re best at.  It’s the codification of the old joke Q: If your keys are over there, why are you looking here?  A: Because the light’s better here.  It doesn’t matter whether it’s something your character would attempt, what matters is how good the character is at it and the fact that if he fails he can screw up the Challenge for the entire party.
  • It discourages improvisation.   Dividing the world into things for which there are Skill Challenges and things for which there aren’t pushes the players towards the courses of action for which there are Skill Challenges and discourages them from trying things for which there aren’t (this is true even if the GM is willing and able to generate Skill Challenges on the fly).   Also, because Skill Challenges are defined in terms of making skill rolls you’re discouraged from thinking outside the box and substituting a non-skill roll approach to the same task (e.g. if you can make a Climbing roll to gain a success by getting to the top of a tree and spotting something, you ought to be able to do the same thing with an innate or spell-given ability to fly, or to send your flying familiar aloft.  The GM could just rule those as successes, but that doesn’t seem to be the way Challenges are designed.  The point isn’t the tasks you need to accomplish, it’s whether you make the required rolls).  Finally, by providing a list of approved skills and their uses to count towards the Challenge, it discourages using substitutes (e.g. if the Challenge calls for intimidating someone, what about persuading, bribing or seducing them instead?) or trying oddball or long-shot things.
  • By reducing everything to a series of die-rolls it eliminates any of the back-and-forth and progressive discovery and reasoning that are the essence of role playing.  This is an old complaint along the lines of if characters can just roll Diplomacy, then what’s the point of playing out conversations?  But it’s as valid now as it was then…if your game allows you to substitute skill rolls for interacting with NPCs and making actual deductions based on your knowledge instead of using them as a supplement or a fall-back if you get stuck, then you’re devaluing roleplaying.  Now 4e has made devaluing roleplaying one of its key components.

Let’s take a look at actual play, by someone wildly enthusiastic about Skill Challenges (note, I’m not criticizing Emptythreat15 or his GM at all; they had a lot of fun, which is what counts, but I’m arguing that they could have just as much fun and more often with a system that doesn’t get in the way of what they’re trying to do):

    • My male dwarf normally gets decent initiative; however, my female is cursed to having initiative rolls of less than 10. So my dwarf waded into combat with the hideous dogs and I thought to myself, ‘I wonder if we could improve the attitude of that boar and have it help us out…we don’t have Handle Animal anymore…but hell, maybe the DM will think of something suitable’.

      So my turn came around, and I explain to the DM that I want to undergo a skill challenge to improve the attitude of the boar, and what do you know, the skill challenge “Taming The Beast” was listed in the rules for the adventure. Sweet.

That right there is enough to make a grown man weep.  The player comes up with a really neat idea, and what’s he excited about?  The module gives him permission to try it.

This was turning out to be no easy task though. 4 successes before 2 failures? This thing was just built for failure. Fortunately for me though, I was allowed to use the better of my Heal or Nature checks. Asking a cleric to make a Heal check is like asking a clown to make a balloon giraffe, so I was pretty excited. Much to my chagrin, my clerics +11 Heal modifier does very little when you roll a 2 on a d20. One failure already. Not good.

The next two rounds, a 15 and 16 consecutively, making 26 and 27 Heal checks, and successful ones at that. The next turn, two have a bit of a safety net, her brother, the paladin, came over and aided her on her checks, giving her a +2. She made the next check with a roll of 10, and spent an action point for her last skill check, rolling a 10 once again. The boar was calm, and finally, so was I.

So what’s wrong with this picture?  Practically everything, though at least this is a situation where it’s conceivable that botched attempts could make the plan impossible to carry out (fiddle with the wounds unsuccessfully long enough and the boar could be too enraged to befriend even if you eventually get it right) .  Note the way the player meta-games which skills to apply, and when to get the second character involved.  Note also how the player was channeled into a solution that involved making skill rolls, even with two characters that have magical healing abilities that would have made perfect sense in terms of game-world and genre logic to use at that point.  I’d also be inclined to hold it against the system that what seems to drive the tension is the roles themselves, rather than, say, the fact that a fight is raging all around them and they’re spending time fiddling around with the caged animal while their companions are fighting for their lives.

Now imagine playing out the same scene without being anchored down by the Skill Challenge system.  There’s nothing they accomplished that couldn’t have been done by straight role-playing, unless you just don’t trust the GM.  The only difference (besides the fact that the player might have made different choices about whether to involve the second character or use magical healing without the meta-game considerations) is that the task would be open-ended…the player would have an additional decision to make if the first two turns didn’t succeed about whether to continue (and perhaps up the stakes by using special abilities) or give it up and help with the fight.  All the Skill Challenge system does here is dumb it down.

Let’s look at another example, this one a bit more elaborate, and told from the point of view of a GM designing a Skill Challenge.  Again, I’m not meaning to disparage what At Will wrote, on the contrary I’m using this as an example because I think it’s a good Skill Challenge, that would be better if you dumped the whole Skill Challenge mechanic.

You have to go read it, because otherwise I’d end up quoting the whole thing.

So, now that you’ve read it, here’s what’s wrong.   There is, as I believe is typical, no justification whatsoever for the mechanism itself.  There’s just no earthly reason that failing to remember something about the history of the area, or to have a flash of insight, should have any bearing at all on whether you can intimidate a child into telling you what you want to know.  But if you try and fail at those two things as well as using Diplomacy on the villagers, you’ve got no reason to talk to the child…it becomes impossible for you to get information out of him.  The fact that Intimidate is the only skill usable on the child is another failing, in my book…it doesn’t matter if one of the characters is the motherly sort with oodles of Charisma and Persuasion or a spell like Charm Person (or whatever the 4e equivalent is)…that approach is something you can’t even attempt.  In the comments At Will makes it clearer that he’d allow you to use appropriate magic to bypass some or all of the Challenge or to roleplay some of the encounters (but he’d make whether that accomplishes anything contingent on the die-roll…all the talk is just window-dressing), but what advantage is there to even using the Challenge mechanic then?

Here’s how the Old School would do it (in this case, Old School refers to all the way back to 3.5):

Kobolds have kidnapped a group of children from the village the PCs are currently in.  The players must gather information to find out who did it, and find their way to the location in order to rescue the children before the kobolds can sacrifice them to a dragon.

If you talk to the villagers, they don’t know much, but they know that they miss their children and when they went missing.  They were out in the fields…; Someone does report that they thought they saw more children than normal playing for a while…

One child who is still in the village seems to know something. The child doesn’t want to confess because he’s afraid they might come after him if he tells someone, but if the adventurers can persuade or intimidate him somehow he saw some creatures that look like small dragon-people.

Somebody who knows this area and its history will know that this area has not seen too many hard times, but goblins and kobolds can always be a menace. Goblins haven’t been seen around this area for some time…

Examining the tracks in the field, and near some of the houses,  reveals that fairly small creatures kidnapped these children, and they headed North out of the forest.  Appropriate knowledge will further reveal that  creatures are most likely kobolds judging by the tracks.  Appropriate knowledge recalls that kobolds favor mountainous terrain.

Gaining a good overview of the territory (by climbing up to a advantageous spot high among the trees, flight, clairvoyance) grants a good view of the terrain, and suggests three areas that the kidnappers might have gone: a mountain slope, a ravine to the South, or into the dense center of the forest.

And so forth.  Essentially, all the work that went into devising the Skill Challenge is valuable, and can indicate the kinds of information that you can get by talking to the various people and examining the scene.   What’s the matter with Skill Challenges is the mechanism itself, not the putting extra thought and care into how various skills and abilities can be used in the scenario.  You want to strip out extraneous requirements that only specific skills be used (except perhaps noting things that could give a particular approach an advantage, such as so-and-so being a coward and easily intimidated), and beef it up with additional information so that they players themselves can reason about the information they uncovered (the combination of the tracks heading North out of the forest and the overview of the area showing only one likely place to the North, or the overview revealing the mountain and the knowledge check revealing that kobolds like mountains).  You don’t want to artificially prolong it, by requiring a certain number of dice rolls before they can proceed–if they think they’ve figured out that it was kobolds and they’re likely to be in the mountains, by all means let them head for the mountains to take a look around.  You also don’t want to cut them short by saying, oops, you’ve failed, no point in trying to gather more information or recall anything further.  You most emphatically do not want to discourage them trying oddball or long-shot approaches by penalizing them if they fail.  If you want to put them under time pressure, then attach amounts of time that it takes to try different things, and let them decide which they are willing to spend time on and whether they can split the tasks up more efficiently, don’t just arbitrarily rule that failing to recall a fact about kobolds chews up just as much of the day (moves them closer to failure) as canvassing the neighborhood and talking to all the farmers.   For role-playing purposes you can easily get everything that is good about them, and avoid most of the bad, by dropping the Skill Challenge mechanic altogether and just using the list as a guideline to the kinds of things the players can roleplay out.

Skill Challenges take what was a reasonable idea, of examining a situation and making note of all the obvious (and some less obvious) ways that you could use certain skills to obtain information, advance your agenda, or solve a problem and formalizes it into a mini-game that basically piths all that was good and fun about roleplaying to fit it within the framework of a board-game.  The entire Skill Challenge system of collective accomplishment and punishment pushes the players more strongly than ever before into treating the party as a Borg collective of drones with various specializations controlled by a hive-mind, rather than a group of individuals with their own psychologies and approaches to life.  Think about it. There is simply nothing in-character that could be said by one character to another as to why he shouldn’t attempt to recall what he learned about kobolds until some other character elsewhere has either succeeded or failed in picking a lock or climbing a tree (or vice-versa).  Skill Challenges make ordinary role-play harmful to the party, and for no reason at all other than it to satisfy the arbitrary strictures imposed by the mechanics.  If nothing else, keeping the Skill Challenges and dropping the “before M failures” clause would be a step back towards making them at least compatible with role-playing.