RAILROADING
All of the dissociations of the skill challenge mechanics arise because, for any given problem, there are multiple possible solutions. It is likely that each of these solutions will require a different set of skills.
For example, if you wanted to solve our “get into the castle” problem you could try:
(1) Diplomacy (to bribe the guards)
(2) Gather Information (to find out who works in the castle) and Diplomacy (to get an audience)
(3) Stealth (to reach the walls) and Climb (the walls)
(4) Architecture (to find out about the secret door), Stealth (to reach the door), and Thievery (to pick the lock on the door)
So here we see four possible solutions, involving completely different skill checks: 1, 2, 2, and 3.
If the DM sets the skill challenge to be 2/1 then the skill challenge mechanics will fail to report success despite success being achieved in one case and report success before success has been achieved in another.
Similarly, in the case of the fourth scenario a failed Architecture check would seem to bollox the entire effort and the skill challenge mechanics would seem to accurately report that. Of course, this is entirely acccidental… and not accurate, either. As we can see, even though solution #4 is no longer an option, the other three options haven’t become impossible just because of the failed Architecture check.
One way to solve these problems is to simply design the skill challenges so that they’re railroads. This, based on their web sample, is WotC’s solution: Instead of merely setting a goal (“get the duke to help us”), their skill challenge specifically tells the players how they will achieve it (“by getting the duke to trust you”).
Explaining why railroading is a Bad Idea(TM) is beyond the scope of this essay. But it’s a Bad Idea(TM).
RULE 0 FALLACY
You can also work around some of these problems by invoking the Rule 0 Fallacy (“this rule isn’t broken because I can fix it”). In this case, when the system is inappropriately reporting failure or success, the DM should simply ignore it.
But if the mechanics are so broken that we need to frequently ignore them, why are we using them at all?
UNFUN WITH PROBABILITY
You can also try to remove dissociations from the system by varying the number of skill checks you require to accomplish a particular task.
For example, let’s consider our castle break-in skill challenge again. Let’s say that the DM sets it as a 4/2 challenge and the PC decide to sneak up to the walls and then climb over them. The DM has them make a Stealth check (1 success) and then requires the PC to make 3 successful Climb checks. If the PC has a 50/50 shot of making the Climb check, then they only have a 12.5% chance of climbing the wall.
Now, let’s change the scenario: One of the PCs decides to distract the guards with a Diplomacy check while another PC sneaks up to the walls with a Stealth check and tries to climb them with a Climb check. The DM has them make the Diplomacy check (1 success) and the Stealth check (1 success) and then requires the PC to make 2 successful Climb checks. With the same 50/50 shot on any given Climb check, the PC now has a 25% chance of climbing the wall.
For some reason, talking to the guard has made the wall easier to climb!
You see similar probability artifacts arising out of the skill challenge system even if you aren’t padding the number of required checks in order to fulfill the arbitrary requirements of the dissociated mechanics.
For example, if you get to the point where you just have to make a single Climb check in order to succeed at the skill challenge, the difficulty of successfully climbing the wall will depend on how many failures you’ve accumulated getting to that point.
If it was a 4/2 challenge and somebody in the group failed on that Architecture check to see if they could find out about a secret door, then you’ve only got one shot at it: If you fail the Climb check, you’ll have accumulated two failures and the skill challenge will fail. With a 50/50 shot, you only have a 50% chance of climbing the wall.
But if your group never considered attempting that Architecture check, you’ve still got a failure to burn. If you fail the first Climb check, you’ll only have a single failure and will be able to try again. With a 50/50 shot, you now have a 75% chance of climbing the wall.
You’ll note that, in both of these cases, the scenario is identical: The PCs are unaware of the secret door (either because they never thought to look for it or because they didn’t find it). But in one scenario they have a 50% chance of climbing the wall and in the other they have a 75% chance of climbing the wall. Why? Because of a mechanical artifact that has absolutely nothing to do with the game world.
That’s the definition of a dissociated mechanic.
Some would argue that this type of probability shift is irrelevant because the PCs will only go through the skill challenge once: Either the wall is a 50% wall or it’s a 75% wall for them, it’s not both. But this sophistry ignores the possibility that this same wall can end up being part of many different skill challenges for the same set of PCs.
And do we even need to discuss why it’s ridiculous for a wall to become unclimbable by everyone in the group just because the guy with the lowest Climb bonus failed his check?
THE BIG PROBLEM
Okay, so we’ve established that the skill challenge mechanics are dissociated. Why is that a problem?
Because, unlike the Wushu mechanics, the skill challenge mechanics don’t seem to actually be accomplishing much. You’re making all the sacrifices inherent in the use of dissociated mechanics, but you aren’t gaining anything in return.
Most notably, the skill challenge mechanics aren’t giving the players any meaningful narrative control. The flow of gameplay is unchanged. In 3rd Edition, for example, gameplay looked like this:
(1) The DM describes a problem.
(2) The players propose possible solutions.
(3) The DM determines whether the solutions will actually work and asks the players to make the appropriate skill checks to resolve them.
With the 4th Edition skill challenge mechanics, gameplay will look like this:
(1) The DM describes a problem.
(2) The players propose possible solutions.
(3) The DM determines whether the solutions will work and asks the players to make the appropriate skill checks to resolve them.
Nothing has changed.
The only concrete benefit of the skill challenge mechanics, as far as I can tell, is that they codify a way for rewarding XP for overcoming challenges. This doesn’t even begin to justify the problems that come with dissociated mechanics, in my opinion.
It takes some real effort to find those Worst of Both Worlds solutions. WotC seems to have really nailed it with the skill challenge mechanics.
On the other hand, you could certainly adopt a system very similar to WotC’s skill challenge system and use it to pass a lot of narrative control into the hands of the PCs. I’m not personally convinced that mixing that type of player-driven narrative control with a combat system that doesn’t even begin to feature player-driven narrative control will make for a particularly effective game (it sounds more like mixing oil and water to me), but it’s certainly not a bad idea to experiment with.
THE OTHER PROBLEM
Since I’m discussing skill challenges, I might as well mention the other major problem they seem to have: From what we’ve seen so far, the skill challenge system can’t handle any inputs which aren’t skills.
For example, what happens if I cast a fly spell instead of using a Climb check to climb over a wall? Should that count as a success for the skill challenge? Multiple successes? Or does the fact that I’m flying mean that I now have to make an extra skill check somewhere else? And doesn’t that create yet another weird, dissociated disconnect between the mechanics and the game world — encouraging me, as it does, to potentially climb a wall and make a Climb check even though I’m wearing boots of flying?