The question was why to integrate SCs if you can roleplay through it. I will name 3 good reasons:

First, in my experience it helps the DM designing an encounter or adventure A LOT and is an EASY tool to hand out XPs and other rewards. That is true for many 4e mechanisms that sound awkward from a players perspective.
4e is a bless for GMs mostly.

Second, it can improve RP, because (as i understand it or as we play it) everyone HAS to contribute in a skill challenge. So you avoid the situation that the group lets the “experts” solve certain problems alone. I see the problems this mechanism has, but the point remains that it supports roleplaying by providing uncertain players with a kind of frame for their actions.

Third, the SCs strenghen the skill uses.
In 3.5 a fighter would hesitate to spend a feat for improving skills, in 4e you are not punished but rewarded for doing this. Using skills is much more important and rewarding now.

Your main objection was that failure would stall the game or prevent players from doing certain things. I still don´t see how, because no one forbids you to try other ways. You can still climb a tree and search the goblins lair even if the corresponded SC failed. But you have to expect other obstacles on your way.

In earlier versions it was more difficult for the GM to determine the success or failure of an action. Now he can rely more on game mechanics.

You say there shouldn´t be the possibility of “global failure”
I think you overestimate the effects of a failed skill challenge.
Failure means there are no rewards, but more difficulties and obstacles, thats it.

And why should anyone roleplay or make skill checks when you cannot fail ? Failure is part of every challenge and even if another member of your group produced the failure: where is the difference to older versions ? You can be the best diplomat in town, if the dwarf next to you insults the duke, your efforts could still be wasted, if you are in a SC or not. (But maybe the consequences for the dwarf would be more serious) If you found some advice in a library (which you would rather than find nothing just because your thief failed to climb a tree), it still could be useful (if you rolled a success), but wouldn´t be enough to find those goblins.
So i can´t see why the M-Strikes are that much of a problem.

@r_b_bergstrom – if that was all there was to their design intent, then they really screwed it up. Guidelines as to xp for skill rolls depending on degree of difficulty and whether there’s danger or consequence attached to their failure…or just general guidelines for xp awarded for all non-combat stuff together, would have been a lot simpler, without all the negative consequences. A key difference between HP and Skill Challenges is that HP is merely an abstraction, while SCs are a mini-game. So it’s not just about choosing a level of detail/realism that you’re happy with, but the consequences (intended or otherwise) of how decisions the mini-game translate (or fail to) into the game world. I’m pretty sure, after reading The Alexandrian’s posts, that they might have had design goals for the Skill Challenge system, but they didn’t do any real analysis or play-testing. Which is a pity, because I don’t think the state goals are crazy.

That said, it’d be very difficult (and require a book of it’s own) to balance all those possible consequences of every individual action. I believe one of the goals of the skill challenge system is to provide rules by which skill tests can be rated and compared against each other and against combat encounters to determine relative degree of difficulty and appropriate XP awards.

4E’s Skill Challenges system is attempting to do for problem-solving, skill use, puzzles, and role-playing what 3.X had done for traps.

In other words, the designers want to give out XP for using your skills, but they don’t want it to be a flat rule (25 xp per skill roll) in which the danger or difficulty involved had no impact. Nor do they want it to be a tremendously complicated system, where ever possible factor and consequence needs an entry in some gigantic reference section. Yet within those constraints, they also needed some element of danger and/or chance of failure. Without some way of doing so, Skill Challenges would become a source of easy no-threat XP.

So they hit on an abstraction – that being the “M-Strikes Rule”, and the Complexity Rating – to determine when you fail and the consequences kick in.

Just like hit points, M-Strikes isn’t realistic enough for some people’s tastes. As I said, I’m one of those folks who’d rather have detailed consequences stemming from specific actions rather than an overall out-of-character ticker. I wouldn’t have made the decision that 4E’s designers did, but I can still see why they made that call, and why they felt it was beneficial to go there.

There’s nothing wrong with a good mechanical toy inside a game system; combat subsystems often are, as are involved character advancement systems. It’s often satisfying to play with an intricate toy just for its own sake. This one just isn’t done very well if its meant to complement and encourage playing a character as a role rather than an avatar. (A much better non-combat mechanical toy is The Burning Wheel’s “battle of wits” debate mechanic.)

The other reason I think the too-common “Skill Challenges are fine if they’re written right” rejoinder is silly: If the subsystem is supposed to make creating and running these properly so easy, then why are so many people “doing it wrong”?

The complaint that Skill Challenges cut across the grain of the game world has been similarly made (and similarly articulately) over at The Alexandrian: Dissociated Mechanics Part 5: Skill Challenges, Skill Challenges: WTF?, and Playtesting 4th Edition Part 6: Skill Challenges. You might find some added insights there if you haven’t already read them.

Part of my complaint is that the Skill Challenge mechanic discourages RP by presenting mechanically right and wrong ways to approach challenges, even when from the point of view of the logic of the game world there’s no justification for the preference of one approach over the other.

I’m at a loss here. I think what you are saying that in some case some actions might be more difficult then others while this is not true in the fiction. If this is the case, why not make those difficulties the same?

Or are you saying that if there is a difference in skills and equally difficult tasks, then it only makes sense to use the better skills? I’d say this is true in situations where skill challenges are appropriate; that is, inherently risky and challenging ones. If skills are used to help others, than roll them as help.

But, what exactly are you getting at? I fear I misunderstand you.

On helping: My guy is triyng to pick a lock, you roll the relevant lockpicky skill to help, on success your character has encountered a similar lock before and can give useful advice, on failure not so much. (If picking a lock is a skill challenge, it better be pretty complicated one with in-built traps that trigger if something is messed up.)

Skill challenges are about actions, not skills.

Skill challenges are structured roleplay.

http://at-will.omnivangelist.net/?p=122

I talk about these at length at the site, so don’t want to rewrite it here.

Definitely take what I’m offering in the skill challenges as suggestions. The results and actions are meant to be roleplayed out, and not just read off the sheet.

I personally find that skill challenges involve quite a bit of roleplay and improvisation from both me and my players –one of the things is not letting players off the hook. A skill role is not the action that there character performed. That is what they give you, and that’s what you give them feedback on.

Just wanted to clear some stuff up. Thanks for the link, and I would appreciate your comments on some of the work I’m doing at the moment.

I know you’re not sold (which is OK, because I’m not selling! :)), but I think we can have some interesting conversations on this.

In any case, I don’t see how it could help with the aspect that if those checks of 18, 17, 16, 14, and 13 belonged not to separate characters, but say two or three characters who have different skills/actions they could take the “correct” actions to perform are the ones that checked against 18, 17, and 16, not the ones that the characters would choose based on game logic and personality. Part of my complaint is that the Skill Challenge mechanic discourages RP by presenting mechanically right and wrong ways to approach challenges, even when from the point of view of the logic of the game world there’s no justification for the preference of one approach over the other. Granted there always exists the possibility of tension between the mechanically optimal action and the RP-driven one, but it’s bad to exacerbate it. In most systems, going with the RP action costs maybe some time and resources, not risking global failure–and when it does risk global failure there is often strong pressure from others at the table to violate character and go with the optimal action.