Choosing to do this allows you to use Bluff and Streetwise in the bandit avoidance skill check.
For the sake of argument, I am going to branch this out into two different SC. In the first SC, the players are doing what the DMG 1 suggests -- they are picking skills and rolling dice. In the second SC, the players are doing what VB suggests -- they do not even know that they are in a SC.
SC1: Either the PCs have no one to ask (as there is no one in the set-up description), or their decision to use Bluff and Streetwise "cause" someone to appear. Or they have nothing to do with the world itself, but are simply the highest stats the PC has.
One PC chooses to bound like a grasshopper. How is this relevant? Perhaps the bandits are afraid of obviously insane people. Perhaps all that bounding makes him tired, so that he rests at just the right time, so as to miss the bandits. Perhaps the nose picker sees an "omen" in the shape of his booger that tells him to go the other way.
The only reason not to use a skill is a "failure of imagination" on the part of someone.
Unless the world around the PCs, and interaction with that world, is a prime determinant, player choice is compromised. Highest stats are all that matter. This is, as I understand it, the majority of where Ariosto's complaint lies.
SC2: The players just think that they are asking around. The skill checks might even be made by the DM in secret.
Asking around might lead the PCs to believe that the best way is to go across the Running River and take the Other Road. Crossing the Running River might require swimming or wading across a ford. It might include other encounters. Perhaps enough people have taken this route that the bandits now have a scout along it that the players must slip by or deal with.
In this case, player choice is very important. Players are not simply given information, but must decide how to act upon it. Moreover, there is not a static DC for the skill challenge; what the players choose to do determines the various DCs, and the various numbers of checks they must make, in order to succeed.
The problem, of course, is that this is no longer a skill challenge as defined in 4e (at least not as initially defined). There is little difference between this and freeform play.
Worse, the residual effect of designing skill challenges to move PCs along from one combat encounter to the next might rear its ugly head -- the poor neophyte DM cuts short the drama of the action because, along the way, the PCs have made three successful skill checks. No longer is there a need to swim/ford the river, or avoid that bandit scout.
If one chooses to forgo the structure in order to fulfill the dramatic potential of events, then why is this considered a SC at all? This is, from what I can gather, the other part of Ariosto's complaint.
In neither SC1 or SC2 is the result better than that from freeform play, and in both cases it may be substantially worse. The less the "skill challenge" resembles a "skill challenge", the better it is in terms of actual play.
Where, then, is the benefit of the mechanic?
And this is just showing the failure of logic in this argument. Being able to imagine a decent way to use a skill in a challenge is not even close to being able to or required to incorporate them all.
Again, we are given two choices:
In SC1, you can say "No" to any skill (and saying No to Streetwise and Bluff makes just as much sense as saying No to Acrobatics and Perception). It is entirely arbitrary. Ultimately, this is DM Fiat thinly disguised as something else. Which is why you and VB cannot agree what skills should qualify.
In SC2, appropriate skills are chosen by examining a world which presents opportunities for skill use, again, by DM Fiat. If we follow VB's advice, the players don't know that they are involved in a SC, allowing the DM to transition smoothly to freeform play as soon as possible....indeed, quite likely before the SC is resolved if the DM is a good one.
This leads us to another observation:
The best skill challenges are not resolved as skill challenges.
In the case of SC2, the mechanic should be there only as a "hand holding" measure to guide the DM and players back into freeform play once they've become stuck. Unfortunately, AFICT, this is not how the mechanic is presented (although I would be happy to hear otherwise).
It is certainly not how this supposed "holy grail" of game mechanics is seen described here on EN World.
*******
The problem, as I see it, is that prior to 3e, D&D hasn't had a good skill system. This is IMHO, of course, but I would go so far as to say that good skill systems were few and far between in RPGs.
In the 3e PHB, Swim was given a "three strikes and you're out" rule. You didn't drown on the first go. And, Behold! Rolling more than once before actually failing (or even actually succeeding) was demonstrated to be more fun/satisfying than just rolling once.
Then, when UA came out, complex skill checks were added. Effectively, these were using the "X successes before Y strikes"; the same formula, in fact, that would be rebranded as Skill Challenges in 4e. Effectively, swimming across a body of water writ large on the D&D rules.
The problem with the mechanic, overall, is that it isn't dynamic enough. Other systems, like "Degree of Success" (DS) and "Opposed DS" in
Tournaments, Taverns, and Fairs, gave as good, or better, methods...and were OGC to boot. (DS already appeared in the 3e PHB in the core concept that some skills, when failed by a certain amount, had negative consequences; DS merely grants various DCs for various degrees of success....exactly as does Bardic Knowledge or Knowledge skills in 3e.)
Ideally, any form of complex skill check should map in a way similar to combat; failures make overall success more difficult, successes make the overall task less difficult. Decisions that the players make
now should lead directly into how difficult subesquent checks are, and the result of each success and/or failure should change the situation.
In the example above, playing through Streetwise/Bluff should not only affect whether or not they meet the bandits, but (in some cases) how the bandits react to them subsequently.
Let us look at Bluff.
The use of Bluff that billd91 suggests assumes that the PCs can, by talking to the farmers/whatever, communicate with the bandits. I.e., by making the farmers think they are tough, they make the bandits think they are tough. It is implied that some system of communication must in place between the bandits and the farmers.
Well, then, failure may not just be another failure to chalk up. It might mean that the bandits have better information on the PCs than they did before.
If success has any consequences apart from frightening farmers, failure must also have consequences apart from not frightening farmers.
The DM must consider how each skill attempt, no matter its outcome, will affect the situation as a whole. And, to make the whole seem like more than just an exercise in die rolling to get from point A to point B, the DM must include the means for the players to discover that their actions have larger consequences.
Again, the SC mechanic becomes nothing more than (at best) a hand-holding measure moving into freeform play, where no complex interaction is resolved by a single die roll....or even three. At worst, it fools the DM into thinking that a situation has been fully presented, and player decisions taken into account, when the mechanic serves instead to dismiss those same decisions if followed verbatim.
**************
Another problem with SC1 is that, should the players choose to harass and intimidate farmers, then they can presumably know whether or not they succeeded at the skill challenge before hazarding the road. If they succeed, they go down the road and get where they are going. If they fail, they do not go down the road, and, unless the bandits teleport to their new location, their failure can still be turned into a success. This is a very different animal than in freeform play.
Of course, our SC1 example has the players doing things like determining whether or not their Perception checks are successes. How likely is it that the players would actually know this, apart from learning what information a successful check might garner them?
Couldn't/shouldn't/wouldn't the DM conceal the results until the PCs have hazarded their outcome by actually going down the route they choose?
*****************
Skill Challenges, as presented, are a step toward integrating skill checks into complex interactions with the game milieu. They are a step towards complex skill use in freeform play.
However, they are not there yet. They have a long way to go.
Their initial design was hampered, IMHO, by a design philosophy that stongly pushed combat as the sole important complex mechanical interaction in the game.
They are currently hampered, IMHO, by trying to make current implimentation consistent with that of the DMG1. That's a pretty speculative opinion, mind you, as I have not read the DMG2, but it is what I gather from secondhand sources (including the examples & discussion thereof in this thread).
I suspect that these opinions will not be popular right now, but will be "widely acknowledged" by the time 5e is announced. Indeed, I suspect we will learn then that they are "widely acknowledged" now.
RC