Color me skeptical.
<snip>
I think power corrupts. The thing that encourages people to behave the worst is by giving them power.
Since we're playing in an imaginary world, everyone's power is pretty much infinite by default, so the way to counteract that is to either have a DM who places a lid on everyone else (and is himself limited by the need to please everyone enough to keep them in his game), or by having a narrativist mechanical system for attaching consequences to behavior.
My experience is that the participants' power is
not infinite. Their desire is to have a pleasant time playing a fantasy RPG; and their power to do that is confined by such things as their knowledge of the rules, their good relationships with other players and the GM, etc.
Players who (for instance) have their PCs go nuts, kill everyone in the village and rob them are no different from GMs who have 1st level PCs enter towers guarded by ancient dragons, or who declare "Rocks fall, everyone dies." They're not contributing very much to the play experience, and either leave the game or correct themselves.
Combat is supposed to be easier than diplomacy. The nonviolent way is hard. A good character chooses it not because it's the easy way, because it's the right way. If it succeeded reliably, everyone would be doing it.
I'm not sure I agree with you with respect to the fiction: negotiation is sometimes seen as
morally harder than combat, because it requires forsaking honour and vengeance, but I'm not sure it's actually harder.
Nor am I sure that combat "succeeds reliably" in fantasy fiction. For everyone but Conan in the Conan stories, combat seems a pretty bad choice. Even in LotR, combat doesn't really succeed reliably. For instance, the Battle of Pelenor Fields was only one because Aragorn brought the Army of the Dead, and that was a diplomatic victory.
As was discussed in
this thread, I prefer that the choice of peaceful or violent means reflect the players' views, in their play of their PCs, as to whether or not violence is morally warranted, than that they reflect choices about mechanical efficacy within the game. (See especially around post 23.)
OK, what if the "authority" making that decision isn't the DM at all but the written adventure she's running? You-as-player will not (or certainly should not) know the difference between a) the DM arbitrarily assigning a DC 75 to the Dip. check required to get by the chamberlain and b) the written module stating "the chamberlain will not willingly allow the PCs to speak with the king under any circumstances as he knows the king is planning to assign the party to raid the keep that houses the chamberlain's secret ... (etc.)" - to the PCs the end result will (or should) appear exactly the same.
And I also pose this same question to pemerton as on the surface it appears that in his game the DM is not allowed to have any secrets at all.
First, I don't run modules in that way. So I see no difference between GM and module-writer - as I play, it is the GM who is taking material and ideas from the module and actually putting them into play. To put it more bluntly, if the module writer authors a crap scene or a railroad and the GM then implements that, the upshot is on the GM.
As to mystery, I discussed this a bit upthread:
I don't think the indie style is especially well suited to traditional mystery play, precisely because the withholding of information that is part of standard mystery play is deprotagonising of players for the reasons I've already mentioned.
An interesting evolution in indie style around this very point can be seen in a debate between Jonathan Tweet and Robin Laws in the Over the Edge rulebook - Tweet talks about techniques for keeping secrets from players, and Laws talks about the benefits to play of having everything out in the open at the table even if the characters don't know, such as possibilities like irony, or ingame unknown but at-table appreciated interlinks, and generally everyone just having a good time laughing at the troubles their and others' PCs are getting into.
And flipping it around - I don't think it's a coinincidence that CoC - the paradigm mystery game - is seen as a poster child for high-quality GM authority, GM force, non-player-driven RPGing.
Now all of the above said, there are techniques that you can use for mystery or intrigue RPGing in indie style, as TwoSix indicates;
In indie-RPGing intrigue-style play, the GM makes up the details of the intrigue as opportunities accompanying player (and PC) success, or as complications accompanying player (and PC) failure. So although, in the fiction, the PCs are discovering things, at the table the players are not discovering something worked out by the GM in advance (at least, not worked out in detail). The GM might have general parameters in mind, or these might be set by some other considerations (past revealed backstory, for instance; or genre considerations), but the details emerge in play, and their function is to keep the players on their toes and spur them to keep going.
(In this approach, there probably won't be many GM-introduced red herrings; these are supplied by the players' own changing speculations about what might be the truth of the situation, most of which turn out (retrospectively) to have been red herrings as the possiblities are narrowed down by more and more backstory becoming solidly established because "revealed" (ie narrated into the shared fiction) as part of the process of resolving action resolution. (Of course, this takes us back to "Schroedinger's NPCs" or, more generally, "Schroedinger's backstory". Keeping the backstory loose and flexible until it is actually solidified via narration is a pretty common "indie" technique.)
For actual play examples of GMing mysteries and intrigues in indie-style, see
these three threads.
I have also talked about my concerns around secret backstory changing the fictional positioning in ways the players can't anticipate. I'm not a big fan of this. Typical examples are things like the Chamberlain secretly hates clerics of St Cuthbert, one of the PCs is wearing a holy symbol of St Cuthbert, and so the Chamberlain is hostile for no reason that the players can easily fathom. (The example of the duke who can't be intimidated, that I discussed 5 or 10 posts back is another instance of this.)
I don't regard this as completely out of bounds, but I think (i) that the presence of the secret backstory creating fictional positioning that the players are unaware should not be
determinative of the scene's outcome in-and-of itself (so the Chamberlain shouldn't just walk away because of his Cuthbert-hatred); and (ii) that the players should have the chance to uncover the secret backstory, and therefore potentially change their PCs fictional positioning in response to it, via the ordinary action resolution mechanics as they are engaged in dealing with the scene (in the duke example we see the possibility of an Insight check - not the most sophisticated way there ever was of handling this sort of thing, but better than nothing).
The three threads linked in the quote above give examples of secret backstory being used in play, and how I handle the balance between "big reveals" and giving the players the information they need to meaningfully and effectively engage the scene.