So the first principal is that it's not your fault you failed even if you roll low... interesting, I find this weird since even powerful protagonists in fantasy fiction a times mess up through their own hubris, flaws, etc. Are there usually limits on this?
It's generally a principle about narrating the task the player has declared for his/her PC.
So, for instance, in combat: if the fighter misses with an attack roll, it is narrated not so much as "you swing and miss" but rather as "your swing would fell any normal person, but the troll swats your blade away with its brawny, steel-hided arm".
Hubris and flaws would tend to come in, in this style, via external circumstances. For instance, the hubristic thief doesn't fail to pick the lock, but is knocked unconscious by the gas trap that s/he triggers; or perhaps the guards arrive just as the lock clicks open.
Is part of fail forward discussing a change in outcomes with the group?
Not necessarily. In the case I described, before framing the PCs into the jail scene I wanted to know which players (if any) would prefer to start a new PC, in which case the old PC would really be dead.
If you want to locate the discussion under a methodological label, I would put it under "scene-framing" rather than "fail forward" - namely, one way to decide what sorts of scenes to frame, as GM, is to discuss with the players what they want. Managing discussion vs uniateral decision-making is a broader part of managing pacing, tension etc at the table. Just after a "TPK" there is generally no need to maintain tension or a rapid pace, and so it's a good spot for a conversation.
Were the success and failure parameters set before he entered this SC? If so what were they?
Success conditions were - I gave the player the option to keep heading for town or to try and return to the tower, and he chose the latter.
Failure conditions were implicit. I discussed my approach to this on some recent thread, maybe even this one. Luke Crane, in the Burning Wheel core book, stipulates that failure conditions must always be stated in advance. But in the Adventure Burner (which is something like a GM's guide for BW) he says that, in his actual play, he often doesn't stipulate failure conditions because they are implicit in the situation.
My approach is similar, in that sometimes I specify failure conditions but often I prefer to let the framing of the situation, and the player's action declaration in that context, carry them by implication. In the context of riding a flying carpet trying to escape from hobgoblin wyvern riders, for instance, where you are dodging and weaving (Acro), throwing flasks of elemental fire at them (Arcana), etc, I think the implicit failure in most systems is a crash or capture - and D&D doesn't handle capture of a single PC all that well (because of its party focus) - and so crash landing becomes the implicit outcome.
Relating this to illusionism (or its absence): the player can
see how the crash, as an element of the fiction, has been derived from the actual process of play. He could see his failed dice rolls in the skill challenge, could see me totting up the failures, and in his ensuing action declarations was responding to my narration of the consequences of the first two failures (from memory, some of the elemental fire caught him in its blast and singed his carpet, which was what helped precipitate the crash).
So then is illusionism more about being dishonest about methodology as opposed to creating the illusion that choice matters on the result?
I think "dishonesty" is too strong - I think the word "covert", used upthread by [MENTION=6696971]Manbearcat[/MENTION], is preferable.
The reason why I say this isn't terminological - it's not about trying to keep the jargon pure.
The reason is that - especially in this thread - "illusionism" has been used to describe a set of GMing techniques and approaches, rather than a broader social dynamic at the gaming table. And "covert" is a good for techniques, whereas "dishonest" tends to have those broader social connotations.
And I would say it's not just covertness about methodology per se - that's too broad for what people upthread have been describing. It's about covertness in respect of the relationship between (i) player action declarations and the mechanics that surround them, and (ii) the GM narrating changes in the fiction as a result.
The two main forms it takes that were in my mind in my list of examples above were the techniques around dice rolls, and the techniques around "secret backstory"/metaplot.
With the first set of techniques, the illusion is created that rolling the dice for action declarations will affect the fiction - whereas in fact it won't (the GM has a predetermined answer in mind). "Say yes or roll the dice" is meant to dispel this form of illusionism. Consider how this applies to looking for a secret door: the players announce "We search for secret doors, tapping and rapping and prodding and pulling, etc". The GM replies, without rolling or calling for a Perception/Search check "OK, you find one . . ." In this case, the GM doesn't pretend to the players that there was a possibility of an alternative fictional state for the game, where the PCs
don't find a secret door.
(A variant of "say yes or roll the dice" is "say no" - a form of hard(-ish) scene framing/backtory enforcement. It comes up in my game quite a bit: the players say "We [ie our PCs] search for traps/secret doors/loot" and I say, without rolling dice or calling for a roll "You don't find anything".)
With the second set of techniques, involving direct manipulation of the fiction, the illusion is created that the effects upon the fiction of resolving players' action declarations will matter to the broader state of the shared fiction. It is a type of covertness about the stakes. In my experience of reading publishes adventures, this variety of illusionism is generally advocated to mitigate against deep changes in the gameworld (eg if the PCs don't stop the evil-so-and-so, eventually another NPC will a day or week or so later).In my actual play experience, this variety of illusionism is used by GMs who (for either practical reasons or aesthetic reasons) don't want to lose control of the campaign world, or have it go in unanticipated directions.
What I'm distilling from the above explanations is, as soon as the player/s choice involves dice and you ignore it, that is illusionism, but if the player/s choice does not require dice then everything is hunky-dory.
Given that I (and not only me) have also talked about techniques that involve manipulation of the fiction (especially bacstory) via the GM, I don't know why you say this.
Based on what has been described, any/all decisions taken by a player without dice are not as important as the ones which rely on dice. If the choice exists to take the left or right passage but leads to the same result why bother even offering the choice to the PC?
I don't think I follow the question.
The only person who talked about going left or right is me, and I explained what role that might play if it is not a genuine decision, namely, it provides colour. It's a bit like a player specifying the style of hat that a PC wears: in most campaigns that will not be a decision that has a real impact on any outcomes in the fiction - it is just colour.
But this has nothing to do with dice or no dice. Not all action resolution mechanics involve dice. For example, choosing to use a rationed power (a Vancian spell, an encounter power, etc) doesn't involve dice, but it can still be a meaningful decision because of its resource management implications. (So the meaningfulness is a type of
mechanical/tactical meaningfulness.)
And here is an example of
narrative meaningfulness that doesn't involve dice: the player declares "I pilot the Tower staight down, breaking through the Demonweb" and I, as GM, say "OK", and then set about describing the resultant Abyssal rift that starts sucking in everyone nearby (including, initially, the PC ranger). (Technique-wise, this is an example of "say lyesl or roll the dice." I said "yes", because I did not see that there was anything interesting that might come from a
failed attempt to pilot the tower straight down as the player wanted his PC to do.)
if you want to railroad the characters to a planned encounter just do it and cut time, instead of giving the PCs the illusion that their decision to go left or right matters.
Instead use the narrative of the journey to the planned encounter to produce the colour desired.
There are two things going on here, neither of which I quite get.
First, there seems to be an assumption that someone on this thread is advocating, or using, illusionistic techniques around "go left"/"go right" decisions. But I don't know you think that person is. [MENTION=6668292]JamesonCourage[/MENTION] attributed the technique to me - I replied with a fairly long discussion of how I have handled travel in my game, none of which involves illusions around "go left"/"go right".
Second, there seems to be an assumption that all colour must come from the GM ("use the narrative of the journey . . ."). Why is that? Sometimes it is fun for the players to inject colour, and they can't do that if the GM doesn't give them the chance ("Do you want to go left or right?"). Especially because, sometimes, what began as mere colour can evolve into something more. Now personally I don't find "going left" or "going right" the most engaging and vibrant colour of all time, but then it wasn't my example.
More generally, I think the tendency to identify meaningful choices in terms of "going left" or "going right" is a legacy of D&D's origins as a dungeon exploration game in which the players are trying to beat the GM's dungeon. It is predicated, too, on the players having access to information that makes it possible to choose rationally - detection magic (from spells, wands, swords, potions - these are very common items in the classic game); rumours; treasure maps; etc.
Once you drop that approach to play and associated paraphernalia (does 3E have potions of treasure finding, wands of metal and mineral detection, etc? they're not a big part of 4e), then choosing to go left or right becomes much less significant. If the players can't know in advance what lies down one pathway and what down the other, the choice becomes arbitrary, not meaningful. As I posted upthread, at that point the reason for a GM to stick to the ap isn't to preserve meaningful player choice (it was just random), but rather to preserve the integrity of the backstory (no one wants to narrate a contradictory dungeon).
