Cool.
Well, the question is, what would the players be choosing from otherwise? And what counts as meaningful?
If those are the questions, then you don't address them. I'm busy working/geting ready for my next session, but I've been following this with some interest despite my lack of comments.
I know that Forge theory isn't all that popular on these boards, but to be honest I think that it helps unpack some of these issues and the debates that have a tendency to recur here.
I like Forge theory because its a serious attempt to address the complexities of what makes an RPG work. I dislike Forge theory because of particular claims that it makes that don't for me stand up in the light of my experience.
[In this case, I think the discussion of illusionism is being distorted in the way that Mallus described - that is, too many assumptions are being drawn from classic AD&D play compared to other approaches to play.
As opposed to what, for example? How I played 'Star Wars'? Chill? Call of Cthullu?
As far as I can tell, from a combination of experience plus reading what others have to say, AD&D has two classic approaches to play: either purist-for-system simulationism (here we all are, a fighter, a wizard, a cleric and a thief in a fantasy world - I wonder what our adventures will be?) or a pretty austere form of gamism (here's this fantasy world generated in accordance with all these random tables, that looks almost like purist-for-system simulationism, except that when we take a second look at it we see that the real point of it is for the GM to run the players through the challenges that the game setup poses).
You've immediately jumped into one of my least favorite assumptions of Forge theory - the notion of concrete mutually exclusive play styles. I seriously doubt anything so easily described represents actual AD&D play, and I certainly protest that a game could have both approaches at the same time. You are hunting for a 'real point' to the play as if the game has a single 'real point' rigorously adhered to, and that prevents you from seeing that it might have neither a point at all or, to the extent it has a point, it might well have many equally valid points sharing time and energy in the game.
In either approach, from the subjective perspective of the players there's no scene framing other than the initial "You all meet in a tavern" because from that point on the whole game unfolds according to the logic of those random tables, plus the GM's decisions about how to apply them and interpret their results.
I'm not sure what you mean by that, but I'm tempted to just say 'bollocks'. I don't think its possible to say what sort of scene framing has likely occurred in either of the cases you describe, nor do I think the scene framing tells us much about the game. For example, if my game begins with the scene framing, "You are all intimate associates of one of the princes of a kingdom, and the game begins at a private party being conducted in his chambers", it doesn't necessarily follow that I'm not going to flesh out the world randomly or that I'm not going to treat the castle like a dungeon where the players overcome challenges. The scene "You all meet in a tavern", is itself a sort of scene framing that tells us something about the player's characters. In particular, it's worth noting that 'Dragonlance', which you've sited as being a step away from this style of play begins with the scene framing, "You all meet in a tavern."
The highwater mark for encouraging this sort of illusionsim/railroading is AD&D 2nd edition, which keeps all the mechanical trappings that supported purist-for-system simulationism and austere gamism, but wants to deliver a Dragonlance/Driz'zt experience in play.
The real strange thing about this is that the Dragonlance/Driz'zt experience (Driz'zt especially) was itself created from games run according to something like what you call 'purist-for-system' simulation and austere gamism. The original Icewind Dale Trilogy reads like a recount of a pretty standard AD&D campaign in novelized form. The incoherence in DL/FR results first from wanting to communicate the results of a play experience (the personal campaign of the designers) rather than the tools that built that experience in the first place, and secondly from not having preexisting tools for communicating end results rather than starting points. It's worth noting, that this desire to communicate the wonderful joys of the designers end results, was a desire on the part of the designers that wasn't necessarily in line with the desires of their audience.
There are plenty of people who played DL who did have the 'Dragonlance' experience themselves, but they did so by abandoning the attempt to have the exact same experience in play as the designers, and instead employed the techniques that they had learned from playing AD&D. The result was recognizably DL, but DL in an alternate universe where the war was primarily naval in nature, or where different characters died and had different roles in the ultimate outcome of the story. What you call the 'purist-for-system' simulation can result in an epic narrative, but what it can't do is transmit the same narrative between groups (because it depends heavily on things happening randomly and players making free choices).
The "solution" to the incoherence between mechanics and goals is to have the GM tweak mechanical outcomes to produce goals. This is why I agree with the Forge-ites that AD&D 2nd edition is an incoherent ruleset.
Maybe, depending on what you mean by a 'ruleset', but not for the reason you cite. And note, the 'solution' you cite still doesn't produce the 'Dragonlance' experience. It produces the experience of observing someone else's 'Dragonlance' experience. The incoherence in 2nd edition is a failure to understand that you can't directly transmit the 1st person experience of gaming out a story. It results in modules and even settings that are novelized and hense, noninteractive.
But I have to admit that, for a lot of players, it seems to deliver something they want (or at least are willing to accept).
Really? No, I don't think so. I think that alot of players want the experience of an epic story, but they want to experience their own epic story and much of the reason early to mid 2nd edition TSR produced all this product that didn't sell was that they were trying to transmit their own epic stories rather than providing the tools to produce epic stories on your own. Oddly, I think DL did this latter job much better than the 2nd edition attempts that followed it, because it still for the most part used the earlier adventure framework (dungeons, encounter areas, hooks, narrow-broad-narrow) rather than the read along style that followed it. It also learned better from its mistakes than latter attempts.
It seems that the players who want this combination of story and physics are therefore prepared to put up with the GM exercising either covert or overt power in order to make the simulationist mechanics produce satisfying narrative outcomes. (This is even the explanation for the use of illusionism that Celebrim has given upthread.)
You seem to have completely missed what my stake is here. Why am I trying to define 'soft illusionism'? What do I get out of having this term in the debate? One of the things I've observed about players who have been burned by 'railroading', is that they become very upset at the slighest hint of DM's exerting authority over the story because they fear that it will put them on a slippery slope which ends in the same sort of play experience that they were burned by before where the were robbed of all meaningful agency. These players tend to define games by a binary 'illusionist' or 'not illusionist' paradigm, where 'illusionist' means 'bad'.
But as I observe games, what I find is that there isn't a nice neat binary either-or thing going on. Like most everything, I believe there exists a continium between the two end points. I'm bringing up these 'soft illusionism' examples where the DM shapes the game world for metagame reasons to achieve some goal, but where most observers would say that player choice is not 'meaningfully' being reduced, in an effort to show that first, no game is or even can be completely free of illusionism, and secondly that illusionism doesn't inevitably lead to 'a railroad' because there are always implicit assumptions about the sort of choice/outcomes that will be available in play and the sorts that will not be available.
For example, if - from the very get-go - it's understood that certain decisions will be made by the GM, then the GM making those decisions is not illusionism (because there's no deceit).
Only if those things are openly acknowledged. One way to get around illusionism entirely is for games to openly acknowledge their conciets. The illusion is still there, but because the veil is openly and always peirced, there is no possibility of 'illusionism'. The illusion is maintained as an open conciet, rather than an unacknowledged one. However, this style of play is IME somewhat outside of the mainstream except in a few narrow cases. One example from the thread being the player who requested a certain object and the DM responded by placing the object in the player's path. That isn't illusionism, because there is an open transaction between the player and the DM. The player has peirced the illusion and tacitly both sides have acknowledged it. Some game system openly encourage these above board transactions. Some even create game resources for managing these transactions and even make that the mechanical focus of play.
In the same way that no one regards it as railroading for a typical D&D campaign to begin with "So, there you all are in a tavern waiting for a patron to show up" - this isn't railroading, it's just the GM kicking things off - so if it's agreed that the GM will do this sort of thing from time to time then it's not railroading. It's just the GM doing his/her job.
I don't know that I would go so far as to say 'no one' regards it as railroading. Probably someone out there is going, "I protest. My character would never hang around in a tavern.", and a negotiation would ensue as to what the realistic place for the player to begin play would actually be according to the player's desires and the approved background of the character. Part of the problem I had with your early reference to 'You all start in a tavern' is that in fact, for 'purist-for-system simulationism' there is this prior to game negotiation over the simulations starting state where the DM and the players negotiate over how the characters will fit into the world at the moment that the game clock begins running and the whole elaborate mechanical clockwork begins to turn. Most true 'purist-for-system simulationist' DMs probably actually do feel that 'You all start in a tavern' is the first step to railroading the players. What if the players would rather start in a rowboat? What about on a pilgrimage? The DM is expected to kick off the game in a game state that represents what realistically would have happened before, where 'before' is something that the player has some input over and 'after' is also something that the player has some input over.
This is not consistent with traditional AD&D play, because the transition from TPK to prisoners wasn't mediated by the rules, except in the very loose sense that the 4e rules didn't mandate that the PCs had to be dead simply because they lost the combat. (For me, this is an attractive feature of 4e - it removes the need for the GM to fudge in the interests of the story.)
Rather, it gives the GM explicit permission to fudge in the interests of story in this particular case. In 1e, the GM has explicit permission to fudge anything he wanted, but strongly encouraged a 'tough love' approach that encouraged the 1e notion of 'skillful play'. In your game, it's clear that you are more interested in maintaining story continuity than in the 'skillful play' that the designers of 1e felt was a crucial part of your game.
Nor had I made any notes about the goblins intending to take prisoners, although there was already a dungeon on the map of their lair. There was no railroading or illusionism. The whole thing was worked out by negotiating with the players, and driven by obviously metagame considerations.
Agreed. However, this says nothing against any of my prior arguments because, since you've moved away from 'the heart and soul of illusionism', you can't actually use this as a counter example to my definition of illusionism whether 'hard' or 'soft'. You are, for the purposes of this discussion, off on a tangent. I have no problem admitting that an open agreement like this isn't illusionism, while still maintaining my prior arguments unchanged.
I'm not 100% sure myself - although Forge theory dictates that each functional game has only one ultimate goal of play, I'm still not fully persuaded of that.
I'm strongly persuaded that that is where Forge theory goes most wrong. Not only IMO does a functional game never have one ultimate goal of play (unless there is only one player, and then only 'perhaps), but since under Forge theory any game which lacks one ultimate goal of play is perforce 'incoherent', it's only a short jump from the assertion that there must be one goal of play to 'Everyone else is having badwrongfun', 'Every prior designer produced incoherent games', and 'You are all doing it wrong'.
Or maybe my game is mildly dysfunctional but we cope.
No, your game is functional because you balance the different desires your players have in a way that your players are comfortable with.
In any event, I'm pretty certain that a game can be played which gives the GM a certain sort of authority over sceneframing but which nevertheless involves neither illusionism nor railroading.
I never asserted otherwise.
All it requires is letting go of an attachment to hardcore simulationist mechanics - some things can happen without needing to conceive of them as mediated via the action resolution rules.
Whoa there. You haven't established that. You've only established that there is
an alternative to illusionism. You haven't established that this is the only alternative, nor have you established that you hardcore simulationism is incompatible with all forms of illusionism, nor have you established that a game can't be both 'hardcore simulationist' at one point in the game and 'illusionist' at another point. See my prior example of 'narrow-broad-narrow', where the default assumption is "During the hook and the conclusion, I will accept a certain amount of illusionism provided that I'm allowed to make the journey between the two points in a hard core simulation where I have a gaurantee of full agency and fair arbitration with the GM in a referee stance."
