Well, let me refocus the question by asking what you’d call the Kindle version of The Intercept. inklestudios.com/archives/1074
But it’s essentially the “same game” either way; the fact that the Kindle version has a huge automatically generated tree is irrelevant to the play experience and even mostly invisible to the author, who was able to design the game without considering that it would one day turn into an exploded tree.
The Intercept for Kindle blows up the whole idea of the stateless/stateful distinction, IMO; it requires thinking about the distinction in a different way.
The idea of “delayed branching” (“delayed-effect branching?”) captures the played experience that both versions of The Intercept share: after making different choices, you’ll read the same text at first, but you’ll experience very different results later on.
Maybe we could call “state-tracking” CYOA (or whatever) with a memory? The game remembers some of what you did before and takes it into account later.
And the Kindle version of The Intercept would be CYOA that fakes having a memory. Or just plain CYOA with a memory. I don’t think that the behind-the-scenes difference between the Kindle and non-Kindle versions of The Intercept is too important here – whether the game carries its memory around in variables or in the page numbers of otherwise identical pages doesn’t affect things for the player.
Mmm. I’m intuitively kind of suspicious of this - it seems as though this kind of game still has state, it’s just emulating it in a weird way. You could, in theory, emulate a fully playable version of Anchorhead in a technically-stateless CYOA, assuming you had room for an exponentially vast number of nodes. The existence of a game like that wouldn’t mean that things like the parser or an object-oriented world model were illusory distinctions.
I’m not sure we’re understanding each other. I don’t claim that Intercept doesn’t have state; indeed, I claim that all CYOA has state, which is why “stateless vs. stateful” doesn’t make sense.
I think most people recognise a qualitative difference between the old paper CYOA books and these modern variable filled stories. In theory Anchorhead could not only be presented as a state-less CYOA, but as a paperback! So could the Intercept. But I at least think that when a node needs to be present multiple times, which are on the surface identical, because of subtle choices in the past, then it’s different from the kind of story where each paragraph only needs to be given one node.
I think the technical meaning of stateless is perfectly fine, and considering how many people have been using it for such a long time, I think it’s fine outside that technical context too. I’m not sure what the fuss is about.
The interesting thing about The Intercept in this context, I think, is that no-one playing it sees any difference between the web- and Kindle- versions. That’s interesting because, even when reading on a Kindle which cannot tracks variables, people still play the game with a sense of state in their heads. And they do that, not because the game carries any label, but because the game design makes it pretty clear that some kind of state must be being tracked; otherwise it would make no sense.
I think Choice Of games do much the same - at least, the stat-based ones I’ve tried (Kung Fu, Dragon). Because I see these values changing, I implicitly understand they’ll affect things later on. As soon as one Choice is greyed out for me, I know that has to be because of tracked variables, right?
The Walking Dead game, IMO, achieved the same thing, just by use of the UI pop-up that said “Clementine will remember that” after the occasional decision. (The Walking Dead is totally delayed branching, too, but didn’t use that term, but simply said “branching”. I think delayed branching as a problem in that, no-one really wants anything to be delayed, do they?)
I don’t intend any difference – “CYOA with a memory” was my proposed new term for what we’ve been calling “state-tracking CYOA.” Just a failure of writing on my part in explaining that. Where I use “state” to mean “anything the game keeps track of over and above what page you’re on”; as I said before, I don’t think it’s that confusing to use “state-tracking” to mean “tracking of states beyond the absolutely minimum state a CYOA must have to be a CYOA.”
But anyway the thought was that the page number controls the stuff that’s immediately present to the player. Anything over and above that is in memory. So maybe that would convey what you want more clearly, without raising any trouble about what a “state” technically is.
I don’t think this is necessarily the case; not every state-tracking story can be flattened into a simple hypertext. Consider as a trivial example a story with two outcomes for potential single choice, decided between using only a random number. Many “RPG-elements” stories use such a device, for combat, skill checks, or other randomized obstacles; D&D gamebooks, for example.
But even though non-deterministic works can’t be turned into functionally equivalent (unweighted) node trees, I’d say they still have more in common with deterministic state-tracking stories than with simple hypertexts.
Here’s another one: “static pages vs. dynamic pages.”
In a paper gamebook, the pages don’t change. Dynamic pages can change, either randomly, or based on your earlier decisions.
The Intercept for Kindle has static pages; The Intercept for web has dynamic pages.
(When it’s software, they’re not literally “pages;” I suppose techies would prefer to call them “nodes.” But “pages” is at least as good a metaphor as “branching,” and I think it’s a bit easier for outsiders to understand if we call them “pages.” Nodes are not literally knots, either.)
I nonetheless claim that we should resist the temptation to refer to the entire product as “static vs. dynamic.” The Intercept is a dynamic gamebook regardless of whether you play it on the web or on Kindle, but on Kindle, its pages are static.