Limitations of ChoiceScript as a medium

[Havenstone]

Well, with Wraith_Magus on a forum hiatus again, my hopes of avoiding mega-textwall by going through his points one or two at a time are dwindling. I’m giving him a lot to respond to as and when he gets back. So it goes.

Another limitation Wraith wrote about was the difficulty of representing meaningful conversations in ChoiceScript:

@Wraith_Magus

You can’t even really have meaningful conversations in a text-heavy game like this, since the actual interactions are reduced to multiple-choice where the player simply picks the choice that will get them the result they want, not the one they would actually feel the need to say in the heat of the moment. (It is, for example, never a good idea to say anything angry in response to being insulted in-game, because that just makes your relationship bars drop, and higher relationship bars are always good. There’s never any negative consequence for being a doormat, and players can always take a breath and keep emotional distance to avoid making rash decisions in the heat of the moment.)

It seems to me this is mixing up a couple of things – the current habits of ChoiceScript writers, and the limitations of computers at simulating meaningful conversation. There’s no reason inherent to CS that “it’s never a good idea to say anything angry.” Relationship bars aren’t part of the engine. It’s easy in theory to write negative consequences for being a doormat or good consequences for an angry response. It’s rarely done… but that’s grist for a thread on “cliches of CS writing/coding” rather than one on limitations of the medium.

On the other hand, the fact that conversations can’t really range beyond the options that writer has envisioned and programmed is basically a consequence of computers’ inability to semantically understand the world. Wraith recognizes this throughout the thread, but he thinks there’d be a better way out if instead of presenting multiple choices we could use the “meaningful abstraction” of a graphical puzzle:

@Wraith_Magus

I’d like to see something more like Puzzle Pirates, an MMO where a diverse array of interactions are handled with abstract puzzle games, making meaningful interactions from things that aren’t usually capable of being rendered in a game meaningfully.

I’ve got nothing against puzzle games; I write as someone who quite enjoyed the hacking pipe minigame in Bioshock. (And huh, the guy responsible for it also wrote one of my favorite COGs!) But applied to the problem of simulating a meaningful conversation, graphic minigames seem to me every bit as much of a hammer seeking a nail as anything ChoiceScript writers try with stats, if not even more so.

It’s not as if we haven’t seen it tried. Anyone who played Oblivion will remember the use of a graphic puzzle minigame to simulate conversation, and, well, yeah. The abstraction was pretty far from being meaningful. To each their own, but I think a passage of ChoiceScript written with a bit less stat-reliance would be much better than a minigame at producing the illusion of a truly meaningful conversation.

Of course, that would require more detailed coding, where the author anticipates and writes out a different range of human reactions… and that brings me to one last, slightly odd point of Wraith’s. He argues that ChoiceScript is inefficient because “trying to recreate everything over and over again when you can just reuse existing code is simply a waste of time and effort better spent advancing the project in new directions.”

And I guess this is where game designers of a certain mindset simply part ways with authors. What has been “recreated” between Choice of Robots, Choice of the Deathless, and Choice of the Star Captain isn’t a bunch of coding inefficiencies: it’s the story. It’s the whole point. It’s worth the time spent coding it because – to come back to the point I made about Minecraft up top – a strong narrative and good characters are pleasures that only human authors can (so far) create. They don’t emerge from any open-world spatial simulation that I’ve ever encountered. True, an open world can include various narrative mini-hooks, and I can enhance my experience as I explore by filling in the gaps with my own imagination. But it’s a different kind of pleasure to the pleasure of a well-realized novel.

I understand that Wraith is interested in a different kind of project, using “spacial simulation in new and different ways [rather] than just writing a new narrative.” But that’s no reason to bash new narratives.

@Wraith_Magus

To use the words of Tarn “Toady One” Adams, he said he “was not interested in making a crappy fantasy story, [he] wants to make a crappy fantasy story generator.”

Personally, I’m interested in making a good fantasy novel, and one which is all the better for allowing you to explore alternate paths through it – to get to know the characters from different angles, to see different bits of the world, to allow the consequences of actions to play out differently. If I end up writing a crappy one, it’s because of my weaknesses as an author, not because ChoiceScript locks us into writing crap.

I do find the “story generator” project interesting; even a crappy fantasy story would take on a certain extra pleasure if I felt it had emerged almost entirely out of my choices rather than an author’s will. But it would be a different pleasure to exploring a well-written narrative like Stanley Parable or the best COGs.