Is the short story the most natural fit for IF?

All the comps are doing it. Vast majority of the authors are doing it. Development resources over development time seems to demand it. IF is highly experimental and the more experimental corners of other media also evolved to focus on short forms. And as a player, I tend to approach IF more spontaneously when I have better information about the size of my commitment, as I do when there is a score (I am way pro-scoring for this reason) or when a short game is the expectation.

By adopting the short form, has the IF form found its natural fit? Or is this a phase of some sort? I’m thinking, natural fit.


I don’t think it’s a natural fit for IF, so much as a natural fit for the kind of people making IF at the moment. If a commercial company decided to make an IF, I imagine that a relatively small team of writers, working 9-5, would probably find a novel-sized game a better fit. But with IF currently being made primarily by individuals in their spare time, unpaid, I think a short story is the only thing they can expect to do well, unless they have the stamina to work on a game for years.

That said, I think that authors can make bigger games if they’re prepared to ignore certain received wisdoms and accepted ways of doing things. Procedural generation and randomisation are ways for small games to offer more content than a single author would normally be able to provide. Limited command sets and/or restricted implementation are ways to create sprawling stories/worlds while limiting combinatorial explosion. User participation is another way to extend the life of games dramatically.

Basically, I think short stories are a good fit for modern IF, but if you just peak over the fence at what other amateur game devs are doing, there are a lot of other forms that look like they might fit as well.

My prejudice is for short games, whether or not they’re interactive fiction. Actually I think IF might be better suited for longer games than other forms of games, but I’ll get to that.

I say it’s a prejudice because, never having bought a game between “Yars’ Revenge” and the Humble Indie Bundle 2, I’ve never actually had a chance to play a long game except nethack and maybe Cave Story, and I suspect Cave Story wouldn’t count as a long game if I were any good at it. (Oh, and I think Legerdemain probably is long, but I haven’t yet made it to the third area.) Absorbing my attention for that whole time while continuously generating new content just seems like an incredibly difficult challenge. And from what I understand of the length of a typical AAA game, they expect me to spend more time on it than I did watching the first season of Heroes. I didn’t even want to spend that much time watching the first season of Heroes. Really good TV shows seem to come in under 20 hour-long episodes per season, and they seem like they can deliver more content than a game can.

To elaborate on that last point: A game can communicate something through its procedures in a way TV can’t; if only by making you perform the main characters’ actions rather than watching them be performed. Can it keep communicating new things through its procedures for twenty hours? I’m skeptical. (Anna Anthropy has a couple of pithy rants against the eighty-hour game here and here.)

So I don’t really want any game, or at least any game that falls short of complete brilliance, to try to soak up that much of a time. (And keep in mind that in the time it takes me to play a novel-length game I could probably read several novels; progression through a game is usually going to be slower than progression through a novel.) Procedural generation is one way to keep things interesting for a long time, though I’m not sure IF is very well suited for it. The other thing that might make long-form IF suitable is just that IF is a medium that’s really well suited for delivering tons of story, and a long (and good!) story can sustain interest. Even so, I suspect that I’d enjoy an extended IF story more in multiple episodes than I would as a single game. Partly, perhaps, because a single game would probably either have to close off a lot of the content to begin with, effectively gating you through multiple episodes, or present you with an overwhelming amount of stuff at the beginning.

I like some long games, and some short games, and I think IF is appropriate to some of each.

One of the big problems I have with epic-length IF (er, epic for IF - I haven’t run across any Moby Dick-sized works) is that it’s easy for me to lose my place, especially if I’ve restarted the game at some point, especially especially if there’s no mechanism for reacquiring information. (There are relatively easy workarounds to this, of course, but most IF don’t use them.) So not only does a long IF work have to continue to draw me back, it has to be so compelling that I want to spend a whole three-day weekend on it, which . . . yeah. That’s gotta be pretty compelling. I do prefer longer games in general, but it is usually not handled in a way that makes them easy for me to finish.

Serials might be a better fit - something like the magazine serials novels used to be published in, or comic book serials. A little story, with a big hook on the end for the next issue, and regularly scheduled releases could be quite compelling, and reuse the same base “engine” each time. (I happen to hate that mode of delivery, but I think it would be a good choice - the engine could be an extension.)

But part of what’s a good fit is what audiences will play and writers will write. I’m surprised there aren’t more gamey-games - I can think of a number of ways that strategy IF could be hugely fun. So could sim games, or RPG’s. I don’t think they’re completely natural fits, but part of that is that I’m not convinced I could get many people to play one. But the idea of plotless IF (or story emergent IF) is still vastly unexplored.

Interesting – like what? Indie games in general are going the same way as IF it has seemed to me. Shorter games, simpler graphics, one idea expressed as well as possible. Mind you with graphics play can be more naturally extended without adding any story; still, they seem much shorter and less ambitiously epic now than they were.

Man… I hear ya. Know how I keep up on the AAA scene? I go to a hardcore friend’s house and just ask him to play and show me game after game. The sad thing about modern mainstream gaming is that despite the 80-hr commitment, once you have watched someone play for a half hour you have seen everything of worth. I can’t think of any counterexamples though I haven’t tried Heavy Rain yet. (Alan Wake though is the same story – let’s make them do the same damn thing for two weeks, but in a cooler atmosphere.) Red Dead Redemption had a rather interesting ending, narratively speaking: I watched it on YouTube. 87

Thanks for the cool-looking links. I’ve been enjoying her Twitter feed. I’m fundamentally uninterested in procedural content generation. What I want is an author to fake me out with smoke and mirrors so that the environment feels somewhat randomised or ‘experienced’ but without actually being random, because random is boring. A lot of game authors don’t realise that they are fundamentally engaged in a deception and that is never going to change no matter how good their worlds get: if they did realise that it is fundamentally a deceptive act, their narratives would get better, because narrative is also fundamentally a deceptive act, and those who get this are those who tend to master it. Maybe they’d be more willing to pull fake-outs and tricks, maybe shake n bake them a little with their other techniques. But no, they seem committed to producing virtual worlds as an end in itself, and they consider a lot of narrative techniques ‘cheating’, and it’s just so dreary to me. Frickin’ fool the hell out of me! Fake it! Trick me! Make me think it’s all random when it’s really all so… not. Anyway, so few seem to get this, that I agree modern studio games are like a wasteland, but it’s one I feel obligated to know about because evolution dictates that if you put enough monkeys in a room, one of them is eventually going to type something interesting. XD

I really like your ‘multiple episodes’ idea. Something I’ve also been thinking about, but I don’t want to announce a first game as ‘volume 1’ of anything. I’ve seen too many part ones with no part twos – it leaves a bad aftertaste of promises unfulfilled. Better to promise nothing but what is already in their hands.


I think that’s pretty perceptive and to the point on why short forms are ascendant.

I also agree with you about engines, there are too many engines already as it is. Not that someone with an idea for one shouldn’t do it, but for the purpose of supporting a serialised story? Just publish the episodes in the accepted engines. Don’t use new tools unless you need new tools because they do new things that are integral to what you’re trying to achieve. Or just do engines the way graphic designers do – they are internal. You don’t really talk about them much. (They don’t even ask people to install them separately, the game is just the game. I can understand why IF splits the streams it makes sense for IF, but really there is too much focus on engines and not enough focus on techniques. I wonder what the response would be like if some well-respected prolific author were to take apprentices?)


Like the ones I said. :stuck_out_tongue:

I don’t see that indie games are “moving in that direction” - if anything there are probably more epic/HD/multiple-idea indie games than there used to be. But, as with IF, single author games will always be more likely to succeed if they’re smaller in scale.

That seems… backwards to me. What exactly is appealing about randomness in and of itself that you would want to be tricked into experiencing it? Surely the trick with randomness/procedural generation is to make it seem like it isn’t random but is actually the result of cause and effect or authorial control (the AI director in Left 4 Dead is one example of an effort to achieve this).

Randomness is a side effect of a particular way of trying to create limitless worlds - but it’s an entirely undesirable one. I’m sure we’d all prefer it if the limitless environment could instead be crafted instantaneously by a master architect.

I wouldn’t, because an architect isn’t a storyteller. If all I’m getting is a sandbox to wander around in – well, that will keep me amused for a while, but really I prefer to do that in the real world. To some extent we are talking at cross-purposes. You are talking about what is required to make a good sim (which is why you’re talking about faking intentionality on the part of the characters where computers find it difficult to generate) whereas I am talking about what is required to make a good narrative (which is why I’m talking about disguising intentionality on the part of the author and making a certain set of pre-designed conclusions seem to have been generated spontaneously from the environment and your choices).

I think of your issue as the classic graphical game designer’s issue – they are all about that – whereas the issue I thought you meant and that I tend to focus on is largely ignored or given short shrift by graphical game designers. So, very different things, yeah.


P.S. ‘Like what’ I meant which games. Most of the indie stuff I have seen getting buzz lately (Passage, Sleep is Death) are very much aimed at short play sessions. Are you talking about stuff like Minecraft, etc.? i.e. stuff that bears little relation to storytelling? That’s what I was wondering.

Okay, I understand your point logically now, but it’s not something I remember ever thinking about a story. “That was really well done - I could believe it was random.” I think people like to see cause and effect in stories - not because it’s realistic, but because that’s part of the appeal of fiction - a chance to tackle our sense of cause and motive. In an interactive story, in which you’d expect the player to be supplying the causes, if not the motives as well, that probably goes double.

Sleep is Death is the kind of thing I meant. The play sessions may be short, but the act of user participation means that the game as a whole is larger than any short story. Essentially Sleep is Death is closer to a collection of short stories.

The Choice of… games also feel larger than most IF games (even though they’re a lot quicker to play), which they’re able to do by playing at a higher level of interaction (making broad choices rather than interacting with individual objects). I like the idea of an IF game where you have high-level verbs for significant actions in the place of the usual “take/open/put” form.

And, yeah, Minecraft was another one. Maybe that kind of thing isn’t related to storytelling (although I think the IF equivalent could well be a game that focuses on characters and motives the same way Minecraft focuses on blocks and physics) - but so what?

When it comes to graphical games, I’m often concerned with the place that storytelling takes, but within the IF community I think it’s the gameplay that bothers me more. Too often it feels like authors come up with an awesome story, but when it comes to what the player actually does in it, they just slap in some puzzles. An IF game that instead featured something like the malleable world of Minecraft (you have a story-based goal, but how you achieve it in the sandbox is up to you) would be an interesting development - something that could make a small game feel a lot larger. It might not be the “best fit” but it would at least be “another way”. I suspect that with enough experimentation there could be a lot of “other ways”.

To sidetrack this for a bit into a discussion of randomness in games, I disagree here. I’m tempted to say, “I just read Alex Kierkegaard making the same argument, therefore it’s wrong.” (Warning: link contains homophobia, caps lock, generalized rampant immaturity.) But that would be silly.

Seriously, one appeal of randomized gameplay is unpredictability – not just that the game will be new for the player each time, but that situations can arise that the designer didn’t anticipate, and that allow for kinds of gameplay that weren’t anticipated either. There was an interesting post from the designers of Steambirds: Survival here about why they decided to dump preset levels for randomness. The whole thing is worth reading, but here are some bits:

There’s also a nice diagram of how randomized design means that the playtesting experiments cover a wider variety of experiences. Incidentally this is not about creating a limitless world; the game always takes place in the same blank expanse of sky (bounded by those rassenfrassen antiaircraft guns).

I think part of what they’re getting at here is the importance of the randomized feel. In this kind of game you don’t necessarily want to feel that the designer is setting up an experience for you – “Oh hey, I see what that powerup is there!” You want to be able to discover things you can do for yourself without feeling that the designer got there first. It might be possible to create that kind of experience with a large enough set of preset levels, but that might be the sort of thing that Paul was referring to when he was hoping that the author would fake him out with something that feels random even if it isn’t. Or it might be the opposite, I’m not sure. :wink:

Now, in my previous post I tossed aside the comment that I didn’t think IF was necessarily suited for procedural generation. What I meant was this: In graphical media the recombination of a small set of elements can stay interesting. Look at a game like Walk or Die (which even people who are bad at twitch gaming should be able to get pretty far in). There aren’t that many different items repeating, but the variety is endless, because having a tree of this height this distance from another tree isn’t quite the same as having a tree of that height that distance from another tree. In prose recombining the same number of elements won’t stay as interesting, I don’t think; Hunter, In Darkness does an excellent job with its procedurally generated bits, but at some point there isn’t that much difference between “Delicate brown crystals grow everywhere” and “delicate dark grey crystals grow everywhere.” The random prose can’t quite sustain attention on its own, or at least not that much more attention than the same amount of prose presented non-randomly.

All that said I’d really like to see more different modes of IF play – more simulation and strategy kinds of game, and I’d love to be proved wrong about procedural generation too. I think I’d love the game with a goal that takes place in a big sandbox that allows you multiple ways through, all the better if I lose sight of the goal.

the problem is the dicothomy between the story and the crosswords puzzle, between literature and game. If the story is good, we want it long. If puzzles are good, we want more of it. But long story most times mean more text thrown at you at once while you have no saying in all that is shown and told. And more game aspect most times mean more rooms to explore and more literal and annoying attention to little irrelevant details, like “take key out of wallet, unlock door with key, open”. I’m glad some IF automate all that and some games actually have a lot of story interspeced with lots of quality interaction. But it’s quite rare still…

I’m inclined to agree. Gameplay-wise, you can get a lot out of playing the same chapter over & over. A change in setting and/or antagonists doesn’t do as much for gamplay as it does for story progression. Witness any Street Fighter game. Gameplay’s phenomenal, but the various settings are nothing more than wallpaper. Narratively, though, an individual fight isn’t as interesting as one’s overall progression toward the Big Bad.

But I-F does have that content generation problem, and lack of gameplay problem.

Interesting thing about randomized text: the first time you read it, it isn’t randomized (even though it is). You can’t tell a difference until you see the same language again with small tweaks in the same context. I think Emily Short’s When In Rome 2 begins with a randomly-created antagonist, but you wouldn’t know it at all unless you replayed the game two or three times at least.

What I think Laroquad is after is how to disguise railroading better.

So can an interesting set of tools with various ways of accomplishing a goal(s), which is what good videogames do. And Spider & Web. Once the player understand their tools, they can make and execute plans. And THAT is what mastery of a gameplay feels like.

Ico and Shadow of the Colossus were about 10 - 12 hours long each, and I felt both were too short. It may be that up to 20 hours hits some sort of sweet spot.

I don’t like “episodes” though. An episode usually needs to close a narrative arc, and that give me the pre-packaged feeling of TV episodes to me. I don’t mind occasional lulls in activity for contrast, but a lull every hour on the hour (or insert your favorite train schedule of choice) is too artificial, and that seeps into the rest of the work. Movies, novels, videogames, I-F, and songs are each as long as it needs to be. Only TV has that hard time limit. “Episodes” are stupid. (nyaah.)

“Movies, novels, videogames, I-F, and songs”

Movies are the only ones where they’re designed to be taken in full without a break. If every movie was 3 hours long, bladders and people’s senses would rebel, so they sit at around 90 mins to 2 hrs, which seems to be tolerated well by earthlings.

Novels, games, songs, IF, don’t have to be taken in one sitting (music concerts excluded) and can be as long or short as the author wants them to be to convey whatever the author wants to convey. I don’t think any of these forms have any kind of natural fit. There are trends, and the current IF vogue seems to be for the relatively short. I figure this grows out of the major event in IF culture being IFComp, which mostly mitigates against long games.

I don’t even know what long means in IF all the time. In the 80s you could have games sitting around for 18 months+ that you were unable to complete (EG Wizard and the Princess!). The game’s unhelpful parser, toughness and the absence of an internet providing you with instant hints and walkthroughs help explain this.

How much content was in that game in real terms versus a new longish (only longish) game like One Eye Closed? Probably less. But a movie that was 78 minutes long in 1960 is 78 minutes long today. Wizard and the Princess’s length was a year and a half in 1981 and… how much today? Your actions, patience or lack of it and tolerance or lack of it will change the perception of the length drastically.

Well, I agree that having the episodes be precisely the same length is artificial (and also, in IF, completely impossible), but novels do tend to have chapters, and the chapters are usually roughly the same length. Maybe all I’m looking for is lulls, though, places where you feel like it’s OK to save and pick it up another day.

Anyhow I agree with the rest of what you say (except that I haven’t played Ico and Shadow of the Colossus), and Wade (severedhand) makes good points too about variability in length. Actually I bet that Ico and Shadow would take me a lot more time to play than they did you.

Not how I would put it, particularly since it is my belief that there is no such thing a non-‘railroaded’ narrative, so I would not use that word, because it takes what is a universal truth about narrative (that it must be led in a particular direction by a human designer) and makes it sound like a pejorative. But it can’t be a pejorative, because there aren’t any other options.

Your observation of what I am after is essentially correct. But what I think of as a ‘railroaded’ story is one in which the bloom of choices is too narrow or too poorly disguised. When you get right down to it, though, every story is either a railroad (in your sense, Ron) or not really much of a story at all. I know there are lots who disagree with this, in fact who have pinned entire lines of software development on disagreeing with this, but their disagreements all still seem theoretical, from my POV, pinning hopes on technologies that have never delivered, narratively. Experience with actual games bears this out: Narrative = railroad. They’re the same thing. So it’s a matter of giving the track enough splits and clothing it in enough ‘choice’ clothing so that it doesn’t feel like what it is. There is no shame in this. It follows the same tradition that gave us ‘characters’ in the first place. Characters are fake – they aren’t real people; they are marks on the page, a trick of the light – an intentionality that isn’t. We allow ourselves to believe they’re real, even though we know they are a trick – and we respect the author when he or she makes it very easy for us to willingly fool ourselves into believing they’re real. We don’t dis that author for deceiving us. If an author’s characters are obviously fake, we don’t criticise that because it’s a deception, we criticise it because it’s a bad deception (a very key distinction for IF that tends to get overlooked). This covenant between author and audience was not given to us a priori from the heavens: it developed by convention. (And BTW not everybody agreed at first that these deceptive novels were a good thing.)

So, I see IF as developing a similar covenant: ‘I know I can’t have true freedom of choice but if you make me believe it anyway, for a while, you’re Aces.’ Narrative is just the art of tricking people with their consent, and disguising the ‘railroad’ is no different. We should approach this task with pride and craftsmanship, and not make like it’s a terrible compromise, because the alternative storytelling techniques we would have to compare ‘railroading’ with in order to reach that opinion, don’t actually exist anywhere that I have seen. 8)

Sorry for the single-sentence-targeted response, but I’ll be back for more later.


Hrm. What about Alice and Kev? Boatmurdered? I’m not presenting these as gotchas – I expect you have answers for them – but I’d like to hear what you have to say about them, to clarify your concepts.

There’s also the “narrative in gameplay” aspect, as in the story of Jane the Tourist (from nethack) that I posted a little while back – Dwarf Fortress is probably intermediate between this and Alice and Kev, because it’s partly about struggling to do well in the game, whereas Alice and Kev is about setting up obstacles to create an interesting narrative. (I haven’t actually played Dwarf Fortress or the Sims.) That – a narrative that’s also a narrative of the player’s struggle – seems more orthogonal to the kind of narrative we’re talking about, though.

If we’re defining narrative as a created story, a told story, then I’d say things like Boatmurdered are written by the users, incorporating elements given them, sort of like improv. And the same is true for Alice and Kev - there’s no actual story there, except in the telling. I could take the same video or events in either game and spin a nearly completely different story, at least insofar as character, motivation, theme, etc. (Would the Great Gatsby still be the Great Gatsby if the events were the same but the characters and dialogue were totally different?) Sometimes these stories are given specific frameworks by voluntary play mode - like the One Dwarf Challenge, or the Apocalypse Challenge.

But Dwarf Fortress and the Sims can provide inspiration and a rich toolbox, and far more meat for storytelling than, say, Nethack, where the events are far less malleable and open to interpretation. I mean, you’re in a dungeon, going down. There are stories to be told about that, but the range of events is constricted to a pretty limited set. Gameplay, graphics, random events, variety, worldbuilding - all these can expand your toolset dramatically. A smoothly enough written game with a big enough toolset may appear to write its own story, because it’s so easy to fill in the gaps, even as it allows for alternate explanations from people who deliberately approach from a different angle.

I actually really like this form of gameplay, but it is different than the game designer deciding that he’s going to tell the story of the Chosen One who will Save the World. (Although I confess that I enjoy playing those games and watching destruction spread in the “savior”'s wake, and imagining what the people around him/her must be thinking.) It’s more play-assisted narrative, like pirate-themed Legos or Barbies. There may or may not be an intended framework or simple narrative (Barbie loves Ken and clothes and lives in her dream house), but there’s more often a player-chosen narrative (Barbie and Snake-Eyes have an ongoing feud over who will control Candyland, interrupted only by sieges from the Lego pirates).

There’s nothing to be ashamed of in giving people tools for telling stories, and that often overlaps with stories we’re telling as authors. I don’t agree that the best stories are necessarily the ones that we are told. Some of them are the ones we tell ourselves, and some of them are the ones that appear to be told to us (". . . and then the zombie elephants and skeletal carp came FOR REVENGE . . .") even if that story isn’t hardcoded or intended or imagined by the player.

It’s a randomly selected antagonist, out of a fairly small pool; the idea being that each antagonist brings with it certain challenges to diagnose and control, and having them randomized lets the player play several times and experience the evolution from not knowing much about these critters to being quite good with them. I was trying for something with a clear narrative arc but a fair amount of gameplay depth; the critter AI is intended to give the player a slightly different playing challenge every time around, and encourage solutions that aren’t specifically things that I came up with myself. (How well this worked, I don’t really know – I’ve gotten very little feedback on those games and have the impression that they weren’t that successful on the whole.)

But I do cautiously agree with the sensation that there is some disappointing gameplay in story-rich games; or, at least, gameplay that doesn’t live up to the story. I found “One Eye Open” oddly gripping, but most of the actual challenges were about collecting lots of key objects and diary objects – an idea old as the hills and sometimes rather repetitive. (There were some exceptions. The laundry machine still makes me go “ew” at random moments.) But quite a few games have puzzles that feel like they arise for purely plot-mechanical reasons: we need to block the player at points X, Y, and Z, and require a visit to P before unlocking Q. And those are good reasons to have puzzles, but not all that’s needed to determine what the puzzle should be.

Sorry for the delay — things got a little hairy there and my sleep schedule is now all messed up.

I agree with this, but I think you’re talking about hindsight, which is different, IMO (and can be specifically written to be different) from the way the events are experienced as you are experiencing them. I’d prefer the player to feel as if they had discovered something that didn’t necessarily have to be discovered, and then in hindsight, to feel that the cause-and-effect of it all made sense in retrospect and is satisfying (but not that it is inevitably programmed to happen to every player of the game).

I agree with you on the first point – gameplay is now often given short shift – but disagree with your conclusion (to suggest a swing toward IF gameplay in a Minecraft-inspired vein). Probably because I don’t see the slippage away from gameplay in the same context. I see it as failure to remain seated on the head of the pin. 87

I’ve said it before, but I don’t really beleve in the dichotomy of ‘puzzle vs. story’, and I’m certainly never going to come down on one side or another. What we have here, IMO, is a Third Thing. A thing which we can currently most conveniently understand by referencing the subsystems within it that we recognise, historically (which we typically called ‘puzzle’ and ‘story’). But neither puzzles nor stories are unitary things, either. Each is made up in turn of several subsystems (for puzzles: interface elements, data objects, if/thens; for stories: pretend people, perched-on-the-shoulder invisible narrators, interior monologues, elisions of time, etc). If we didn’t previously have a concept of a ‘story’, we might be able to have an argument about whether these newfangled novelists should focus on writing great pretend people, or on making interesting juxtapositions of time – but this is a false dichotomy, because there is no reason you can’t do both; it’s all part of writing a story. Our concept of ‘story’ just happens to include both those techniques, by convention alone. We don’t think of them as polar opposites even though elided time can interfere a great deal with establishing a contininuity of character, if you don’t do it carefully and thus fall off the head of that pin. The same goes for IF – the only reason we divide it up in terms of ‘puzzle’ and ‘story’ is these are the pre-existing forms that we can identify within IF as subsystems, and because the combination of skills involved is not common in our specialised, stratified society, and so writers tend to come up short on one side or the other. But it didn’t have to be this way. It’s an accident of history. There is nothing intrinsically unified about ‘puzzleness’ or about ‘storytelling’ that should give these forms a valid claim to being irreducible pieces, whereas IF somehow is reducible to into its component parts, so that we can have debates about which part to weigh more heavily, which is the ‘important’ one.

There is no important one. IF itself is an irreducible form, because the elements it combines create a Third Thing not present in any of its subsystems, and that’s why as a medium, it just refuses to stay down and die. If interactive fiction were really just all about story, then it wouldn’t really be necessary, because we have that pretty much covered. To the extent that IF is going to continue to be exciting as a medium, we have to steadfastly refuse to pick a winner between puzzle and story, as I see it. Even better if we simply stop looking at it in those terms, because it might be blinding us to precisely the territory that is exclusive to IF: all the magic that can happen in the middle. So, I’m for the Third Thing. That we don’t even have the perfect words for it yet is incredibly exciting, because it’s just that new. New and irreducible things are not that common in history — particularly in the history of art.

So I’d have to disagree that taking cues from Minecraft is the answer, although I am in no way opposed to experiments in that vein. I would just perceive them as attempts to learn about one aspect of IF by focusing on it, kind of like when you write a film that all takes place in one room; and that’s a good thing for a particular work of art. But the medium itself is located in whatever it is that you get when you combine the various subelements of what we think of as puzzle and story, but which is impossible to describe by reference exclusively to either. And that puts Minecraft-like building block games off on the margin of the spectrum of the most fertile IF fields to explore, IMO. The magic is in the middle – that’s where all the brand-new-to-the-world crossover effects will be found.

Yeah, Sleep is Death is an interesting example in this context, and it’s not alone it seems, so I’ll skip over to something Matt mentioned.

I don’t have answers for them, actually, because I haven’t played them. I take it that Boatmurdered is some sort of an add-on for Dwarf Fortress, which is a game I read about briefly but didn’t try. Anyway, thanks for the links!

Reading responses to you, it sounds like there are a lot of parallels between the sorts of games you’re talking about and Sleep is Death, which is another where the players create the story using the game as just a set of building blocks, and I do see this as a different kind of thing. Not that it’s not IF, but it’s more like… the difference between Pong and Breakout. Pong was fun because they recruited a human to be your opponent for you, making the sorts of decisions that the computer was finally able to make in Breakout, but 6 or 7 years before affordable computer components were able to make them. So the reason games like Sleep is Death work so well is that, like Pong, they sidestep the entire issue and go for a different sort of experience. But that’s not same as solving the issue, and my main interest is still in solving the issue, and in the work of those who are attempting to do so. I have nothing against collaborative storyteling games though, and they are like IF in the same way that Pong is like Breakout – which is pretty damn close but the difference is actually pretty crucial to the game designer and the engine. You have to actually change what you are trying to achieve to go from writing Breakout to writing Pong, instead. I hope this makes sense, I feel like this paragraph has now vanished into a rabbit hole in the early '70s. XD

A great point that overlooked in my first read. Randomised text requires replay to detect as randomised – so, it also requires replay to have any effect at all different from non-randomised text. Both the postivie and negative effects of randomised text require replay to detect (though, not necessarily to function). This is interesting I think in the context of the discussion that followed about the length of a work.

Thanks for mentioning Wiz & the Princess. I remember it fondly, although it may have had the suckiest ending, ever. XD But we’re talking about completion-of-puzzles time, here, obviously, which is so hard to quantify. Maybe that’s why shorter for IF, works better now. It means there’s time to spare – that’s kind of important, in light of the ability of puzzles to expand.

Anyway, that wasn’t the quote I had in mind related to replays, and now I can’t find it – but someone upthread included replays in the estimation of the ‘length of an IF work’ and I wanted to disagree with that somewhat. It really depends on how the replays are meant to be played. It takes a very special game to get me to replay – I am not a big replayer. Once I get what is clearly meant to be ‘a non-failed conclusion’, I will generally judge the game on that ending. I won’t replay it in case there might be a more satisfying ending. I won’t replay it to find out what character X might say about character Z if I talk to them, this time, before I meet character Y. I would only do that if the author of the game specifically seeded me with that desire with some parting clue, which is a legitimate technique that I wouldn’t mind seeing used more often. (Yeah, sorry peeps – you don’t get replays for free, you have to actually work narratively toward them just like any other paragraph, or I’ll take the first opportunity to ‘close the book’.) If those techniques are used, then I could agree with counting replays as part of the length, but if they are not specifically narratively seeded, then no, replaying most IFs I have seen is the same really as rereading a novel – you see different things in it the second time, but it’s basically the same thing so most of the narrative tension has run out of it, and it’s just less interesting to me, unless I’ve become such an uberfan of that author that I feel like pouring over every word, which doesn’t happen much even in the elder arts.


Sorry, I didn’t make myself clear. These are both recountings of specific stories that take place within sandboxy games – Alice and Kev in The Sims, Boatmurdered in Dwarf Fortress. So they’re not things you play, they’re things you read about other folks playing. I haven’t played either of the games myself either, let alone participating in those particular sessions; the links just let you read about what happened. (Did the story happen in the sessions themselves or was it created in the editing and presentation on those websites? I suspect the former, but I don’t know.)

So I’d say you’re right that the player is creating the story, but it’s more that the single player is creating the story with the aid of the Sims/Dwarf Fortress, and the story develops in ways they don’t necessarily expect, rather than that another player is GMing the story as I think happens in Sleep Is Death. The player of Alice and Kev was specifically trying something unusual in order to create a story; I don’t think that was true to the same extent with the players of Boatmurdered.

[edited a couple times]