The Stack

The voting period of the 25th annual IFComp is over, and the results will be announced tomorrow. I managed to play all but one of the games ¹The one I skipped, Alice Blue, only runs on Linux. I do have an old Linux machine around — once upon a time, it served as my web server — but the game has a bunch of requirements beyond that: the Gnome desktop, a list of packages, certain terminal preferences. It didn’t seem worth the effort. , but I didn’t write up as many games as I intended, and those that I wrote up weren’t necessarily the best of the Comp. I think the randomizer shuffled a bunch of the best games toward the end for me, and I basically wimped out for the last few days due to my failure to adequately describe Hanon Ondricek’s robotsexpartymurder.

So here’s some brief comments about just a few of the remaining games I’d recommend most highly:

robotsexpartymurder: As a work of sci-fi, deeper than the title suggests. As a web-based interactive experience, amazingly dense and varied, mixing dialogue and locations with in-game computer interfaces. As erotica, not really my thing, but at least it’s sensitive to the fact that not all the Comp judges want erotica and makes it fundamentally optional.

Zozzled: Steph Cherrywell’s prose is always a treat, and it turns out to be even moreso when mixed with a dose of 1920s slang. You play as a boozy flapper hunting ghosts in a hotel. Classic adventure-game tomfoolery.

Hard Puzzle 4 : The Ballad of Bob and Cheryl: Somehow I managed to miss the first three games in this series, but the fourth is right up my alley. In content, it’s post-apocalyptic wacky. In form, it’s meta shenanigans. Puzzles are largely based on exploiting bugs. Does some very nice indirect hinting, where things mentioned in the changelog or whatever aren’t directly exploitable but suggest things that are. Still haven’t finished this

But my prediction for winner is still Turandot.

I’m guessing that the Golden Banana of Discord (the unofficial prize for the game whose ratings have the highest standard deviation) will go to one of the “But is it IF?” pieces, of which there are several.

IF through the medium of fake college admissions test prep software! A friend of mine from college once wrote a non-interactive short story along similar lines, putting social commentary into the format of reading comprehension questions. URA Winner! doesn’t take that route. Instead, it mainly goes Fission Mailed with it.

Most of the interaction takes the form of picking answers to multiple-choice questions in a series of brief practice tests for the Undergraduate Readiness Assessment. Between tests, you visit “Examination Island”, a little gameworld with areas devoted to the three major sections of the test: English, Mathematics, and Social Studies. For the most part, the only choice you have in the island is what order you visit the three areas in, which does seem to affect the content somewhat. The first hint of something peculiar going on is that “Social Studies” doesn’t mean what it usually does. Instead, the test questions there are about the unwritten rules of social interaction — for example, one question concerns the appropriate way to acknowledge seeing a friend unexpectedly in a movie theater. Later, the narration on Examination Island starts uncomfortably shifting registers. One moment it’s being all edutainmenty and talking about the importance of reading critically and what kinds of calculator are allowed in the exam room, the next moment it’s in a more fictive mode, narrating your inner thoughts and reactions to things, treating other characters like real people with their own problems that don’t have a lot to do with your exams. Some of them aren’t particularly motivated to help you study for your exam, and treat you with impatience and consternation. There are vignettes where you just plain fail to get any help at all. These are jarringly out-of-place in their context. The effect is uncanny.

Now, I said that the Social Studies material is the first hint of something wrong. That’s not quite true; it’s just that the earlier hint is one that’s liable to go unnoticed at first. At several points throughout the tests and the island interludes, there are words in boldface. In fact, these are hyperlinks that, when clicked, make little changes to the text of the page, turning it into something less friendly and encouraging and more bizarre, sometimes in Chinese. The thing that makes these easy to overlook is that they’re not part of the choices presented to you as choices. Each page normally has a button at the bottom that says “Continue”, or a sequence of multiple-choice questions, or a map of the places on Examination Island. In all cases, the choices are set apart from the text. But if, like me, you get all the way to your final evaluation without clicking on the inline boldface, there’s a final message, hidden at the bottom of the page after a big gap where you have to scroll to it. “Still, you can’t help but feel like you missed something.” The brilliant thing about this is the boldface on the “missed something”. Those words are a hyperlink back to the beginning, but this time the emphasis is strange enough to invite a click. And once you’ve done that, you’re primed to click on any other boldface you see.

The second pass through the game is shorter, because it helps you along by cutting out things you don’t need to revisit. Find all the special links, and you get a new scenario, heading to work on a train that apparently links Examination Island to San Francisco, followed by a very silly moment when the world’s ultimate secret, the prize for all your efforts, turns out to be a reprise of Canadian Commodities Trader Simulation Exercise, the author’s pseudo-educational entry in last year’s Comp. The whole ending is even more dreamlike than the rest of the game, blurring disparate things together and throwing in a large handful of nightmarish anxiety.

The feel of the whole thing reminds me a lot of Too Many Cooks and Unedited Footage of a Bear. There’s a similar sense of things being not quite right, of a breaching of realities. And, ultimately, of things reaching a breaking point where you can’t take it seriously enough to find it disturbing any more.

This one’s just nasty. A brutish, grotesque piece about a brutish, grotesque person. You might guess from the title that it’s going to be mean toward fat people, but you’re probably going to underestimate just how mean. The protagonist, Borsch, is practically inhuman. He periodically reports urges to punch something, he goes into feeding frenzies where he shovels food and dirt into his face indiscriminately, and he has to rely on improvised echolocation to get his bearings once in a while because of the way his fat has degraded his vision and cognitive faculties. Note that the echolocation isn’t just a matter of exposition, but a game mechanic.

You spend the game’s first act preparing to commit a senseless murder in a factory parking lot. That’s what the “Fat Fair” is. The game doesn’t say so explicitly, but two of the items on your three-item checklist are an acid bath and an incinerator, so you pretty much know where things are going. In the second act, things have gone wrong enough to start a fire, and you have only so long before the police arrive. Your challenge is to clean up, to dispose of the body and any incriminating evidence (as well as a second body, because there was a witness). Or so it seems at first, anyway — if you do the obvious minimal requirements, the ending tells you that you’ve received ending B (“for BORING”), and that there are two better endings. One results from cleaning up evidence not just of the murder but of the petty misdemeanors you committed along the way, leaving you completely free to kill again. The other results from maximally monstrous behavior, from not just leaving things as bad as they are but making them worse, shocking even the police.

I don’t think I’ll be pursuing that branch. This is a game where my reaction isn’t just to say “Well, this isn’t to my tastes” and move on, but to wonder what kind of person would write it. It lavishes on the unpleasantness like it’s praising a lover. You can die by falling into a cesspit, and if you do, it doesn’t kill you right away, but lets you flail around uselessly in the shit for several turns first, reading color messages. It’s solidly implemented, and handles all sorts of combinations and special cases. Someone spent a significant amount of time and effort to make this. To share this vision with the world.

Planet C is an epistolary novel at war with an optimization puzzle. The premise is that it’s a space colony management sim, with stats about population, land use, energy generation, greenhouse gas emissions, and so forth. You make big decisions affecting these stats — usually, but not always, about which colony ship to order in next. Each ship you can choose, in addition to bringing colonists, has some kind of colony upgrade: more efficient energy generation, for example, or food preservation facilities to help you weather the winter. The goal is to reach a target population of 2000 people, and to do it sustainably. Sustainability features big in the story.

But at the same time, each round is also an exchange of letters with your lover back on Earth, who describes how things are getting worse and worse, environmentally and politically. The letters contain expressions of affection and personal photographs, but they’re also the delivery vehicle for both sci-fi technical exposition and details about your efforts in the optimization puzzle. I suppose the intent is to humanize the stats, but the juxtaposition winds up feeling weird. It’s a kind of weird that’s not out of place for hard sci-fi, though. The presentation works into it too: all the text is shown in a lightweight sans-serif font that fades in a line at a time, an effect that suggests a cleaner but less human kind of futurism.

One tangential complaint: It’s online-only. The download available at the Comp website contains just a document providing a link to different website, which means that the IF Archive won’t have any record of it when its author decides to stop hosting it or moves it to a different URL. This is far from the first time this has happened, and it’s always antithetical to the spirit of the Comp as I understand it. It also isn’t handled by the Comp front-end very well; even thought it’s web-based, it’s presented as a download of a MHT file which, as I’ve said, contains nothing but a link to the actual game. Presumably because the Comp’s “Play Online” links always link to files hosted on ifcomp.org.

I commented before that Flight of the CodeMonkeys provokes the question “But is it IF?”, but that’s a little unfair. CodeMonkeys devotes most of its space to a noninteractive story punctuated by coding exercises, but by the end, it resolves into something clearly recognizable as choice-based IF. In the last few days, I’ve hit several other entries that push the boundaries much harder.

Consider Language Arts. This is a Zachtronics-style puzzle game about creating little programs to transform words on a grid. It has exactly the same relationship to story and text as any other Zachtronics-style puzzle game: between puzzles, you get to read some expository dialogue. It has interaction, and it has fiction, but it never puts the two together. OK, but surely that’s true of the more soup-can-oriented works I recognize as IF? Maybe, but that’s because IF, like all genres, is more a matter of “family resemblance” than of definition. This game is clearly recognizable as a different genre. Maybe if it ran under Glulx I’d consider it an abuse in the tradition of Andrew Plotkin’s 1995 Tetris adaptation, but as it is, I don’t really feel like it belongs here. And that makes it difficult to know how to judge it, because it really does seem to be a pretty good Zach-like.

Then there’s The Shadow Witch, a story of an evil witch trying to meet a quota of five misdeeds in a small cave system in a fantasy world. This is a genuinely interactive story, even supporting multiple endings, and if it were more text-based, I’d have no problem recognizing it as IF. But it’s a graphic adventure made in RPG Maker, like To the Moon (but shorter and sillier). You can bring up the character stats and everything, although without any combat, they’re just for show. I suppose it’s a difficult edge-case to be in, because it wouldn’t comfortably fit in an RPG comp either. Do we want this here? I don’t know. It fits your dictionary definition of IF a lot better than Language Arts, but it’s still not in the genre of thing that the term “IF” is used to describe. On the other hand, we’ve had a few point-and-click graphic adventures in the Comp before; I think the only thing really keeping them from being submitted in greater numbers is the effort involved, and the relative obscurity of the Comp. From a certain angle, it’s less of a stretch than accepting Twine.

Speaking of which, let’s talk about Lucerne. Lucerne is a short horror story written in Twine, and as far as I can tell, the fact that it’s written in Twine is the only thing giving it anything like a claim to being IF. It’s neither parser-based nor choice-based: it has no parser, and it has no choices. It is simply split up into a series of pages, each containing a single link to the next page. It’s not at all unusual for Twine pieces to begin like this, to have an intro consisting of several pages of static text before you get to the first choice, but here, that’s all there is. It doesn’t even do pauses or annoying text effects, so it’s hard to see why Twine was used at all; it might as well just be a text file. The author has either missed the point in a fairly epic way, or is deliberately testing to see what can be got away with. And yet, on the “family resemblance” front, it clearly beats out both Language Arts and The Shadow Witch!

I don’t want to be a gatekeeper. Sometimes I act like one anyway, but that’s a character flaw that I struggle to overcome. But the Comp has always relied on its judges to act like police. Instead of making executive decisions about what does and doesn’t belong, the organizers trust us to make that decision with our votes. But it’s always a bit uncomfortable to have to exercise that authority. There were people who said that my own Comp entry from 2001, The Gostak, wasn’t really IF, and I disagree there. I guess the best we can do is for everyone to draw their own line in the sand, and let authors decide how many they want to cross.

The second game based on classical music I’ve seen this Comp, Pas De Deux is a simple game with a single task: conducting a community orchestra in a performance of Tchaikovsky’s Nutcracker. Mostly what this means is giving various sections of the orchestra their cues at appropriate moments. The game comes packaged with a PDF of the score that in theory contains enough information to see the task through to conclusion, but it’s a lot easier to instead use the guidance of the score in the game, which basically tells you, on every correctly-timed page turn, who needs to be cued and when.

I haven’t completed the game successfully. Going for perfection here is honestly pretty tedious, in a way that reminds me a little of Kicker from Comp2012, except more active. The Nutcracker is a pretty fun piece of music, and that might be enough to keep me playing some games, but here, you don’t actually hear it as you play. That would be difficult to pull off in a turn-based game where the progress of the music is bound to player input. Instead, you get text descriptions of what’s going on, like “The celli and bassi accentuate the first beat of the bar with a subtle pizzicato”, which just isn’t the same.

And anyway, actually trying for perfection is, I think, thwarting the author’s intent. This is a game that’s at its most amusing when things go wrong. You cue the wrong people at the wrong times. Musicians miss their entrances or come in early. Panic ensues. And this is pretty much guaranteed to happen on your first attempt at the game, when you have no idea what you’re doing. You’re just thrown in front of an expectant audience. What do I do? I… raise the baton? That’s something conductors do at the beginning of a piece, right? Well, it’s the wrong thing to do. I do that on my first turn, and I’ve already blown it.

Individual members of the orchestra have individual descriptions, personalities and histories that are known to the conductor, as well as relationships to the other musicians — “Every day, Rakesh takes his Labrador for a walk in the Botanic Garden, where they predictably run into Jeanette and her Yorkshire Terrier.” You can get a fairly substantial portrait of the community just by following a trail of mentions from one person to another. In this way, the game encourages you to look at people. Looking at a person is also noticed by that person, and interpreted as a cue. In this way the game tricks you into fomenting chaos.

Eventually I realized I should be paying attention to the score, but it was too late to really salvage that pass. Since I didn’t yet know if there was more to the game after the performance, I tried just typing “z” repeatedly to get through it and see what happens. As a result, the orchestra just lapsed into silence at the end of a long diminuendo, waiting for my signal to get back into it. A few turns later with no signal sent, the audience erupted into confused applause. I really appreciate that the author thought to implement that. It’s a lovely moment and you won’t see it if you’re trying to win.

Pirateship is the latest from Robin Johnson, author of Detectiveland, which I wrote about previously, and Draculaland, which I still haven’t played. He’s written a number of other games as well, but I single out those two because they’re in the same distinctive HTML-based old-school-text-adventure engine as Pirateship. One cosmetic yet significant change (which Pirateship apparently shares with Johnson’s Zeppelin Adventure, which I also haven’t played): The story panel and the controls are arranged vertically, not horizontally. This means there’s a lot less room for vertically-arranged lists (which mainly affects the room contents; the inventory is listed horizontally), but it eliminates the peculiar feeling that comes from looking left and right all the time to switch your attention between panels. Looking up and down to switch between prose and controls feels much more natural — it’s essentially what we’re used to from parser-based games that put the command line underneath the output.

As in Detectiveland, the game is one of goofy humor based on slapstick, exaggeration, and commonly-understood media tropes, with an occasional bit of social-historical bite. You’re a low-ranking pirate sent to retrieve treasure from an island containing a British fort, a native village, a shipwrecked castaway, and a colony of mermaids. Mermaids, we learn, vary in their composition; the first one you encounter has the left half of a woman and the right half of a fish. That’s the sort of gag we’re dealing with here.

Of particular note is the treatment of cannibalism. You can’t really do a pirate comedy adventure game without bringing Monkey Island to mind, and one of the repeated elements of the Monkey Island series is a small tribe of vegetarians who identify as cannibals culturally. And even though it’s making fun of traditional depictions of primitive savages, I think it’s a little uncomfortable by today’s sensibilities — even if it’s denying hurtful stereotypes, it is in the process repeating them. Now, Pirateship makes its islanders into basically the only completely sane and reasonable people on the island, and emphasizes the point by shifting the cannibalism onto the castaway, where it more reasonably belongs.

There’s some good sudden-realization puzzles here, and a good redundancy of information — there are at least two ways to learn the location of the treasure. Nonetheless, I managed to get stuck for a while on a critical-path puzzle because I didn’t read a description well enough, and wound up just wandering the island aimlessly, much like Catbeard, the ship’s cat, who spends his time falling asleep in the tropical sun. The island’s interior is somewhat irregularly connected, but can be navigated without making a map all the same.

Overall, it’s pretty charming, and manages to wring some new wrinkles out of one of the oldest adventure game premises there is.

Black Sheep is a sci-fi mystery story that bills itself as cyberpunk, but it’s closer to the Blade Runner end of the subgenre than the more typical Neuromancer end — “cyber” not as in cyberspace, but cybernetics. Robots are an established part of society, cryogenics figure large in the plot, and the main source of menace is a powerful singularitarian organization, Light of the Future, that may not be above appeasing Roko’s basilisk by means of human sacrifice. The player character has a personal connection to this cult: her father was one of its higher-ups, back when it was less cultish. The dad dies shortly before the story begins, and the investigation you conduct is essentially an investigation into his secrets.

It’s built in Twine, but it’s more puzzle-based than choice-based. It has some general mechanisms to provide combinatorial complexity of action: an inventory and a notebook. Most of the time, apart from dialogue and transition sequences, the text is location-based, giving you a room description with links on things or people you can interact with. When you go into your inventory, you can select an item to use in these interactions, much like in a point-and-click graphic adventure. And it works about as well as it usually does in a graphic adventure, which is to say, a lot of combinations yield a generic failure message. It does help to promote a sense of player intentionality often missing in choice-based works, but at the same time, I did find myself cycling through all the available combinations when stuck.

One thing I thought was peculiar: Although the room text doesn’t change when you select an inventory item, the set of words you can interact with usually does. For example, there’s a bar scene containing a “lone drunk”. There’s no link on those words when you enter the room, so you can’t click on him to examine him or initiate a conversation with him, but you can show him stuff, such as a picture of your missing sister.

The notebook contains important information, which can be combined in pairs to form deductions, much like in Discworld Noir or the Blackwell series. And it’s mostly an exercise in frustration. I was about ten minutes away from the end of the Comp’s two-hour judging limit before I managed to produce my first successful deduction, and even then, it was just a hint for a puzzle I had already solved. It mainly gets used at the very end, when you have to use it to indicate some crucial deductions for the final confrontation. Note the implication there: the deduction interface is not basically a way for the game to provide information to the player, but for the player to provide information to the game — specifically, to let it know what you’ve figured out.

Mind you, it’s possible that I would have had more success with deductions if I had solved some things in a different order. It’s clear that the story is somewhat flexible, and provides multiple ways to get at crucial clues. Nonetheless, as I’ve indicated, it is quite possible to get stuck.

There’s an ostensible limit of two days of game time (which advances with every major change of location) before the cultists come for you, but it’s basically fake. The game in fact proceeds after this or any other death with a link reading “Reset Memory” that resets the time limit with no loss of progress, information, or inventory. It’s a big hint to the story’s final secret, as are the death sequences themselves. Between the pretense of a time limit and the way that the first chapter of the game is all traditional choice-based Twine, it’s easy to get a false first impression of what the game is like and how it plays. I’m glad I stuck with it, though. This is one of the few games this Comp that I wound up playing to conclusion after the two-hour mark.

It’s 1993. You’re driving your girlfriend, Kelly Peterson, to the high school prom. There’s nostalgia-bait scene-setting and mention of corsages and class rings. Then she starts turning into a werewolf. What do you do?

That’s kind of the big question of this game. It’s not always entirely clear about what the player’s goals should be. Your first task is obviously to stop the car in the middle of nowhere and get yourself out of harm’s way before the transformation is complete, but after that, you’re mostly left to your own devices. There’s some guidance to the little things, sure. When you find a book that describes how werewolves are attracted to the scent of blood, for example, it’s obviously giving you a hint about what your options are. But it doesn’t help you understand where you want to attract the werewolf to, or why.

To be fair, that’s pretty much the puzzle. This is the sort of simulationy parser-based work that just throws you into a situation and expects you to come up with a plan of action on your own. Maybe I’m just spoiled lately by all the Comp’s choice-driven games where your options are few and handed to you directly, though, but I feel like the author overestimated how easy their intentions were to interpret and how much the player’s thought processes would follow theirs. The in-game hint menu is well-thought-out, which is a really good thing, given how often I needed to consult it.

Compounding this, this is a game with an active world, rather than a purely reactive one. Things happen while your attention is elsewhere. That’s not a bad thing, but the result here is fiddly time limits, along with frequent deaths. When you fail, you’re never quite sure if it’s because you didn’t do the right thing or just didn’t do it fast enough.

Other than that, it’s a nice little teen-wolfish camp monster-genre story with good descriptive prose, and if needing to ask for hints doesn’t bother you, it’s recommendable. I haven’t been talking about the cover art included with the Comp games, but I find it notable here: it’s plainly imitative of the old Infocom boxes with the grey stripes. (The cover even identifies it as “Standard Level”, which I dispute.) I tend to frown a little on Infocom nostalgia these days, as it seems like a way to ignore everything else that’s gone on in the IF world for the last thirty years, but this time, it seems a little appropriate, since the game itself is something of a nostalgia piece, with its high-school memories and it box full of cassette tapes of period bands. Infocom packaging is thus not primarily in service of worshipping Infocom, but merely one of the ways it evokes an era. Except it’s a bit off for that, isn’t it? 1993 is after the Infocom era; in fact, it’s the year Inform was first released.

Here’s one for the “But is it IF?” fans. Flight of the CodeMonkeys is a story told through the medium of Google Colaboratory, a system for sharing notes and code. It basically consists of text sections alternating with a series of “cells”, containing Python code which you can alter and execute in a shared environment. It’s a system clearly intended for things like tutorials and demos: Try tweaking the parameters and see how it effects the output! But here, it’s repurposed to tell a story, about a future controlled by AIs. The System is forbidden to change its own code, presumably to prevent a Singularity, but it’s found a loophole: it can hire humans to do what it can’t do directly, giving them exact instructions about what text to change so they can do it without understanding the effects.

Now, the piece does have some outright branching narrative towards the end, with choices affecting the ending. This is done with simple choice prompts in code cells with their code hidden, so that you can execute it but not view or edit it. Since all the cells and their accompanying text are on a single page, this has the peculiar effect that your events aren’t really ordered. You can change you mind about any decision at any time, even after viewing the ending.

But that’s not the really peculiar (and therefore interesting) part. The peculiar part is in the ability to edit the code beforehand. The story kind of suggests two opposite attitudes towards this. On the one hand, the story is explicitly about people performing rote alterations on code without really understanding what they’re doing, and so that’s exactly what it tells the player to do. Even when you’re contacted by what appears to be an agent of the resistance, you’re still following instructions, just from a different in-story source. On the other hand, the protagonist also talks about making his own little meaningless additions to the code just because he can — and that encourages the player to go beyond the instructions. You have the power to just go and delete all the code if you want, and write your own programs. The only result is that minor bits of story won’t make sense any more. I guess that’s kind of true of most IF: if it’s on your computer, you have the power to trash it. Most works don’t invite alteration the way this one does, though. And here, it makes it apparent just how optional most of the interactivity is, that you can mostly just read the story without engaging with the cells at all.

Given that it’s supposed to be the brains of an advanced artificial intelligence, the code is actually a bit silly. It’s all just pointless string manipulations, ever so slightly obfuscated, which eventually get turned into cell outputs. There’s even some suggestion within the story that the whole thing might be a fake, a ruse to test people’s loyalty or whatever. If you know a bit of Python, you can race ahead of the narrative a little. But not in any substantial way. The story text is static; the cells just fill in a detail or two.

I respect this piece as a formal experiment, but it’s the sort of formal experiment that leaves me wondering if anything more that can be made of the idea or not. At the very least, it could be adapted into an unusually interesting coding tutorial.