So anyone who’s played my game, or read my review thread, will probably be unsurprised that I’m taking advantage of my one last opportunity to write way too many words for IF Comp 2020. I know some folks tend to dislike post-mortems, but as a first-time author I at least found it really helpful to read a bunch as I was getting my feet wet, so hopefully this will be of interest and/or of use to at least a few people.
Before getting into the meat of the thing, I should note that I’m assuming anyone reading this has already finished the game, or isn’t planning on doing so. Here there be spoilers, in other words
The basic concept
“Where do your ideas come from?” is usually a vexed question to answer, but atypically this time it’s pretty easy: a pun on “Eleusinian Mysteries” popped into my head late last year when I was playing the Ancient Greek Assassin’s Creed game, and along with the title came a writing style (stolen wholesale from P.G. Wodehouse) and a genre (puzzley shaggy-dog story).
At the risk of over-elaborating the one bit of this I’ve managed to communicate in a reasonably compact manner, I’d been casting about for an idea because I was going through one of my intermittent phases of wanting to enter IF Comp. I’ve been into IF since in the early aughts, though my engagement has waxed and waned. Last year I stumbled across Jimmy Maher’s blog and devoured it, in the process going back to many of the old Infocom classics I’d missed out the first time round and rekindling my interest in (primarily parser, though I’ve been enjoying getting up to speed on the choice-based movement) IF.
The pun occurred to me since I’ve had a specific interest in the Eleusinian Mysteries ever since I read Larry Gonick’s incredibly lovely Cartoon History of the Universe – it was one of my favorite books when I was a kid, great at presenting real honest-to-god history in an engaging, funny way, and the idea that there were these ancient rites so secret that we still don’t really know went on with them has stuck in my head ever since.
(Entertainingly, as best we can tell, the Mysteries were celebrated towards the end of September or the beginning of October, so they’re really a perfect fit for IF Comp!)
The last bit of the alchemy was Wodehouse – funnily enough, I actually haven’t read that much of him, but my wife is a super-fan and we’ve read probably half a dozen of the novels aloud to each other. His way of writing about a world of hapless clubmen seemed a perfect fit for the story of an accident-prone toff navigating his way through a rather silly series of challenges to qualify for initiation into a secret society.
After I’d come up with the idea and started implementing it, I realized that I’d struck on a surprisingly large number of common ideas with past games that have done well in the Comp (Mike Spivey noted this as well in his review). Beyond generally being a comedy parser game, Alias “The Magpie” does the Wodehouse thing, albeit riffing more on Blandings than Bertie Wooster; there’s a party with lots of drinking, as in Zozzled (I even used the word “sozzled” in the poem I wrote for the game!); and there’s a prominently-featured pig, as in Lost Pig. I hadn’t actually played any of these at the time, but I spent a chunk of the early part of this year catching up on some of the games I’ve missed over the last decade or so of not being as active in IF, and it was unnerving to see so many coincidences – I do worry about how the game might come across those who are aware of those other games, though hopefully my decision to target the rather narrow slice of an audience Venn diagram involve those who are big into P.G. Wodehouse, intrigued by Ancient Greek cults, and enjoy old-school puzzle games undercuts any accusations of mercenary motives.
Design and implementation
Once I had the idea, the initial design went pretty quickly. I spent an evening or two late last year doing research – nothing too fancy, basically re-reading some old books I had on hand and doing some basic Google and Wikipedia-diving. Just as I knew about the Mysteries from Gonick’s Cartoon History, his book had also gotten me interested in the mutilation of the herms, so I knew from the off that I wanted to include that as the climactic bit of action – and I of course had to include Alcibiades, one of the most intriguing characters of Ancient history (lightly disguised as Alky, of course. Puffy and Machon are made up).
I also had my structure reasonably well set from the off – it made sense to use a series of puzzles involving the initiation rituals as the centerpiece of the game, but since I’d also be teaching myself Inform as I went, I wanted to start with something simpler. So a cave-set Act I, where the player would need to engage in very simple-to-code, Baby’s-first-IF-game puzzle solving to gather the materials for a linear but more complex-to-code and socially-focused Act II, made sense – and then I put in a black box for an Act III that would somehow involve vandalizing a bunch of herms and wrap things up.
I’ve run a bunch of tabletop role playing games in my day, so I did my design work pretty similarly to how I’d write and prep an adventure for an RPG. I primarily worked from a text file with brainstormed ideas at the top, and then at the beginning of implementing each Act, I sat down for an evening or two’s worth of work to sketch out a basic outline for the section. This involved writing some intro and outro “boxed text” to get the writing style and scene transitions figured out, and coming up with puzzles and solutions at a fairly high level. After I was satisfied with what I had, I did a last pass over the text file to write up what locations and objects I’d need to implement, and from that jumped into implementation (I’ve attached this “design doc” to the bottom of this post in case it’s of interest).
I found Inform 7 reasonably easy to learn and work with – I’m not a programmer, though I did take a few classes back in my undergrad days, and I think that combination of being familiar with coding, but not wedded to any particular language, might be something of a sweet spot for learning I7, since I understand it does things very differently from most modern languages. The documentation is idiosyncratic, but the examples were super helpful, as were the very friendly folks on this board. This is not to imply that I am currently a master of Inform – to the contrary, my code, even the most recently-written stuff, is embarrassingly cobbled-together and violates all sorts of best practices. Like, at one point I ran into a bug due to the order two Every Turn rules were firing. I’d just read the bit in the documentation on how the order is determined by the number of conditions in each rule, and took the admonition that you should under no circumstances use this quirk to re-order things as an invitation instead, and stacked a bunch of wholly unnecessary conditionals into one of the rules to shift the timing (this later caused problems, because of course it did. Programming!)
Anyway, once I had the spine of the thing, a couple nights a week I’d dive into the IDE, and work on turning my notes into code, sometimes focusing more on coding or more on writing as the spirit moved me. After making some good initial progress and finishing up a first cut of Act I, I wound up putting the game on the shelf for about six months due to a death in my family. When I came back to it in late June, I had both a lot of time due to COVID (I don’t have kids), and a strong desire to bury myself in a project set in a world where nothing bad ever happens.
Things went pretty smoothly from there – I got through Act II as I’d initially designed it, and then my wife helped me brainstorm ideas for Act III based on her favorite bits of Wodehouse. Things ballooned, though, and Act III proper became all about learning to use Scenes on the coding side, and creating a more open puzzle environment with lots and lots of alternate solutions on the design side; but then I added an on-rails sequence so I could learn how to do a semi-interactive cut scene; and a more ambitious finale that opened up alternate endings and involved coding a complex puzzle using several identical objects (which was very hard, would not do this again!) Again, I designed each of these in a day or two, then spent probably two or three weeks coding each up, except for the cut-scene bit which was only a couple of nights’ work.
I had a working first cut of the whole game by the beginning of August, and after playing it myself a bunch of times and inflicting it on my wife (an IF newbie), I was able to recruit a quartet of beta testers on these forums. They were really lovely, dedicated folks with lots of great feedback (which I tracked using a Trello board). After making upgrades – most of which were about making Act III easier and better-clued, since that’s where folks were getting bogged down, though there were also some great suggestions about alternate solutions and better signposting things like the comedy puzzle – the indefatigable Mathbrush volunteered as a tester of the near-final version in early September. There was definitely some feedback I should have implemented but didn’t, due to a mixture of creeping burnout and fear that any larger changes would lead to unforeseen consequences due to my ramshackle coding practices (probably I should have done a full rewrite in late July, once I had a better idea of what I was doing, but that hadn’t seemed like very much fun). So I spent a reasonable chunk of the last week reworking the opening poem into a slightly less-awful approximation of dactylic hexameter (the meter of Homer!) rather than doing anything useful.
Once it was out in the wild, two kind players flagged bugs I’d missed , so I made two small mid-Comp updates to fix them, and I’m planning on doing a post-Comp release that smooths out additional niggles.
Since this was a teach-myself-Inform game, I went into it thinking that the puzzles would not be the strength, and instead I’d be relying on writing style and jokes as the primary thing keeping players engaged. So beyond the general themes for each Act, I didn’t have a grand theory of puzzle design, and instead just went through my brainstormed list of stuff related to the Mysteries and cross-referenced it with topics in Writing with Inform and/or my memory of puzzles in other, better games. That’s why there’s a hidden-exit puzzle, a timing puzzle, a try-a-nonstandard-verb puzzle, a conversation puzzle… all the classics, really. I was surprised by how well the Mysteries lent themselves to a silly adventure, though – like, the comedy-show bit in Act II is inspired by a real practice of joke-telling, and provided a wonderful opportunity for me to learn how clothing systems worked once I decided to (ahem) dress it up a bit.
Anyway, most of the puzzles are I think unexceptional and, hopefully, unexceptionable. I saw one reviewer describe them as Infocom-like, which I suppose is about right, since by design they’re fairly traditional inventory manipulation challenges. There are definitely a few solutions that rely on cartoon logic or reading the author’s mind – like propping your eye open with a toothpick to stay awake through the vigil, or getting Puffy off the roof by tormenting him with your awful singing and lyre-playing – but while these amused me enough to implement them, I tried to be real about assessing where things were getting too wacky and making sure there were alternative, more reasonable solutions on offer (one lesson learned is that I shouldn’t have showcased these in the walkthrough or hints, since that can make players think they were expected to make these leaps of logic!)
To my mind, the one real stinker in the bunch is the krater puzzle, where you need to tie a rope to a vase to pull it out of a pile – the physical constraint here doesn’t make a ton of sense and is hard to communicate, and it isn’t funny or interesting enough to justify the friction. In fact, the genesis for the puzzle was pretty much that I wanted to learn how to make a rope puzzle, and then crowbarred it in. The other dishonorable mention probably goes to the wheel-repair puzzle, especially the requirement to use honey as glue – my testers did OK with it so I left it in despite feeling a bit borderline on it, but again this was probably the wrong call.
On the flip side, there are two and a half puzzles I’m pretty pleased by. The two are among the first ones I thought up. First, there’s the pennyroyal puzzle, which I like because the answer is mostly hidden in plain sight, but the player might not notice it at first since they’re prompted to focus on finding the right herb by color, not smell or taste. The second one is the “Mighty Potnia” bit, because I think it makes complete diegetic sense while being very silly (“insufficiently plosive, old man” is probably my favorite joke in the piece. Second-favorite is probably Quite Sufficient Qua Cave So Far As You’re Concerned).
The “half” is the finale puzzle, which I think provides an appropriately farcical climax. It’s a bit more R-rated than would have been ideal – thus the “unmentionable” euphemism, though one of my testers had the great suggestion to allow “thingy” as a shorter alternative – but also I think walks the line of being ridiculous while arising fairly naturally from the plot. The reveals of additional unmentionables I think are also among the better gags in the game. The sequence was a lot of work to code properly, since there are many potential places unmentionables can wind up (in both inked and uninked varietals). Keeping track of them all and providing the right output during the magistrate’s summing-up was challenging, but I thankfully haven’t noticed any bugs in the transcripts I’ve looked at so far!
I’m only giving myself half-credit, though, because I think I got the pacing wrong. The idea is that you should fail at least once, get a hopefully-satisfying ending, and only then, if you’re so inclined, go back to fiddle around and try to optimize the puzzle, which is intentionally very hard to solve completely – it wouldn’t be much of optimization puzzle otherwise, and besides, it wouldn’t have really aligned with history for Alky and his friends to completely escape suspicion!
But paradoxically, I think by trying to be more lenient, I made the puzzle feel harder, and made it drag on too long instead. There’s a time limit, after which the magistrate breaks in, but I made the limit very generous, and prevented actions like examining, taking inventory, and so on from advancing the timer. That means that instead of the scene feeling like a chaotic dash with steadily rising tension, with ideas left unexplored when the first playthrough wraps up, I think most players run out of stuff to try while still waiting for the magistrate to come in, which sucked energy out of the sequence and made trying again feel less fun. I’d noticed this in testing, so I added an escape-hatch where you can remove the barricade and jump straight to the ending without needing to hammer the Z key, but I think I misdiagnosed the problem. I’ll probably tighten up the time limit in the post-Comp release.
Reception and lessons learned
I haven’t fully gone through all the reviews or transcripts yet, and this is already way, way too long, but I’ll just close by saying that it’s been really gratifying to see people play the game! Per the reviews spreadsheet, there are only a handful of games with more reviews, all of which are shorter and/or by much better-known authors, suggesting the blurb and/or my wife’s lovely cover did a good job of convincing people to give it a whirl (I was very anxious about these two things!)
I think I was expecting it to be a bit more polarizing than it’s so far turned out to be, though that might change once all the votes are in (as of this writing, all 8 IFDB raters have it at 4 stars – and I just noticed that means that on the “first effort” list, I’m two spots ahead of Zork. Since I went to Caltech, I will take this completely unjustified moment to gloat at my triumph at those overrated jerks from MIT). I was prepared to get a few ravingly positive reviews, and a larger number of people for whom it just didn’t work; instead the range has seemed to go from “really good, with some niggles,” to “not bad, with some significant flaws”. I’m expecting to wind up in the top third or so based on this, which was my personal goal, so here’s hoping!
People justly dinged the game for being overly verbose and a shaggy-dog story, which, [gestures to mountain of words immediately above], fair cop – though beyond a logorrhetic style, I think part of what motivates this criticism is that important objects aren’t set off by a line break from the often-fairly-long room descriptions. I did this on purpose to make sure people didn’t ignore the environmental descriptions, which are shortened on second visits and do have exits broken out as a concession to playability. But given that it is a reasonably long, large-ish game, probably this was the wrong call.
Most reviewers thought the game was at least a bit tough, though a few found it quite easy. There were definitely places where the puzzles could have been improved: as mentioned, the finale timing was off and the krater puzzle should have been cut, beyond several places there was a need for a few additional synonyms or opportunities for better clueing (there was at least one reviewer who got very frustrated with the puzzles trying syntax that really should have worked, except I’d messed up the coding, which of course feels awful). But overall I’m pretty satisfied with where things landed on the difficulty front, since that’s more or less what I was aiming for (though I think this is a target difficulty that works better for a stand-alone game than a Comp one, given the time and attention constraints).
On the flip side, it seems like the humor worked for most folks, and despite a few hiccups I think most reviewers thought it was reasonably well-tested and implemented (which makes me laugh knowing the hellscape of rusted bear-traps, raging dumpster fires, and cursed homunculi that lurks on the other side of the “compile” button).
All in all writing the game, and participating in the Comp, have been a lovely, lovely experience that’s meant a lot to me in a year that’s been quite hard. I’m grateful to all the folks who helped me make The Eleusinian Miseries, and all the folks who’ve taken some of their time to play it and let me know what they thought! And of course I’m already kicking around ideas for next year’s game, now that IF Comp’s got its teeth into me. Till next year!
Eleusinian Miseries design doc.txt (17.1 KB)