Swigian - An Epistolary Postmortem

Intfiction forum, 5/13/2016

"Post by Mathbrush - The real pattern that I noticed is that long games that were bug free got high scores regardless of content. You could write essentially any crappy game in the world that was long and didn’t crash and get high up votes.

Post by Juhana-Without any further data to support this conclusion, I’m afraid it might be a good old “all dogs have a tail, therefore if something has a tail, it is a dog” fallacy. Did you actually find any such games?

To me it seems much more likely that because making a long and bug free game requires quite a lot of time and effort, it’s probable that a lot of time and effort was put into other aspects of the game, too."

Personal Correspondence, 6/19/17

"I thought you’d be interested in an experiment I’m running. I thought about my hypothesis for IFComp games (that any long, bug-free, not-frustrating game will do well). I’ve decided to put it to a more stringent test.

I thought that having detailed room descriptions and object descriptions would be important as part of the ‘not frustrating’ part of my hypothesis, but Detectiveland was really bare boned.

I’ve decided to enter an additional game. I’ve written it up in about 5 or 8 hours. It is 150 moves long, and is minimalistic. Only standard verbs, no item descriptions, basic, generic puzzles with no common theme, minimalist text, a storyline full of plotholes, etc. So the puzzles, story, and writing are all barely there.

But it’s not repetitive, as the puzzles vary from place to place, and I’ll make sure it’s bug free. I’ll enter it under a pseudonym.

Hopefully, it will do poorly, perhaps even as bad as Recorded from 2015. That would show that you need to do Something besides write a long game.

It would be interesting if it placed in the top 10 or 15, though. If my theory is true, then reviewers don’t consciously realize what makes them like games; they talk about story and puzzle and plot, when I think it’s really just length without frustration. If it does well, I think it will cause some real cognitive dissonance as they attempt to explain what they like about the game.

Anyway, I thought you’d be interested in knowing about the experiment. It’s called Swigian, and is based on Beowulf."

Beowulf, c. 1000 AD

One began
in the dark of night, a Dragon, to rage.
In the grave on the hill a hoard it guarded,
in the stone-barrow steep. A strait path reached it,
unknown to mortals…

…its murder-fire
spread hot round the barrow in horror-billows

…there laid within it his lordly heirlooms
and heaped hoard of heavy gold

Now the wrath of the sea-fish rose apace…

Personal Correspondence, 6/19/17

"Your feedback in general is very good, it’s just hard to know what direction I want to go with this. I only want to fix things if they’re in the formula: typos and bugs need to go for sure. But non-linearity isn’t in the formula; for instance, Photopia and All Roads are nothing but a series of ‘leading you by the nose’ puzzles with just a little less direction than mine. The game would be much better if it were non-linear, but I don’t think that’s what I’m testing.

The puzzle design is similar. What do I mean by ‘frustrating?’ Removing the exits and writing them in-game makes the mood and writing better, but is that part of the ‘formula’?

Anyway, you’ve given me a lot to think about. I think my rule will be that I will only fix something bad if I can’t find any top-3 game that does that bad thing. Detectiveland doesn’t have item descriptions, so I’ll skip those; the games I mentioned earlier are nonlinear, so I’ll skip that; but I think literally every top 3 game changes the default responses (even Cactus Blue and The Play are in a specific narrator’s voice) and does a good job with synonyms, so I’ll fix those."

Personal Correspondence, 7/22/2017

“Swigian is an experiment; I want to see how far my IFComp theories can take me. I’ve claimed before that puzzles, story, and writing are not as important as length, bug-free-ness, and mechanics/setting, so I’m entering a pseudonymous game with as little puzzles and story and writing as possible, but everything else as good as I can make it.”

Personal Correspondence, 9/19/17

"Would you be willing to look over a short game I’m entering anonymously? I’ve had a theory for a few years that the comp is primarily judged by length, setting and polish over writing quality, story quality, or puzzle quality.

So I’ve made a game that is long (takes 150 moves to beat), with a mildly interesting ‘Beowulf’ setting, but with minimal writing, no real story, and minimal puzzles. It only takes about 30 minutes to play.

Would you be interested in testing it?"

Personal Correspondence, 9/20/2017

"I’m entering an anonymous game that is purposely ‘stripped down’ with minimal writing/story and no real puzzles. It lasts about 30 minutes; it takes a lot of moves, but each of them are easy and quick.

Would you be willing to try it?

I’m entering a ‘real game’ too, but this is a project I’m interested in for theoretical reasons."

Personal Correspondence, 9/26/17

"Thanks so much! And you’re not late at all. I have this game (which is under a pseudonym) and a real game coming out that I’ve worked on forever.

This game is called Swigian. The blurb is “I don’t like to talk. Let’s build a fire.”

It’s intentionally bad in Some ways, but I want it to be as bug free as possible, with many synonyms and no guess the verb.

Thanks so much!"

I spent about half an hour yesterday convinced that this was Ryan Veeder writing pseudonymously, possibly entering a speed-IF or a “heck with it, I’ll write something in three days” sort of game.

I think you may have accidentally made the game better than you intended, though it’s hard to say in what way. There was something about the simple puzzles, the sense that the game was cleverer than it looked, and the way more and more of the story was revealed as you went along that I found distinctly Veederesque.

In the ending, the response to eating the final snack is genius - one of those single lines that has sudden implications for what’s been happening all along. And that wouldn’t have been possible with long descriptions or proper EXAMINE.

I did mark the game down for what I thought of as bad implementation, rather than just sparse implementation, in parts though - there are places where you can’t go back for no apparent reason, there are grammatical errors from the parser, and there are missing synonyms.

Definitely an interesting experiment, and I’m curious to see how it scores.

I think the worst grammatical error is ‘the yourself slips off the hook’.

I’m interested to see how it scores, to. If my 8 hour game does better than my 150 hour game, it will really change the way I write.

I think your theory about “not frustration” is right.

I’ve observed that people who liked a lot Tuuli and Swigian, in the end, they preferred Swigian because it has no frustration at all. I think we could state that nobody wants to be stuck and frustrated in the XXI century. Of course NOT! Who would like that? Even those situations are considered bad design nowadays.

To be clear, I don’t want to relieve the problems of design and implementation of Tuuli. The game fails at several levels and doesn’t work for everyone.

So yes, I think Tuuli works pretty well in pair with Swigian to confirm your theories :wink:

Yes, I think this was accidentally well written, as it were. And a few reviewers did pick up on the gratuitousness of some of the puzzles (I saw some remarks on the canopic jars). Some people liked the falling rock puzzle, which at least is unusual in its implementation–it wasn’t my favorite, though, in part because that was a place where I really rubbed up against the underimplementation. (Though the specific idea that I got stuck on–because you’re on the east side of the ledge, you should know without having been told that you can go west–reminds me of a part where I got stuck in Losing Your Grip because the game didn’t explicitly tell me that I could go south from the balconies along the east and west walls.)

The one place where I think you may have successfully hoaxed a lot of people is in the story. It seems as though many reviewers are thinking about some sort of hidden depth to the story, or something that would be unlocked by deeper knowledge of Beowulf, that you’re perhaps claiming isn’t there.

Also I worked out the authorship by anagramming your pseudonym but I didn’t connect it to your post about what works well in IFComp. Very clever!

This really makes me feel a whole lot better. I gave it a 5 in my video exactly because it felt long but incomplete. When I looked at other folks’ scores of the game I felt like I was missing some kind of genius to the work that everyone else saw. I was inventing strange narratives to explain the disconnect between the general craftsmanship of the game and the lack of investment in the prose: may be it was a genius young programmer with extreme ADHD, who didn’t have patience for descriptions or someone with a limited command of English, painstakingly translating every phrase he used. I never imagined the whole thing as some kind of social experiment. I guess I did miss a major puzzle in the game.

I gave it 5 too, and have already been in contact with craiglocke re my thoughts. What was there was pretty competent, but for me too sparsely implemented, and far too linear.

Perhaps what hurt the experiment as an experiment was that it was clear the ‘bad’ elements were there by choice. So I wasn’t trying to judge it by ordinary parser-puzzler standards but on its own terms… whatever those were. That, and the fact it had been sort-of-recommended to me by two other people, probably made me much more lenient than I would have been. I ended up giving it a 6, which was a 7 marked down for some (as I thought) unintentional clunkiness.

And some bits seemed so solidly clever and atmospheric. If you’re putting those in even when trying not to, that’s got to say good things about your writing.

All the same, I can’t shake the feeling that I’ve just got that cognitive dissonance Mathbrush mentioned.

(emphasis mine.)

Another person who gave the game a five here; glad to know I was not, in fact, missing something. I agree that the problem might have been making the gimmick too obvious: it moved the focus from the length of the piece to the minimalism, and made it seem like a purposeful concept than bad execution of a normal game. The best analogy I can come up with is if you were trying to prove that the wrapping of a present made someone like the gift inside more, so you give someone an empty but nicely wrapped box. At that point, the lack of gift inside overshadows the wrapping completely.

I also had no idea the “long games score better regardless of other qualities” idea was a thing. I think I am personally more inclined to like short games than long ones; a long game has to really hold my attention to make me want to play the whole thing. With Swigian, I quickly lost interest due to the sparseness. I wasn’t inclined to want to explore the game world to solve puzzles because the game world wasn’t interesting to me. In general though, something I’ve often heard is that it’s harder to write a good short story than a good novel, and I think that applies to IF games as well.

I wouldn’t suggest you change your writing regardless of how Swigian places. Your choice of subject matter muddies the water regarding your experiment, I think.

[spoiler]For one, you’ve got a neat bit of discovery where you realize that you’re Grendel. That’s a cool moment.

For another, you’ve then got outside-the-game text doing the heavy lifting of setting and purpose. For people familiar with Beowulf, they can lean on it to fill in the lacunae your prose left behind. It’s like how fanfic can lean on what you already know about the work being, er, fanfic’ed. If you weren’t riffing on existing text, I expect the results would have been very different.

For a third, the sparseness worked with your chosen character. Of course Grendel will be oblique and gruff. Your chosen text makes your prose a deliberate artistic choice instead of a social experiment.[/spoiler]

Yeah, I agree that a lot of the game came off as deliberate. I’m not familiar with the Beowulf story, but the cover art, the title, the blurb: they all signalled there was a plan here. And “I don’t like talking. Let’s build a fire.” is straight-up good writing! (The original “I don’t like to talk. Let’s build a fire.” is a lot weaker to me)

The descriptions didn’t come off as overly bad or under-implemented IMO because they came from a distinct character POV; I praised Ventilator last year for the same thing, which is that long vivid descriptions aren’t necessary as long as they communicate something
(like about the protagonist, or their goals) and they don’t come off as generic. And it seemed to fit the story: build a fire, something’s after you, keep moving, survive! No time for sight-seeing; focus single-mindedly on the task at hand.

The puzzles it felt did sort of increase in complexity overall as the game went along, from the point A-to-B fire to later on the jars, the rocks, etc, which sort of helped generate some purposefulness. As far as lack-of-frustration goes, I’ll compare it to Absence of Law’s robot section, which similarly felt like it was trying to keep the player always moving: that puzzle seemed to have a lot of moving parts and fun interactions which I didn’t feel like I was getting a grasp on before it shuffled me along, and I found that a bit of a let-down. The puzzles in Swigian generally still took comprehension and intention on the player’s part to get past even if they were basic, so they came off better in that respect even if there’s maybe more to like in the AoL one.

But yeah, in the end, everything remained cryptic, and the stylistic choices and the seemingly increasingly arbitrary puzzles never get justified. There seemed to be an underlying intent behind everything which sustained interest for me for maybe two-thirds of it, but then it sort of levelled out where it felt like a climatic reveal could’ve tied everything together.

I think I ended up giving it a 7 for the miss congeniality awards?

My goal when writing was just to not repeat the same thing too many times, but I still ended up doing it. Though I finished it in 8 hours, it was extremely draining; coming up with new stuff was exhausting, but I kept going by saying ‘it doesn’t matter if it’s horrible, that’s the point’.

The game is split into 15 sections, each of about 10 moves, and each awarding the player a point.

Two of the ‘puzzles’ are literally going down a 3-room hallway, taking one action, and coming right back.

It was originally going to be a game as mundane as possible, where you are in your house and have to pick up and put away tools and use them (screwdriver on a loose screw, etc.). But I switched it to outdoors; it became a woodsy game. But it was so boring that I thought it would be fun for the character to be taken somewhere else, and that reminded me of Beowulf.

The middle part of the game is a dragon’s lair; the things you throw off the cliff are sharp and pointy, so he can’t land. That’s why the ground is on fire outside. The dragon was the final enemy in Beowulf, and Tolkien said it was one of only 2 real dragons in English literature (or something like that). Also, the shark is from the sea-monster in Beowulf.

I didn’t include the setting in ‘purposely bad’ because just about any parser game has an interesting setting. In this comp, 8 Shoes on Shelves and the Richard Mines both had really interesting settings, but were criticized for other features. In fact, there is very little to distinguish the Richard Mines from Swigian besides their length and the number of synonyms for things. (I personally enjoyed both those games)

The beta testers really helped fill in all of the responses and synonyms; the beta testing lasted several months. There was no way I could make a ‘polished’ game without extensive beta testing.

When I say I want to copy this style in the future, I mean that Color the Truth, Absence of Law, and Ether have big text dumps and are full of stuff, which I’ve realized can be overwhelming and overstuffed. I’d like to make more games that have a high action-to-text ratio.

For what it’s worth, that’s exactly the kind of game where I can believe some voters gave a 7-8 on the strengths and others gave a 5-6 on the weaknesses. (Not fatal weaknesses, but a failure to pull together which might be remedied in a post-comp release.)

My sense of the whole “long games do well” thing (both in IFComp and other festivals) is that voters tend to reward visible effort. (Other things being equal!) And game size is a proxy for author effort. We all know it’s not a perfect proxy; sometimes the work is invisible behind the scenes. You might go through a branching-narrative game once and have no idea how densely it branches. But we still have this judging bias.

By “we” I mean “I”.

Anyway, it sounds like your plans ran into a double whammy of (1) putting in a lot of effort that showed, even if it was packed into a few hours and (2) habitual good writing.

EDIT-ADD: To be clear, I have not played Swigian; I haven’t played any of the comp games this year.

Since reviewers responded differently to the two games, I think it’s worth digging into what did distinguish The Richard Mines and Swigian, especially given your purpose in writing Swigian. Reviewers who reviewed both indirectly highlighted differences other than length and synonyms.

I beta-tested this. The experiment’s not over until we see the results. Swigian has merits of its own, but there are multiple games I think it shouldn’t rank above under any circumstance. We’ll see what happens.

I’ve often thought of the good short book vs average long book and how people rate them. I tend to rate good short books highly because they don’t waste my time, and I think many of us like long books because we feel accomplishment getting through them. They sort of tell us, I want you to read me and get through me, and if that’s not abrasive, we see that as someone wanting our time and using it well. While with a short book, we may be mad there isn’t more, that the author is lazy. While the truth may be they recognize they have nothing more to say and hope the reader can move on quicker to the next cool book.

Plus there may be reticence on the part of judges to say “it was just too long, you know,” knowing how much time a person put in on a work without expecting vast fame and fortune. It seems a bit rude.

I didn’t actually play Swigian, but it looks like you caught onto some extenuating things that made it good despite your intentions.

I think 21st place is the best possible place for Swigian. While I would like all of my games to place as high as possible, I didn’t want this one to place close to my real one.

I think this shows that my hypotheses were off; that game writing cannot be reduced down to this specific formula, even with my inadvertent benefits. I think it also does service to all those authors whose excellent work over days or weeks got them a higher place.

Do you think it’s appropriate to add “comedic tone” to the formula? There are some exceptions, but it does seem to be a Thing.

CMg has an interesting theory about that; I’ve always thought it was just comedy that was part of the formula, but he thinks that people just like to do weird or crazy things in games. I think he might be right;

Yeah, even in the most dramatic parser games, players will attempt downright silly actions. Just because they can. Because the command prompt is an invitation. If these actions fail, they might not be surprised, but they’ll still be slightly disappointed. Even if the action is totally inappropriate for the game.

If a silly action succeeds, however, the game earns points. Even if it is inappropriate.

Games whose plots accommodate such silliness therefore have a natural edge, because more silly actions are likely to be implemented. Even a game like Coloratura, which is serious in general, revolves around bizarre and often inherently silly interactions, like the puzzle involving the frozen steaks.