Very few top games in IF have been optimisation ones. (For example, Verdeterre-esque games are a form of optimisation: restart the game, figure out how to maximise your score, then continue. There are other firms as well, such as IMO All Things Devours or The Lurkening.) The problem with them is, it’s hard to get players to play again once they’ve played through once, if it’s a large enough game that that is applicable. It also has to be, to a certain extent, addictive (note the last statement - not actually addictive, but you know what I mean).
So what makes them so rare? And what kind of things do you like to find in an optimisation game, and what things to avoid? Think about anything: from Orevore Courier to Suspended.
I think the key is minimizing the pain points as much as possible. If you want players to try over and over, you need to make it easy to do so. Labyrinthine Library has a first phase where you set up your puzzle, and a second phase where you test it (and win and lose). So I added a special command that sets you back to right before the phase change, so you can modify your setup a bit without starting from scratch.
I tend to like optimization games more than optimization puzzles, largely because there’s a greater commitment to the game’s worldbuilding and core loops. Nelson calls this a game’s “magic,” though I prefer a craft term of my own choosing: “cool stuff.” This is a matter of scale, of course. A small optimization puzzle could find a home in many games.
I think Suspended works well because it has an excellent gimmick (multiple playable characters with different abilities). It fully commits to the gimmick with different descriptions and responses for the various characters. It is vaguely horrific, which is part of its appeal. Suspended has excellent framing, in other words. The different perspectives encourage experimentation.
As Daniel suggests, minimizing friction that doesn’t add value is important. Suspended allows the player to “go to” locations, ask multiple characters to act at the same time, and so forth.
If everything else is interesting, then a score may be all that is needed. I kept playing Suspended until I got the highest score (without sacrificing a robot) and felt pretty good about it.
I think it’s harder for large optimization problems to work in games where that isn’t part of the magic. I don’t like the endgame of Trinity, for instance, because I want to explore it, not run through it. That isn’t to say that such set pieces never work, but the bigger they get, the more challenges they may have feeling at one with the magic of the game.
Games where you’re intended to play over and over, learning from failure and doing things a little differently each time, until you get the “best” outcome (or the best outcome you can—some of them have leaderboards where people compete for the highest score).
The distinction is valid, but this can lead to a matter of taste. Scott Adams had a good early attempt at an optimization game with Ghost Town, where there is an in-built mechanism of bonus points that required hard sequencing of the whole game after completion, but it was completely optional. Suspended has already been mentioned as well, and there the optimization is required more due to the intrinsic restrictions in game timers. The point is that optimization games tend to be more difficult (sometimes very much so) and therefore they are a bit of an acquired taste. This makes them less popular, but they have their public (Ferret is way older, but I wouldn’t class it among the lot).
Suspended went away with murder due to the excellent framing and the innovative game mechanics for the time (and it is still quite difficult to play end to end, to this date). Others have fared variously. Examples can be given: A Beauty Cold and Austere, Damnatio Memoriae, Sugarlawn, Transparent or Junior Arithmancer are all, from a certain point of view, optimization games, and to me, they explore the trope with very different degrees of success. I’m being deliberately vague as some are authored by people in this forum.
Academic mode on: Fred Brooks’s classic 1986 paper No Silver Bullet is credited with introducing the concepts of essential complexity (the complexity which is fundamentally part of the requirements to be satisfied or the problem design) as opposed to accidental complexity: the one which results from the implementation of the design and the operational constraints such implementation carries. Optimization games delve in essential complexity, since they are designed to be solved in a very specific way that satisfies all boundary conditions. Optimization puzzles can be accidentally complex, because the player is not always conscious of all the variables that have such constraints and can confuse not being able to solve the puzzle because some of those fail (e.g., the PC’s lamp runs out of batteries while trying to solve it, as opposed to the puzzle involving a mechanism that actively drains the lamp). Good design makes it obvious when complexity is essential to the game. Pete Killworth explored these concepts as applied to IF in the venerable How To Write Adventure Games.
Everyone who played it is probably thinking of Museum Heist as the archetipically optimization game, but I disagree, because its complexity is more accidental than essential. This is debatable and I won’t expand too much on it due to spoilers. It is still frustrating and requires clockwork play over many attempts, which is something that I can be in the mood for and actively seek. But not too often, and I don’t expect the game to be exactly popular. It may be well-known, but that is a different metric.
I think having to replay is something many players don’t like. And that means less people want to write one.
An optimization game I love is “An unexpectedly green journey” (I probably got the title wrong, I’m talking about that orc life sim.) Because there are so many choices and they matter much. Maybe it’s only half an optimization game and half RPG.
Well spotted, because it matches @Draconis’ definition: it is designed to be played many times until you get a “satisfactory” outcome. However, it has no puzzles and no constraints, and does not match what immediately comes to mind with the term “optimization game”.
I’d say Aisle is a conceptual outlier because its purpose is to provide a proof of concept for a certain idea (in the boundary between interaction and exposition). But all its complexity, when I think about it, is essential. Probably has to be on a class of its own, like the Basque language.
I can’t explain exactly why, but Suspended is to me THE optimization game. The Prongleman Job is great with a special twist (no more spoilers). The more I think about it, the more great games I remember. None mentioned, none forgotten
I think the length of the game is a big factor—if it’s fairly short, I’ll be much more likely to replay than if it took hours just to go through once. I can’t think of any optimization IF I’ve played that is that long, though, so I have replayed to try to improve my score/outcome in all the ones I’ve played. These games tend to appeal to me because I like being able to experiment, find what works and what doesn’t, and then try to improve.
I think this might be because these games are ultimately pretty mechanical to play—you end up focused on making the right moves, and the narrative fades into the background. Usually the narrative is only there in the first place to create a framework for the central puzzle. So while they may be quite fun to play in the moment, I don’t think this type of game tends to be the most memorable. (Suspended seems to be an exception to this, but I haven’t played it yet so I can’t speak on it personally!)
I also think they’re generally long games, at least for the optimization part. Sugarlawn certainly was. So at least for IFComp, you probably have barely gotten to the “real” puzzle of the game when you hit the 2 hour judging limit…
I think Victor Gijsbers’s review has a nice rationale for how the different layers of the game might get crunched together for someone playing the game for IFComp and how that might be to the game’s detriment.
So… I suspect they need more room to breathe than the IFComp gives them.
Ah yeah, Sugarlawn is a game I somehow encountered early in my IF-playing career, so I’d probably be much better at it now, but because it was so complex I only replayed a few times and was far from completing it. I definitely never encountered the protesters mentioned in Victor’s review.
Adjacent to this is the idea of the accretive player character, where the player is supposed to fail repeatedly and gain more information from each playthrough.
If you want people to learn from multiple playthroughs, you have to minimize the friction involved in starting a new run, making it simple and easy to try new solutions. So you can’t drag your audience through tedious, unskippable exposition every single time, but you also don’t want it to be easy for them to miss important plot details. It’s a tough balancing act.
Maybe they don’t end up being popular because they’re difficult to execute well?
I’d also point out that Captain Verdeterre’s Plunder came in 4th in the 2013 IFComp, and Sugarlawn took 4th place in the 2019 IFComp (and 1st place in the author-only voting). The Curse of the Scarab was 3rd of 7 in the 2020 Ectocomp Grand Guignol category, which is also not shabby. The Lurking Horror II: The Lurkening has 4.5 stars on IFDB (26 ratings and 5 reviews), so does All Things Devours (92 ratings and 13 reviews).
So this thread feels like a little bit of a red herring to me: I don’t know that it’s fair to say that they’re not popular. They are somewhat rare, but I suspect that’s more about them being somewhat tricky to make and not lining up with the interests or skills of the average IF author?
Yeah, I think “hard to make” belongs in the mix somewhere. At least, it would be hard for me. I also think that building a game that really shines upon replay is a risk that some authors may not want to take. It’s a gamble if the reception context involves playing multiple games in a constrained time window (competitions).
If I look at my favourite optimisation game, Curse of the Scarab, it is really dense and really fun.
In the circumstances of repetition, long prose is probably one of the frictions people in this thread have mentioned. This is a game with:
shortish prose
not a high number of locations, but you find more the further you get
very dense locations. There’s a ton of use for each location, even the first. The uses are both immediate and longterm. You won’t know or care about some of them until you know the game better.
mobile and hectoring, but predictable, enemies, whose range is small
limited resources (light etc.) put finite scope on the game, and make sure it can’t drag. It takes a while to discover the full scope, but once you know it, then you know what limits you’re working within
This is an interesting thread, especially seeing all three of the games I’ve written mentioned as optimization games in some sense. Personally, out of my games I only think of Sugarlawn as an optimization game, but I also think a game needs a time constraint in addition to an objective function in order to be an optimization game. So it depends on your definition, I suppose.
I actually found Sugarlawn easier to write than my more traditional puzzle-heavy IF game, A Beauty Cold and Austere. The least fun and so biggest obstacle for me in writing IF has always been the “polish” aspect - trying to think of all reasonable actions a player could try and coming up with appropriate responses. With Sugarlawn I didn’t have to worry much about that; instead, it’s pretty clear which actions a player needs to take - almost solely movement, picking up, and dropping. The tricky part - the optimization part - is figuring out which of these actions to do when. (For example - Do I run to the second floor and get all the valuables there now, or is it more efficient to go through the south wing first?) It was really nice spending less time worrying about heading off potential player guess-the-verb problems.
The other thing about writing an optimization game is that it was easily the most fun I’ve had playing one of my own games. When you write a traditional IF game, you know the solutions to all the puzzles in your game. But with Sugarlawn, the optimization scheme was complicated enough that I didn’t know what the best solution was. (I still don’t!) I fondly remember finishing Sugarlawn and then sitting down with the game for a few hours, trying to come up with the best solution I could.
So, to anybody out there considering writing an optimization game: Writing and especially playing it yourself may be more fun than you think.
