Choice-based puzzles

What interesting puzzles have been done or could be done in choice-based games without a parser?

In a printed gamebook I read the other day there was a puzzle based on a classic logic riddle (involving two paths, and two goblins from two different tribes, one of which being known to always lie and… you know the rest). Guess that kind of logic puzzle is always possible, as long as enumerating some possible answers can be done without making it too simple.

Mazes is always a possibility of course, but who wants to do that again?

In gamebooks based on numbered paragraphs (common in printed books, less common in choice-based computer games) there have been some puzzles involving figuring out a number and then turning to that paragraph. Difficult to do with pure choices because it relies on that particular implementation detail.

Perhaps it would also be interesting to mention some examples of puzzles that work well with a parser but could never work with choices (assuming that we do not count “guess the verb” as a puzzle that ever “work well”)?

I suspect that most parser-based puzzles are at least possible in choice-based games, because you can replicate a surprising amount of parser in a choice-based format. (For example, I can picture a choice-based implementation of Spider and Web without too much of a headache.)

Two of the XYZZY award nominees for Best Puzzle this year are choice-based games:

Earning One Million Dollars in ULTRA BUSINESS TYCOON III (Porpentine)
getting out of the lab in Chemistry and Physics (Colin Sandel and Carolyn VanEseltine)

Curious examples, because they are not typical choice-based games. UBT is more of a simulator, and C&P is reliant on the same model (rooms and objects, and travelling between the rooms at will) as parser-IF, a model which is often different from choice games.

Also, though I can also imagine a choice version of S&W and indeed almost every IF game, I can’t imagine that choice version being as fun, or having as much appeal. Even if you construct it so that you can choose, say, that on some objects you can use some tools, and come up with a list of those tools, and select one, and then it either works or it doesn’t… it still limits the scope of what you’re doing, and doesn’t have a key factor in the experimentation of most parser IF, which is manipulating an item in different ways (prompted by clues in the text, or by common sense) to see how it reacts, what its properties are, et al. Yes, this could all be made into a choice-based system, where say the description of an item would have all the properties aforementioned… it would be more streamlined, there would be more handholding, and in a sense it would be dumbing it down.

But I guess all this proves is, some games are better as parser IF and others are better as CYOA, and adaptations often suffer.

A horrid thing in most CYOA, or choice-based, is the inability to save your game. That dictates the length of the game and the difficulty - existence? - of the puzzles. If you’re going to have to solve that puzzle every time you boot up the game, just to return to where you were last time… well, personally I didn’t finish C&P because of this. Too much possibility of failure, failure meant restarting everything, and to people who say “Life doesn’t have a save feature” I reply “That’s why I play games”.

Finally - I was actually thinking more of “Trapped in Time” for puzzle mechanics in a choice-based system. It expounds on the “figuring out a number” example pelle brought up (which I actually saw in action in some Fighting Fantasy books).

Oh, one more thing: Jon Ingold’s blog might be of interest as he’s recently shared his thoughts on Choice games and how they might be better at IF than parser IF.

In another thread I pointed to Paradox Factor … ox-factor/ which contains systematic time-travel puzzles. Actions you take in the past affect future events.

The one counterexample I can immediately think of is All Things Devours.
EDIT: Although I’ll have to look at Paradox Factor.

Conversely, Alabaster is effectively a choice-based game already (albeit one with a huge number of choices per node). I don’t think that one would have been as enjoyable without the “threaded conversation” system guiding my options.

Several techniques I’ve seen work:

Chemistry and Physics, ULTRA BUSINESS TYCOON III, Quit Your Job Simulator 2014, and Bigger Than You Think all include puzzles that depend on inventory, where relevant actions become available only when you have the right thing. (This is so common that I’m sure there are many other examples – those are just the ones I happen to be thinking of at the moment.) Because the player has to both find the inventory item and revisit the situation in which that item could be used, the solving experience here is not really so very different from the solving experience in conventional parser IF: the player has to realize that there is some potential associated with the found object and take a second step (go to the right place to use it) in order to apply it. In some of these examples, (QYJS2014, for instance) you get an obvious option for using the object as soon as you have that object in the right location. In others (Bigger Than You Think), there’s an inventory system whereby you can always attempt to use anything in your possession – so that eliminates even the possibility that the player will accidentally stumble on being in the right room with the right thing, and requires that the player has actually thought of using it there.

ULTRA BUSINESS TYCOON III at one point requires the player to type in a keyword to progress, much like a gamebook having a password system.

Many Choice of Games stories allow the player to build up stats over the course of play that then affect success during later attempts. These are sort of on the borderland between RPG and puzzle, but you could view the whole system as a puzzle, in which the challenge is to work out what sequence of actions will place you in a position to be successful. A lot of dating-sim-esque visual novels also work that way: you have to figure out the characters well enough to see which choice sequence over the course of the game is going to optimize your connection to this or that possible romantic lead.

Save the Date and Bigger Than You Think track your actions on previous playthroughs and uses that to open up new possibilities when the same nodes are revisited; this tracks a bit with replay-focused parser IF like Slouching Towards Bedlam, where there are important actions possible within the first moves of the game, but no one who hadn’t played at least once already would think of trying them.

It gives me a headache. :slight_smile: I intended the tone of the game to be being surprised by possibilities, and thinking of unexpected options is part of that. Of course, I’m too close to the question.

(I’ve thought about a graphical implementation of S&W, but that’s a complete rewrite with completely different mechanics!)

My (parser) game that I suspect is most amenable to a choice-based rewrite is Heliopause. I’m a little afraid to try, in case it works… like I said, too close. Anyhow Heliopause isn’t really a puzzle game.

On the topic: I have to bring up Meanwhile, which uses some “find and remember the code” mechanics, but is primarily built on finding and understanding the story. Flipping randomly through the book (or app) is possible, but it won’t make sense until you’ve seen and understood all of the narrative branches (albeit not necessarily in the “canonical” order).

(I wish I knew how to do that trick. Meanwhile covers its ground so perfectly that it leaves the impression no other story can be built the same way. I’m sure that’s not true, but…)

There are games that use systematic mechanics to build interesting puzzles out of small, accumulative actions. I don’t know that field well enough to name titles, but I assume that’s what Dan meant when he mentioned Paradox Factor. In a sense the graphical Grow games (by Eyezmaze) fall into this category. (They’re more about exploration of crazy results, but they’re based on a simple combinatoric mechanic.)

howling dogs and Impostor Syndrome both contain puzzles that I can’t describe without spoiling them, so I’m going to describe them under a spoiler.

They each contain giant fields of links where only one or two links will take you to the Good End, though that’s a bit of a reductive description anyway. In howling dogs all the wrong choices had an anodyne response that took you back to the original link-field, and there was a link that let you move on (and thus fail the puzzle); there were so many anodyne responses that no one was likely to find the true one by link-spamming them all. In fact, I didn’t realize there was a puzzle until I saw it mentioned in a review. In Impostor Syndrome choosing a wrong link just takes you straight to the Bad End.

My advice is not to try one of these puzzles unless you have an interesting way of varying the implementation (though don’t let me stop you from thinking of one!); they’re much less effective when the freshness has worn off. And it’s significant that neither of these is really a puzzle game; getting what I not entirely seriously called the Bad End is not like losing or getting stuck in a puzzly game.

You can definitely do code-based puzzles in a choice game. Meanwhile (which is a book) does this. It also has an optimization puzzle which combines with that. [On preview, zarf mentioned this, and also the point about S&W in the next paragraph]

Puzzles that don’t work in choice-based formats; I’m surprised Carolyn cited Spider & Web because its most famous puzzle, which may be the most famous puzzle in post-Infocom IF, is commonly cited as one that wouldn’t work without the parser (I agree). My Faithful Companion (one of the parser-based Best Puzzle nominees!) wouldn’t work in a choice-based format either; the central mechanism requires that the player be able to perform a wide range of actions, and… well, mild spoiler

the nominated puzzle requires that you think of doing something which it’s implicit that you can do, and when you think of it it’ll be obvious that you ought to be able to do it (or try it), but if it was presented as an option that would spoil much of the puzzle. More details here but it’s a very short game, why not try it?

spot on

anyway, Hallowmoor was a pretty good clickable text-adventure with some rather nice puzzles

doesn’t matter much as most of them are pretty short anyway (specially as they mostly don’t rely on puzzles, just mindless clicking or branch exploration)

that said, COG stories on mobile start from the exact point you left (a feature I also enjoy in a few mobile parser IF terps)

and twine stories usually grow the url with each link visited. That works as a savepoint.

OK, let’s talk about Spider and Web. There’s just no way to discuss its central puzzle without spoiling it horribly, so that’s what I’m going to do.

[spoiler]I’m assuming we’re all talking about the puzzle where you “say tango,” using your voice transmitter to activate the acid pack. Your inventory is mostly visible on the desk, but just the voice transmitter and the acid pack are missing. You must have the idea, on your own, to “say tango.”

But, really? You can’t say just anything in this game. You can’t even say anything except “yes” or “no” to your interrogator. You can’t even “say tango” to the interrogator before the crucial moment; if you do, the game plays dumb, saying “Talking to yourself might draw attention.” And it’s not even in italics, which is how the game narrator normally suggests hints like these.

In a recent blog post, Jon Ingold wrote: … teresting/

Concretely, that means that there would be an on-screen option to “say tango” in a not very obvious place. For example, you could put it in a wall of links, like howling dogs or Impostor Syndrome. You could put it in an “inventory of thoughts” like Chunky Blues. You could have an on-screen verb bar, like old LucasArts or Sierra games did, with verbs like “examine” “use” and “speak.” If you speak to yourself, you could have the tango/waltz options there, or put in a wall of options there, if you’re trying to be tricky.

You aren’t meant to solve the puzzle by brute force, of course, but ultimately, there’s a finite list of things that you know you can do in this game. Eventually you try it. As for me, as soon as I saw the acid pack in my inventory, I tried restarting the game to “say tango” at my first meeting with the interrogator, and I kept trying it, bull headed, until it eventually worked. It wasn’t exactly brute force, but it didn’t really feel like the moment of triumph I think other players say they had in that game. (I even had to post here to make sure I knew which puzzle people were so excited about.)

Or, if you must, use a hybrid system, where when speaking, you have to type the word you want to use. “yes” “no” “tango” and “waltz” would be the only valid options, but at least “tango” wouldn’t be directly visible on the screen. Is that all there is to a parser?[/spoiler]

Through an odd coincidence my computer went down right after the guard brought my equipment, so I had a few hours with a hard-copy transcript and no working interpreter. Four things were missing, and one of them could be activated while in the chair.

Matt w:

There’s a puzzle in howling dogs? Please tell me more, I must have missed it completely. I merely played it to the end.

You remember the page (in the queen sequence) where there are a bunch of links, and each one says “How interesting!” and sends you back to the original page? One of them doesn’t.


So the “good end”, and indeed the path that continues the story proper and helps give it coherence, is buried in a mountain of links in a block of text amidst links that give you no reason to believe they’ll be any different from one another. Indeed, no reason to believe that any of them would be more fruitful than the link that let you move on.

I suddenly have much less appreciation for “howling dogs”.

Responses within the spoilers:


Of course not. Nobody is saying that the parser ought to allow you to say just anything. The strength of the parser as far as these puzzles go is in hiding which choices are available, not in giving you unlimited choice.

Well, that’s not exactly how I remember it – how I remember it is that the italicized comments come when the narrator is actually lying about the narration you’re giving to the interrogator (and the interrogator isn’t catching it). There’s a system here. It’s true that the game doesn’t give you the chance to play your trump card before the moment when it becomes clear why it should work, even though the underlying world model strictly ought to allow this, but this is pretty common in games of all kinds (there are lots of point and clicks where you e.g. can’t attempt to enter a code until you’ve found the clues for it).

All these would make the puzzle much worse. zarf did do something like this in Bigger Than You Think, where you have to click a link at a non-obvious time – but there’s a very good reason for that link to be there. Cluttering up the interface with a wall of extraneous stuff so the solution doesn’t stick out too much is less satisfying than letting the player pick out an option for themselves.

Or you figure out what’s going on and you have an aha! moment.

Well, this seems like the equivalent of trying to eat the door and complaining that you didn’t get a convincing response. Trying something that looks like a solution every chance you get without waiting for the clues about why it would actually work [I mean, the sequence that tells you that you’ve had time to sneak into the interrogation room and put the acid pack in place] does seem like brute forcing to me, and that’s not as satisfying as solving it without brute force. The argument isn’t that it’s impossible to brute force the solution, it’s that the satisfying way to accomplish the solution is to realize “Hey, the acid pack is in place and I can activate it during the interrogation,” and if the option to speak the code word is up there as a hyperlink that realization will be much less satisfying.

For my part, Spider and Web was about the third IF game I played and I was absolutely appallingly bad at puzzles. I used a walkthrough a lot. But when I got to that point I had the aha! moment, “I can set the acid pack off!” Marred slightly by forgetting which keyword meant “on.” And that was satisfying.

Really, I think your story proves that the puzzle wouldn’t work with hyperlinks. You think the puzzle might as well have been a hyperlink – you also didn’t enjoy it very much, used a brute force approach, and tried to break it as soon as you could. I think it’s a great puzzle, think it wouldn’t work as a hyperlink, and had an aha! moment solving it; and I think this attitude is shared by a lot of people who enjoyed the puzzle. Doesn’t this suggest that whatever is great about this puzzle wouldn’t captured by a hyperlink approach? If you were like “This puzzle is great and here’s how I’d preserve its greatness in hyperlinks” that’d be convincing but to say “This puzzle is meh and we can preserve its mehness in hyperlinks” is less so.

I think it would be mildly suspicious to suddenly switch modalities and have the player type something in when it would be far easier to make links for “yes” and “no.”[/spoiler]

And a comment about Faithful Companion (which spoils the puzzle so again, I hope anyone who is reading this will try the game first):

[spoiler]You’ve used a key to unlock a door. The next room has a mechanism that would be trivial to work if the ghost weren’t interfering with it. It’s basically a red herring; you can spend a fair amount of time trying to work out a way to trick the ghost into helping you with it, but AFAICT it’s impossible. You have to lead the ghost back outside and lock the first door again. maga put the insight well; you have a key in your inventory and it can do two things.

There’s no guess the verb here, I don’t think; once you realize you want to lock the ghost out “lock the door” will work. But if “lock” were presented as an action it would be a giant spoiler.

I hope it’s not too immodest for me to drone on about this puzzle because it was created accidentally – when I started coding I thought you would be able to trick the ghost into helping you, and when that didn’t work I had to check in a panic that locking the ghost out would work.[/spoiler]

I don’t know if you said this after playing the game or on your memories, but I disagree. If you read the text closely you can see that one of the links is not like the others.

The howling dogs thing made sense to me.

[spoiler]I looked at the thing and was like “I bet one of these links goes somewhere.” and so I did a bit of clicking, then stopped and thought about what word might be intended to stand out. Took a few tries but it worked.

I don’t particularly see it as brute force.[/spoiler]

More howling dogs spoilers:

[spoiler]I have to agree with inurashii here. At first when I heard about the puzzle I sorta groaned inwardly, but as I replayed that sequence I realized that finding the right link was more obvious than you’d might expect. If you read through the big mass of links carefully you can find the one option that’s not quite right, so it doesn’t really require brute force. It’s actually pretty clever once you’ve figured it out.

That said, I can’t imagine another game using it nearly as well (haven’t played Imposter Syndrome yet).[/spoiler]

More on Spider and Web.

[spoiler]First, I want to defend my honor a bit.

I don’t think it’s fair to call that “eat the door,” or even to call it brute force. Once I saw the acid pack, I said: “oh, hey, this looks like it would be really useful during the interrogation. I wonder if I’ve already set it up? Maybe I can even use it during the first interrogation scene? Let me restart and try it!” It was even the right answer, in the end. So, I did have the “aha” moment, but wayyyy too early. I concede that the resulting feeling was not awesome.

(FWIW, I honestly feel that the game would have been improved by having a better message there. “You say the keyword, but the voice transmitter isn’t here.” Or, better, let the interrogator hear you use the keyword, then say his “you’ve planned this!” rant and kill you, as he already does if you admit that you’ve been in the interrogation room before.) Or, even better, let it work.

Anyway, replaying the game, I find that I misremembered the scene. The voice transmitter is present on the desk, and, indeed, that’s why trying to use the keyword too early doesn’t work. You have to trick the interrogator into bringing the voice transmitter into the interrogation room, so you can use it to escape.

So the simplest would just be to use an on-screen verb bar (“examine” “use” and “speak”) where if you “speak” to the “voice transmitter” you’d say tango and it does its thing.

Would this “make the puzzle much worse” as you say? I claim that the thought process is the same. (“Can I use something on the desk somehow? Oh, the voice transmitter!”)

Maybe my cold heart just can’t appreciate “aha” moments; I’ve been playing parser IF for years and I honestly can’t remember having a satisfying “aha” moment, here or in any other IF game. That requires the answer to be hard to discover but obvious in retrospect. In my experience, every puzzle I’ve faced has either been obvious beforehand or non-obvious in retrospect (when I brute force the puzzle or look at the hints).

I agree with Emily that the best puzzles are “systematic,” where the strategy to approach them is semi-obvious, but the solution isn’t accessible until you learn more about the system. Like the board game “Mastermind.” Your first guess is semi-random, but the responses give you clues that lead to the right answer. At no point in such puzzles do you smack your forehead and say “aha! of course!”[/spoiler]

Both. I’d played the game once a while ago, and once again three or four days ago.

Read the text closely? Fine, I guess this is where I seem to disagree, and I’m clearly in the moniroty so I won’t belabor the point, but I was given a mass of links, all of them gave me the same “no plot advancement” response", and up until this point the game was spectacularly linear (with some interesting variations, to be sure, but always linear and advancing in the same direction).

So why the heck should I, as a player, decide to read this huge unappealing mass of links carefully? What reason have I to believe it’s not an aesthetic effect?

I mean, in the context, all those tons of things all eliciting the response “How interesting!”, which demeans them to the status of casual, unimportant, inconsequential light conversation? It made perfect sense, it’s a good aesthetic effect on its own, like a cloud of bubbling society full of frivolous things hiding whatever may lurk beneath. It’s not gratuitous. I had no reason at all to believe it might hide something critical for the game.

Again, I see I’m in the minority and most people didn’t seem to have a problem getting past this. Maybe this sort of game just isn’t for me after all.