Adventure Gamers - Forums

Hi! I signed up just to reply to this, so forgive me if I spew out a lot of thoughts all at once.

I’m a mystery writer, editor (mostly developmental - I help people make sure their plots make sense, especially in thrillers and mysteries) and puzzle designer, so I spend a LOT of time thinking about this sort of thing. And I’ve just been hired to help with my first adventure game, so I’m very excited and this is very much “on my mind” (It’s very early days, and I’ve signed an NDA, unfortunately, but I don’t think it’s giving away any secrets to say that I’ve spent a lot of the past week thinking about different sorts of “open” puzzle design.)

To start with the mystery example… I always used to think that the mystery and the adventure game were a perfect fit for each other. After all, traditional mysteries are basically already games. In a “fair” mystery, there will always be enough information to solve the case before the climax. So Poirot is a good example here, but Holmes isn’t. That’s not a criticism of Holmes stories, they’re just a slightly different genre. (For adventure game purposes, we’re obviously interested in the Poirot model. The player gathers clues and then puts the pieces together to solve the mystery.)

And to take it further, a lot of police work also seems to have obvious mystery game analogues… searching crime scenes, comparing fingerprints etc. etc. They seem perfect for turning into puzzles.

It seems like it should be childplay.

But of course it isn’t. I think the apparent similarities are, if you’ll excuse the expression, a red herring. As Iznogood rightly points out, playing a mystery adventure game or a police procedural game doesn’t feel ANYTHING like being a detective. There’s no real sense of deduction. By splitting the process of solving the crime into smaller parts and giving you confirmation of success at each stage, you remove any sense of achievement.

But how do you fix it? I’ve thought about this on and off for years. I agree that the Poirot example seems like the ultimate goal. You’re it’s at the very least a long way off, but I think it’s worse than that. I think it might actually be impossible. How do you check to see whether someone has solved a mystery, without reducing it to a quiz you can trial-and-error your way through?

Laura Bow, dagger of Amon Ra tried one solution: make the quiz so long that brute force is impossible. But that was a non-starter, because many of the questions asked were totally arbitrary, and a single wrong answer lead to failure, which makes no sense in the context of a police investigation.

And what does failure look like? If, at the end of Death of the Nile, Poirot had got it wrong, how would that play out? At the end of Cruise for a Corpse, selecting the wrong killer results in instant death (if I’m not misremembering), but in-story that makes no sense at all.

But if not a quiz, then what? How do you present your evidence, accuse your suspect etc. etc.? I’m not trying to dismiss the idea - it’s what I’ve always wanted to see.

More broadly, I just think that interactivity and mysteries might not mix. This is very counter-intuitive, I know. How can the genre that is most like a game not work as an actual game?

I think the reason lies in what TimovieMan said: the player is NOT a great detective. And neither is the reader in static fiction: In a mystery, the reader is rarely asked to identify with the detective. The narrator is usually the assistant - Watson, Hastings or whoever.

When I write a mystery or advise my clients about clues/complications/twists etc., I have a rule of thumb: a third of people should be able to solve the mystery in advance, a third should solve it just before the detective does and the final third should be completely surprised. If you try to surprise everyone, there’s a good chance that some people won’t understand what happened. It’s always possible to be too clever! In the thirds model, everyone gets to feel something good - some people feel smart, some people get that “phew, of course” feeling of getting it just in time, and some people get a great shock.

But in an adventure game EVERYONE has to be able to solve it. (Unless you want a game that 2/3 of people will never complete.) If you don’t solve the mystery in a book, the book carries on regardless. Playing along is a big part of it, but it isn’t the whole point. In an adventure game it would be. So how do you match up these two problems? In the world of the story, ONLY Poirot is able to solve the case. How do you write a mystery worthy of Poirot, but also structure it so that everyone will get there in the end, all without leading the player by the hand. I wish I knew…

With regards open puzzles in general, though, I think you’re right that you want lots of different ways to reach the same solution. This works best with information/research based puzzles like the password you mentioned. Say you know the password is the name of X’s dead wife’s maiden name. In a traditional adventure format, the name will be written on the back of a photo or something. It doesn’t feel like investigation. It feels like unlocking a door. But if the name is there AND you can get someone drunk and let it slip AND you can steal someone’s address book and find it AND you can search in the library catalogue using details you’ve overheard to bring up a newspaper article etc. etc. Well, now we’re getting somewhere.

The point is that those puzzles can individually be obscure, intermittently clued or require items that you might have missed. But if you have enough of them, then the player is bound to have some way of progressing, and it will feel like they worked it out for themselves. The key is variety and options, and not being tied to the A unlocks B unlocks C combines with D to unlock E model.

I don’t know if people play a lot of IF (“text adventures”) on this site, but The King of Shreds and Patches does quite a good job of this

Iznogood - 04 March 2013 04:50 PM

Lucien21 - 04 March 2013 04:21 PM

The problem with these types of things is that the more option you put into it the more variables there are in the game and the more dialogue, story parts, animations etc etc you need to build into the game that most people will never see.

I agree, i think that this is the main reason why we haven’t seen more of this. Not because nobody else have thought of it before me, or because it can’t work properly.

This is the main issue. A moment there’s a very entrenched view on both sides that people must see ALL the content. Designers are reluctant to “waste” resources on things that people won’t see, and gamers feel that they’re being “robbed” of content if their actions somehow gate off sections of the game. But this is false reasoning on both sides, especially from the designers: making a great game that many people see half of is obviously better than making an average game that a few people see all of.

Anyway, I think a lot of modern adventure games are stuffed with padding anyway. I think with strong design, you could have the same amount of dialogue but repurposed so that it was all actually useful.

I missed After a brisk nap’s post which covered a lot of what I said. I haven’t played Resonance, but that’s the sort of puzzle I mean. Actually, Blade Runner is a great example of what I’m talking about. In that game there were usually three or four ways to find the information you needed to progress the investigation to the next stage.

Iznogood - 04 March 2013 05:11 PM

The beauty of the Password example, is that it actually requires less work from the developers then the more traditional design, all they have to do is make sure there are clues the player can find, and that the user can type whatever they want.

I’m not convinced of that.

If you make it possible for players to solve the puzzles “before they’re supposed to”, you have to make sure that the rest of the story and puzzle logic doesn’t get broken. Let’s say you’re able to access a file before you speak to the person who was supposed to give you a crucial clue to the password; now you have to remember to write dialog for the case where you already have the information, or the conversation won’t make sense.

Or let’s say it gives you an address to a new location, where there’s going to be a forced puzzle (e.g. you’ll get trapped and have to find a way out); now you have to make sure that players can’t go there without the information/inventory items they need to solve that puzzle, or you’ll have created a dead end.

Or imagine that something was going to happen once two conditions were satisfied, like having found an inventory object and having accessed this file. For example, the phone next to the computer starts ringing. Before, the puzzle logic could have guaranteed that you had the object by the time you’re able to break the password and access the file, so you’d be in the right room when the phone starts ringing. But if you access the file first, you have an arbitrary and unfair trigger after you pick up the inventory item, where you’ll be stuck until you randomly return to the room with the phone.

Of course, it’s possible to structure the game to avoid cases like this, and more non-linearity is often a good thing, but it’s clearly something that creates more work for the designer. You have to take more and more possible cases into account.

It is more work, but - if I can be snarky for a moment - it maybe feels like work the developers should be doing. At least with this structure you’re forced to sit down and work out exactly how to clue a puzzle and what role it’s playing in the logic of your story. With the current gates/keys method of puzzle design you can just chuck together any old tat!

Bladerunner got round the problems you mention by not having any items (except the gun, which you couldn’t lose). If the unit of puzzle currency is “information” rather than “items”, I think it’s much easier to avoid dead ends (although maybe not non-sequiturs in dialogue). But maybe having no item puzzles is too far from most people’s idea of what an adventure game is?

Iznogood - 04 March 2013 04:38 PM

I am not claiming that everything i wrote is revolutionary, and not seen or discussed before. Also my main point is not so much in the password puzzle, that is just the smallest baby-step you can take, in order to reach my Poirot example. Also i do believe it changes the experience, perhaps only in a small and subtle way but still ...

I focused on the password example because I found it more understandable and more compelling, while the “Poirot” example just didn’t sound like a very good idea to me.

Part of the reason it doesn’t strike me as a big departure from traditional adventure game puzzles is that I’ve played a number of games with puzzles like that. For example, most of the games published by Wadjet Eye (the Blackwell series, Gemini Rue, Resonance) have some kind of free-input puzzles (e.g. search engine puzzles, phone book puzzles, password puzzles, decryption puzzles), or at least puzzles with multiple solutions depending on which clues you came across/noticed. It’s nice and feels natural, but really the “openness” is barely detectable as a player. It’s probably not going to jump out at you as different.

After a brisk nap - 04 March 2013 03:54 PM

does that really sound like fun to you?

No but that is also the point, it is not meant to be funny, you are meant to be punished for your stupidity not being a great detective, by having to do some leg-work instead simply because you are an idiot failed to do the brain-work. Of course after a while it has to become obvious that you are on the wrong track, otherwise it would be terrible.

How does this make the game better?
How far should players be allowed to stray down the wrong path before they’re guided back? If the goal is to make adventures more accessible, isn’t this the exact opposite of that?

This is also not my own idea, i have just adopted it. In The Secret World there are investigation quests that work similar to puzzles in AG. In one of the early ones, it works almost exactly like i described, if you get your thinking wrong, you are sent on a wild goose chase, and only after you have first decoded the place you need to investigate, and actually investigated the place, only then are you even allowed to redo you thinking. (I would omit the last part, and allow you to rethink things anytime you want).

Well, multiplayer online games work a bit differently from AGs, with grinding and teams and all of that. Is the “wild goose chase” as completely pointless as it would be (by the sound of it) in an adventure game?

The point is that i don’t want to be told the story, or at least not who the murderer is, i want to discover that for myself. But this is also a general challenge that exists in all interactive story telling.

Also not all games might be suited for a more open story, there are some stories that require a more sequential approach.

And my point was that in a mystery story - to simplify and generalize a bit - the appeal is the mystery of the case (encouraging speculation), the surprise of the reveal of the solution (or satisfaction of having been proven right), and admiration for the detective’s brilliant process of deduction.

If you’re the one working it all out, step by step, then it’s hard not to dissipate the tension of the mystery, it’s hard to achieve any level of surprise (unless you ultimately solve the case with a single, smoking-gun clue), and since we’re not all Poirots or Sherlocks, the case won’t have been so baffling or the deduction so brilliant that there’s much to admire.

This really goes to somanycrimes’ argument that mysteries and adventure games aren’t as compatible as they seem on the surface. The types of design you criticize (playing as the assistant rather than the detective, for example) are attempts to address some pretty fundamental problems with the combination.

somanycrimes - 05 March 2013 12:24 PM

It is more work, but - if I can be snarky for a moment - it maybe feels like work the developers should be doing. At least with this structure you’re forced to sit down and work out exactly how to clue a puzzle and what role it’s playing in the logic of your story. With the current gates/keys method of puzzle design you can just chuck together any old tat!

Bladerunner got round the problems you mention by not having any items (except the gun, which you couldn’t lose). If the unit of puzzle currency is “information” rather than “items”, I think it’s much easier to avoid dead ends (although maybe not non-sequiturs in dialogue). But maybe having no item puzzles is too far from most people’s idea of what an adventure game is?

Oh, it’s stuff the designers should work out, definitely, and you have to do that work even in a fairly linear game. But like I said, as you introduce more unpredictability, the number of potential paths through the game explodes, and it becomes much harder to check them all and provide appropriate interactions, responses and dialog for every possible situation. It may well be worth the effort, but I think Iznogood is dead wrong that it’s less work overall.

After a brisk nap - 05 March 2013 12:51 PM

And my point was that in a mystery story - to simplify and generalize a bit - the appeal is the mystery of the case (encouraging speculation), the surprise of the reveal of the solution (or satisfaction of having been proven right), and admiration for the detective’s brilliant process of deduction.

This is a very good point that I forgot to mention. The point of most mysteries is that you’ve got all the information in front of you, but it’s only by looking at it from an unusual angle that you can see how the pieces fit together. The best mysteries work by turning assumptions upside down. Once the solution is revealed, everything that didn’t make sense should suddenly coalesce into a new image, like twisting a kaleidoscope.

That’s just not possible to replicate in a game where the player has control over when the solution is sprung.

I think I disagree that it doesn’t sound like a good idea (or at least a nice dream), but only in the most general sense. I love mysteries, and that “drawing room showdown” is an experience I’d love to be able to experience or create. But I think the reason why it works is because the author is in complete control. As soon as you abdicate some of that control to the player, all the dramatic force is gone.

I usually point to the Phoenix Wright games as a series that manages to solve this problem quite successfully. In case you don’t know the games: you play as a defense attorney, there’s always a seemingly air-tight case against your client, but you have to show that he/she’s innocent, identify the actual killer and prove it in court.

At each step, you’re addressing one element of the case (e.g. a witness saw your client kill the victim). You have to uncover contradictions in the evidence or in the testimony (the games have a cool mechanic for this based around cross-interrogation). Once you’ve done that, you learn new information, which often changes the case completely. So you’re making small breakthroughs and getting small rewards all the time.

After you’ve gone through a few rounds of this, there’s a break where you do some more investigating. While you’re doing that, new setbacks usually happen (e.g. the prosecution finds another witness, or another person is murdered, or your client decides to confess), so even if it looked like you were about to win before, by the time you get back to court it looks like you’re just as screwed. But as you gradually work through the layers of lies and misleading evidence, you start to discover the truth, and finally you’re able to make a slam-dunk case against the real murderer.

So in this design, you are doing all the “star” detective work (investigation, deduction, facing down suspects, final showdown with the killer). You’re making real progress and figuring stuff out all along, but because there are so many twists and turns, you can’t work out the solution until quite late in the game. And because the cases are always so baffling at first, it does actually seem impressive when you finally solve them.

The games don’t quite manage to provide sudden, surprising reveals (by the time Phoenix Wright produces the smoking gun, you’ve usually worked out who the killer was and how he/she did it for a while), but all the other character do a good job of acting as if it’s all really shocking, and that helps a bit.

I think there are four key elements that make this design work so well:

1. Baffling cases. Often some kind of locked-room mysteries, where your client is the only possible suspect. As initially presented, they seem open-and-shut, and your job looks impossible.
2. Playing defense. Your “official” job isn’t to solve the case, but to defend your client. The prosecution controls the pace of the case in court, you just have to knock down each piece of evidence or testimony as it is presented. This breaks the case down into manageable sub-tasks, while still keeping the stakes high.
3. Constant confrontation. Throughout the game you’re facing up against lying witnesses, unscrupulous prosecutors and a hostile judge. In many ways, every cross-interrogation is like the climax of a classic mystery story where the detective confronts the suspects and pokes holes in their stories. It allows - indeed requires - you to demonstrate your deductions, keeping the game and the player in sync and on track in solving the mystery. And because the other characters are against you, it makes sense that they knock your claims down if you get it wrong.
4. Twists and turns. Layers upon layers of mystery. Nothing is what it seems, almost every revelation changes the case radically, and opens up a new mystery. That means that even if you see through the current level of obfuscation, you don’t automatically know the solution to the whole case.

Plus it helps that the games are comedic with absurdist elements (e.g. interrogating a parrot), which makes it easier to swallow some of the more outrageous developments. Other detective adventure games would do well to look to PW for inspiration.