How I design adventure puzzles (and how do you??)

[RootBound]

So this grew from the thread about my current game-in-production and it made more sense to start a new conversation here.

What is your process for designing puzzles? Here's mine (which I'm looking to refine)

Designing puzzles is hard. Maybe that's why there's so little participation in the (shameless plug) Puzzle-Making Practice competition. My own process varies from game to game, so I'll stick to the current one.

The first question I ask is: What kind of game am I trying to make? Classic third-person P&C? First person? Visual novel? Even platformer? In the early stages of design, it may not always be clear what subgenre is the best fit for your project. And if you're not sure, puzzle design will feel extremely haphazard. (This won't necessarily change later.  ) But the kind of game you make narrows the kind of puzzles that can be put in the game (or does it? Can you, for example, solve a murder in a platformer? I'd love to try).

A second part of this is who are you designing for? Experienced adventurers, beginners, or a wider audience? How difficult do you want the game to be? Do you care more about story or obstacles, or both equally?

Once you have a subgenre (or two, or three...) and a target audience, then what?

What I do is:

1. Start with story. What is the player/protagonist trying to achieve, story-wise? It doesn't need to be complicated. It could be simply "Get to the other end of the dungeon," and that's the entire game, room after room. It could be "Solve a murder." It could be "Become president."  Here's what's important: the overarching goal will help determine the obstacles.

2. Where does the game take place? Setting has a huge effect on puzzles. Solving a murder in the wilderness is going to feel different from solving one in the city, and different from solving one on a spaceship. Get a good feel for your setting and the places for obstacles will become clearer.

3. Push-and-pull between puzzle and setting. This is where it helps to be flexible. Once you have a goal and a setting you can both refine the setting around the puzzles and also refine the puzzles around the setting. Play with them like clay until they fit together--add rooms, delete rooms, change what type of obstacle is in a room. But...

4. Do the obstacles make sense? Why is that key under the sink? Who hid it there and why? Or what about that bookshelf in the basement? Would the homeowner really want to go down all those stairs to get at it? If you don't know the ins and outs of everything, the player won't be satisfied with either the setup or the solution.

5. Is the puzzle solution too clear or too obscure? This is where I struggle. And where test players come in. As @CaptainD suggested in the other thread, it's really hard to gauge difficulty when you're the one making the puzzle. How do you give clues without giving the whole thing away? How do you know when not to give clues and trust that the player already has enough information?

I'm really interested in how others design puzzles in your own games. Do you have a set of principes you follow? A process of steps like (or unlike) the one I listed above? How do you figure out difficulty, clues, and teaching the player the tools without too much hand-holding? How do you make sure the puzzles make sense within the world?

I'm not so much interested in what to avoid (although that's important) but more so in what you gravitate toward. How do you do it? Are there reasons you do it that way?

Thanks in advance for any answers!

[Danvzare]

Hmm, interesting.
I'll gladly share my process, which I hope helps. 

Firstly you need an idea. This part is easy, because ideas are like dust. They're f*cking everywhere.

Next you need the story. Keep it simple, and divide it into bullet points.

Then you figure out how the player character will reach each of those bullet points, by subdividing them.
Then you subdivide those new bullet points, again, and again, and again.

And there you go, you have your puzzles.
If you're worried that the puzzles might not make sense, just add signposting. Anything can make sense if you give it context.
Showing a beetle to a llama to get a newspaper might not make sense, but it does if you establish an aggressive llama sitting on a newspaper as being afraid of beetles.


Something a lot of creators mistakenly do, is separate the story from the puzzles, usually with the puzzles being roadblocks to proceed in the story. Which I believe is the wrong way to go about making an adventure game. A better way, is for the puzzles to BE the story, so that if you were to remove the puzzles you'd essentially be removing the story. Think of it as if the puzzles would be required to tell the story, even if it was adapted to another medium such as a movie or a book.


So here's a quick example of my process.
First, idea: Let's say it's something small and simple. A guy at an outdoor concert and he has a small challenge he has to overcome.
Second, story: Ok, how about he needs to use the toilet, but there's a huge queue. We're keeping the story simple, and that's pretty simple.
Third, expand: Alright, so there's a queue, so let's say there's three people in the queue, and he has to do something for each person.
Forth, expand more: First person is afraid of spiders so you need to get a spider, second person wants to be paid off so you need money, third person wants a signed autograph from the band so you need to get that.
Fifth, expand even more: You get the spider by finding some gloves in a trash can and pick up a black widow. You get the money by stealing a water pistol from a child and mugging someone. You get the signed autograph by getting a pen and writing on a piece of paper.
Sixth, expand even MORE: You find the gloves in the trash can by distracting the garbage man... ect, and so forth, and so on.

Simples. 

I hope this helps.

[heltenjon]

I have designed far more puzzles than I've been making actual games, I'm afraid... Very often, I would think like Danvzare describes and pretty much follow the same method. But not always.

Go backwards!
Sometimes I know where I am supposed to end up. Then I can make the story about which hurdles the player needed to pass before getting there. Trade it Forward was written pretty much this way. If successful, this may make a natural "flow" of doing something to get something that is needed in the next puzzle and so on.

This is what you have to work with!
Some action movies have been scripted like this...we have this location and these vehicles - what can we do with these toys? MAGS invites this way of thinking, and the Puzzle-Making Practice Competition, as well. Then it's about how to include everything in the description and try to see a jigsaw of a story that includes all those pieces. Use it All (Or Forget Your Ball) is a result of such an exercise. This method is fun, but the puzzles may end up being difficult or far-fetched. Like Danvzare said, signposting will be needed. (Or, like in the case of Halloween with an Unexpected Guest, I didn't think of puzzles when I posed my kids' toys in Halloween positions, and as a result, the game has next to none of them.)

What can go wrong?
This method is fairly common when writing stories. Let's say our character, Petunia, is going to take the bus to town. Now, what can go wrong? She could miss the bus. She could get on the wrong bus. She could have mixed up the time table and the bus doesn't come. She could get on the bus, but have no money, or have forgotten her bus card. And so on. All these situations come with their own set of problems, and in turn, solutions. Pick the one that sounds most fun, and off you go.

I have played this game before.
Remember when you had that great idea in a game, but it wasn't the correct solution? You weren't supposed to redirect the river to get water for the town's fountain, but had to perform the traditional rain dance instead? Well, how about when you write your game, you can redirect the river?

Yap it up.
Oh, the best way. Chat with your friends and put obstacles in the way for each other, and try to come up with solutions to the obstacles your friend proposed. If you have a half-assed puzzle that doesn't quite work, your friend may have just the right idea to make it tick.

[Danvzare]

I have designed far more puzzles than I've been making actual games, I'm afraid... Very often, I would think like Danvzare describes and pretty much follow the same method. But not always.

I'm going to be stealing some of these methods from now on. 

[CaptainD]

@RootBound - my process is pretty similar to yours really.

1 & 2 - Whatever puzzle is in the game should fit with the story somehow (and preferably, wherever possible, it should even add to the storyline or lore of the game universe in some way, rather than simply be an obstacle for the player to overcome, although it's not really possible to do this with every single puzzle (unless you have rather few).

The tone of the game is also very important when it comes to the puzzle solutions, too. When I designed "The Corruption Within", I decided from the outset that I wanted everything in it to be at least vaguely believable that it could happen and be solved that way in real life. (Much to the disappointment of those who thought the game was going to be a gothic horror, I guess.) With my Captain Disaster games, much zanier solutions are obviously possible (even desirable), albeit they still have to fit in with the internal logic of that game universe.

3 - Definitely agree, sometimes a puzzle is simpler than you originally intended for various reasons, and you feel the need to add another small puzzle for that location, or add an extra step or two for an existing puzzle. This is where tester feedback is absolutely vital, which I'll talk more about later.

4 - Even in highly polished commercial games I've seen this plenty of times - there's an obstacle of some sort, but it makes no real sense that it should be there. Occasionally, the game doesn't even attempt to make an explanation as to why this is the case. I try to think of the general setting obstacles in the initial design stage (albeit it is a very iterative process) because adding something later on is, I think, where these incongruities often arise. Something is there because the game designer needed to add a blocker, not because it actually makes a great deal of sense in terms of the overall storyline.

5 - To me, the real key to this is having good testers, and communicating with them effectively as to what did and didn't make sense to them (and allowing for the fact that different testers will obviously have different sets of things that make sense or not!). This is often difficult to do via written communication only, and sometimes I've not been able to understand the issue until they sent me a gameplay video or I've had a chance to actually chat to them. This happened with one of the puzzle sequences in CD3 - we had communicated several times, and I'd made some adjustments / added pointers etc, but until we did a Discord call and he talked me through it while actually showing me in the game itself, that I truly understood the issue.

Additional comments:

I've always liked the idea that a game should contain enough hinting to give the player enough information to solve the puzzles, as I dislike having to look up hints to progress (albeit I dislike being stuck in one part of a game even more), and I try to incorporate this in a multi-stage way, to allow for the puzzle to be solved by those who like their games on the tougher side, but also if someone is struggling, the game allows the player to receive some more pointers. I've tried to do this in a few places in CD3, so will be curious to see how people find it. (I'm fairly convinced there will be the normal groups of people saying it's too easy or too hard!) If in doubt, I would always tend to err on the side of making it easier, since I don't want players to become frustrated and give up on the game.

One final comment I have is mini-games - yes, I know these are pretty much love-them or hate-them parts of adventure games - I feel they're a nice addition as long as 1/ they fit the story and 2/ aren't so difficult or out of line with the rest of the game that they just hold up or annoy players. For these I have a pretty simple strategy - I prototype them until I'm happy with how they play, then get a couple of testers to look at them in isolation to make sure they're understandable and completable, then I add them to the game and put in final art / sound. This seems to have worked pretty well so far. I create them as a room in the project itself so that no importing is needed, and if I decide to get rid of the mini-game, all I need to do is delete the room (or just never have it be accessed in the game).

Anywhere, there's my tuppence worth!

[RootBound]

Thanks so much for all these thoughtful and detailed responses!   This is a really great discussion! 

@CaptainD I think the multi-layer hinting sounds really smart. In practice, though, that sounds really difficult to design without simply including a "hint system." Is this demonstrated in the newest Captain Disaster demo? Would love to see how you're implementing it.

Anyone else have thoughts about hints or hint systems and how to nudge struggling players without making it too easy for the pros? Do you have a favorite game that did this well?

I remember one of the later Myst games, 3 or 4, had a hint system, but it was very immersion-breaking, just a gui, not integrated into the game world. The first-tier hint was sometimes helpful but the second-tier hint was usually not, and the third-tier hint was just a walkthrough. It was too easy to just give up and use the hints whenever a puzzle got difficult. More ideal would be stopping short of a walkthrough and integrating the hints more into the game itself. Which is obviously a lot more work.

I think the way I've tried to do this in the past is to make the rules of the world as clear as possible and have the puzzle solutions deduce-able from those rules and from the tools available. I'm trying to be much more intentional about that in my current project, making sure the first puzzles teach the player the major rules and tools of the world. Later puzzles should then (in this method) apply these rules and tools with increasing complexity. All of this is of course MUCH easier said than done.

[Danvzare]

Thanks so much for all these thoughtful and detailed responses!   This is a really great discussion! 

@CaptainD I think the multi-layer hinting sounds really smart. In practice, though, that sounds really difficult to design without simply including a "hint system." Is this demonstrated in the newest Captain Disaster demo? Would love to see how you're implementing it.

Anyone else have thoughts about hints or hint systems and how to nudge struggling players without making it too easy for the pros? Do you have a favorite game that did this well?

I personally have three ways of doing hints.

• Firstly, if you try and combine two items that aren't supposed to be combined, you'll be given a hint as to what you should be combining those items with: "Nah, I'd rather cut up the pillow than draw on it."
• Secondly, if you look at the item, you'll get a hint as to what you should be doing with it: "Hmm, I've always wondered what was inside of a pillow."
• Thirdly, have characters basically be built-in hint systems, with their dialog options being literal hints: "Any idea of what I could do with this pillow?" "I don't know. Just don't destroy it like you did with your last one."

And of course, add more hints for more obscure or harder puzzles, and less hints for the easier and simpler puzzles. If it's a part of the game where the player has a choice to solve multiple puzzles (like the Three Pirate Trials), then you can usually get away with less hints, since if they get stuck, they can just go do something else until the hints they did find, finally clicks in their mind.
I like to think that the best part about this method, is that they won't be able to distinguish the hints from the red-herrings and meaningless blabber. So the hints are likely to just sit at the back of their mind, until they suddenly recall it, thinking that they've come to that conclusion on their own. (In other words, planting the idea of the solution in their head, so they think it's their solution and not yours.)

[heltenjon]

Anyone else have thoughts about hints or hint systems and how to nudge struggling players without making it too easy for the pros? Do you have a favorite game that did this well?

I liked the phone ringing with various hints in Strangeland. You could just not pick it up, but on the other hand, it was just as logical or illogical as the other things in the dream logic of the game.

In M*A*S*H: The Adventure Game I suggested for Viktor that he made Hawkeye a sort of hint central in the game, someone you could ask for help with all your other problems. I think he did a marvellous, in-character job of it that fit the game very well.

In BYO-Hero, there is a hint system that runs on time - if you have been stuck for a while, more and more light bulbs light up, allowing you a hint after trying for yourself for a time first.

Challenge of the Tentacle has a hint system, but will withdraw points from your score if you use it. This is also a nice balance between providing the help if necessary, but urging the player on to try alone first.

All of these work fine. I think my favourite way is to have an in-game character (or more of them) provide clues, as that doesn't break immersion.

[cat]

I like the concept where a character is the hint system. Tales by AprilSkies did this very well. The ZDF Neo Royal game had a character called "Captain Obvious" that was basically the in-game help. (Unrelated side note: Aeons ago, there was a MAGS topic about useless superpowers. I thought about making a game about Captain Obvious but couldn't think of any story or puzzles. I'm glad that ZDF made use of him later )
I generally like when you can talk to NPCs about your current progress and the dialog updates accordingly. Nothing is more boring and lazy than characters who's dialog doesn't change with game progress.

About designing the puzzles:
I usually come up with a setting first, and then see what fits in there. I'm not a person who can write epic stories anyway, so my stories are a result of the setting and puzzles.
It also helps to make a game together with someone else. You can bounce ideas and spot problems early on.
For Augustin, the setting made it clear that there has to be some music puzzle. I also did a bit of research about the era and so the other characters and puzzles fell into place (with the help of Creamy).
On the other hand, for one game I thought it would be fun to code a catapult and so Miez and I built a whole game around this minigame
I generally seem to have the "problem" that my puzzles are rather on the easy side. So I do not have the need for a hint system.

[CaptainD]

@CaptainD I think the multi-layer hinting sounds really smart. In practice, though, that sounds really difficult to design without simply including a "hint system." Is this demonstrated in the newest Captain Disaster demo? Would love to see how you're implementing it.

I don't think there's anything in the demo, although truthfully the puzzles in that are pretty easy anyway. The two specific examples I can think of relate to tougher puzzles that are standalone in nature - what I've done is make it so that if you try and fail once, the game points you towards asking for help, who will give you a hint. If you try again and fail, you can get a slightly more pointed hint, and then a third. The game checks that you have tried the puzzle (entered the room it's housed in, essentially) before it opens up the possibility of the next hint.

I haven't had much feedback on how well it works because the testers either worked out out themselves and / or didn't realise the hints were available - I have made it a little clearer in the dialogues and responses that this is the case now.

But it's pretty arduous to implement in-game and I think the easier option would be to have a consistent hint option in the game. One I remember quite enjoying (I'm not actively against hint systems in games, incidentally, I just try to avoid them in my own games) was Runaway 2 (or maybe 3), where you could phone a character to get a cryptic hint. That character was already built up in the series to be a bit of a crackpot so it worked well in-universe.

[edmundito]

I don't have much experience designing good puzzles, but as I was working on a small game recently, I researched how to approach it. Essentially, I think of two types of mysteries: Riddles and Puzzles.

- A riddle is something mysterious where the player does not know the solution at first, but works towards it.
- Puzzles are situations where the player knows the solution (e.g., open the door, jigsaw puzzle) and works towards finding a solution.

To offer a good puzzle, the player needs to understand the goal first, and then work their way towards it. Often, I notice that the design ends up being a riddle with the player not exactly sure what they are solving, which can make it difficult. You have to be deliberate on what it is, and the player needs to know if they're solving a riddle or a puzzle.

Here's some notes that I wrote based on The Art of Game Design: A Book of Lenses (Schell, 2008) chapter on puzzles (chapter 12):

- Make the goal easily understood: Right away the player needs to understand the end-result by looking at it
- Make it easy to get started: Player should be able to start moving pieces right away
- Give a sense of progress: Reveal changes slowly, instead of one big answer at the end. That would be a riddle instead of a puzzle.
- Give a sense of solvability: Make sure that it feels like it can be solved.
- Increase difficulty gradually: Make first solution an easy steps, then add more steps as it gets more complex.
- Parallelism lets the player rest: Give parallel challenges to take breaks in between, like crosswords
- Pyramid structure extends interest: A series of small puzzles that give a clue to a larger puzzle
- Hints extend interest: Well timed hints. What Ben (Chandler) told me once: give some kind of clue when the player gets stuck
- Give the answer! Give out the solution in some way. Show the final product.
- Perceptual shifts are a double-edged sword: "either you get it or don't is hard" - example. monkey wrench - changing perception to make it work.