The Perils of Puzzle Design

Words: 2667 Approximate Reading Time: 20-25 minutes

I’ve mentioned in multiple other essays that I enjoy puzzle games. That feeling of discovery – the “aha!” moment – grants a certain sense of accomplishment that be tough to match. But enjoying puzzle games does not mean I enjoy puzzles of all types. Indeed, many puzzle games have created a sense of annoyance rather than enjoyment.

And the problem here is that there is a language to puzzle design – particularly within video games – that needs to be properly understood and utilized. If an element gets miscommunicated, or not communicated at all, then the entire system can fall apart.

Recently I played a game called The Painscreek Killings which provided some excellent examples of this problem. Painscreek is a mystery game in which you explore a small abandoned town. You are an investigative reporter sent to explore a series of killings that took place years prior, because the town is about to be sold off and demolished. As you explore the town you read journals from the town’s previous inhabitants to slowly piece together what happened.

The game’s systems evokes the design of Myst and similar puzzle-adventure games. Which means it involves a good deal of running around to find key items – often literally keys – to then use in appropriate locks. There are, though, genuine puzzles. Usually some variant of a drawer or safe needing some numerical code to be entered, which you then have to either find written down or in a few cases gather clues to deduce what the code must be. I’ve written before on how “hide-and-seek” as a feature is a common design decision due to particular limitations about how people engage with these games, and Painscreek succumbs to that same issue. But I want to focus on the actual puzzles that the game has to offer. Because those puzzles are what ultimately helps the game stick out.

And the examples that I want to look at highlight the problem of video game communication. That the language of design – both the general language that has been built up among video games in general and the specific language of a given game – needs to be something consistent and something that the player will understand. When the design deviates from that language, the game struggles to communicate information. And in a puzzle game, where information is particularly important, this creates an especially big problem.

Especially so because one of the first things that the game begins with is a note that you as the player may need to make logical deductions from incomplete information. You will have to solve puzzles. By putting this note at the front of the game, the game developer is telling you how to play the game. And this initial note does not match up with how the game actually works.

Explaining these issues requires going into some particular examples, and I think it best to try and explain with as much detail as I can. Which in turn means that there are some puzzle spoilers for Painscreek.

Playing Darts

Perhaps the first puzzle I ran into that was intensely frustrating involved darts.

Specifically you are exploring one person’s house, and you are opening drawers in his desk. One of the desk drawers is locked with a combination lock requiring a four digit code. In one of the other drawers you find a scrap of paper that contains a reminder about the code: “Subtract the yellow darts’ score from the red darts’ total.”

That’s a bit convoluted, but it gives us something to look for.

If you look further around the room, you might see a red dart lying on the floor next to the door. If you then close the door, you’ll see a dartboard on the back, with darts in it. The red darts are located on the inner bullseye and the triple-20 spot, while the yellow darts are located in the single-9 and single-2.

So your first reaction might be to do the simple thing: add up the points for the two red darts, then subtract the points from the yellow darts. Which yields 101. Since it’s a four digit code, though, we’re lacking a digit, but we can just put a 0 at the front or back. So either 0101 or 1010. Neither code will work, though.

Hang on…the reason we noticed this in the first place was that there was a red dart on the floor. Why would it be on the floor? Because it fell out! So we need to figure out where that third dart was. And if we look closely at the board, we see a conspicuous black pixel in the single-6 area. So do the math aga- yeah this doesn’t work either.

You might alternatively notice those same black spots on a couple other points (which are a bit harder to spot because they’re in areas that are already black), so maybe you try those. To no avail.

But wait, there’s one other bit of information that we should remember. When you start the game, the very first place you’re directed to go is the town’s inn. And the inn also contains a dartboard. Next to the board is a set of instructions for how to play darts. The first two-thirds of the instructions explain the scoring parts (telling you about singles, doubles, and triples). The last third then explains the actual rules: you don’t start at 0 and count up, you start at 501 and count down.

So that’s it! You start at 501, subtract the score from the red darts, and then subtract the total from the yellow darts. This puzzle thus really asks you to remember different clues that you’ve seen at other points and put the information together to- I’m just kidding, none of the possible combinations you use this with will work.

So then what’s the secret? As you explore you’ll later find a photograph of the dartboard which shows that the third dart was originally located in the double-18. So if you take the score of the red darts – counting up from 0 – and then subtract the yellow darts, you get 137. Lo and behold, 0137 works.

Why is this all so frustrating?

Firstly, there’s no real puzzle or deduction involved. There’s some basic math, but that’s all. There are a number of opportunities presented for the player to actually solve something, but nothing is done with those opportunities in favor of just locating the piece of information that tells you what to do. You simply go from spot to spot until you have the key – literal or metaphorical – that unlocks the next problem.

Secondly, the game puts in information that is not only unhelpful, but actively distracting. The rules sheet for how to play darts is something you not only can clearly read, but like many other things you can grab to zoom in on and read more clearly. Some portion of this rules page is helpful – you definitely want players to know how to read darts scores. But the last bit tells you how to calculate scores in a way that leads players down a false trail. This information then implies that the information is important and should be remembered: you don’t start at 0 and count up, you start at 501 and count down. It may seem small and innocuous, but since the point of making a game like this is to get players to pay attention to details placed in the game world to solve problems, the information that you put in is going to be assumed to be relevant to at least one problem.

Playing Chess

In another home is yet another desk with a locked drawer, also requiring a four-digit number combination. So as with the above, we need to locate the numbers and the order to put them in.

As you explore the house and other parts of the game, you’ll find these chess pieces with numbers written on the back. You’ll find six total – a king, a queen, a pawn, a rook, a knight, and a bishop. Of course, since there’s six pieces and six numbers but only four digits on our code, it means either that those numbers are for a different lock, or that we need to know which of those pieces are irrelevant.

As you continue progressing through the game you’ll ultimately open up a different safe in the same house which contains a diary, and in that diary is a clue about the locked desk: K = King/Knight, B = Bishop, Q = Queen. This arguably is pretty easy to grasp, but is unclear. What are these K, B, and Q that the note is referring to? There’s a fair chance you’ve not seen anything like that.

But it turns out that if you go back to the lock on the drawer, crouch down, and then zoom in on it, you can see the letters KQBK etched above the numbers, which gives you the order for the code. Since K means both king and knight, there are technically two possible codes since we don’t know exactly what the order is supposed to be, but at least if we get it wrong the first time we can just reverse the first and last number and get the code.

But this puzzle ends up being poorly designed because it hides a piece of information in a way that the player hasn’t been taught to look for. While the player is given instructions on how to crouch early in the game, it’s mostly not necessary. Just about everything you need to interact with (picking up items, activating locks, opening doors) can be done while standing. And no other lock in the game contains this kind of clue where you need to look at the lock itself to glean information. The clue is entirely unique in its placement, and in a way that the player hasn’t been given any reason to look for it.

Just as the information you give to players needs to be considered, the way in which that information is provided also needs to be carefully chosen. Any given puzzle element should be either introduced in a way that makes sense and signals to the player how to play effectively, or else should build upon design choices that came before. Unique elements may come out of the combinations of rules, but presenting the player with something entirely unique is a recipe for disaster. Because the player has no reason to think about the intended solution.

To put it a different way, one way in which we can think about puzzle design is starting from the conclusion and working backwards. We have a combination lock, so we need a code. We then need to hide the code somewhere for the player to find it. But when reviewing the puzzle, we need to ask how each step follows logically from the one preceding it. How will the player know how to look for the code? When they find it, how will they know what it belongs to? When they know what it belongs to, how will they know how to use it?

Even if some parts of the equation are communicated well or obvious, it’s important for every part to be something that is well within the player’s grasp. Which means either something the player can easily locate or deduce from the information provided. If a necessary clue or object is hidden and the player is simply expected to keep looking until they find it (or give up), then the puzzle has failed as a puzzle.

Information, Communication, and Design

One of the most common complaints about these kinds of games is what is referred to as “adventure game logic.” In its most egregious form, adventure game logic involves stumbling through and combining objects or bits of information to solve problems. Especially so if the problem could be more easily solved in some other way.

But the root of adventure game logic lies in poor design. And the failure of design is one of information and communication. It is where the developer has created a problem with a solution that makes sense only to the developer. Where the developer has seen the invisible connections between bits of information.

As I mentioned above, one thing you need to do in reviewing a puzzle is ask “how is a player supposed to know how to perform each step?” One of the hardest parts of reviewing something that you created is that you already know the solution, and so the answer seems obvious. Of course you combine this rope with this inflatable kid’s toy and this magnet to create a hook to grab a key. It makes sense…to you. But when viewed from the outside, it’s absurd. Absurd beyond even the necessary absurdity of the games themselves.

Games in general and each game in particular adopts a language for how it communicates to the player, and understanding both the general language of design and the specific language of your own game is necessary. Because that is how the designer and the player enter into a dialogue, and most importantly how the player knows what to do and how to play effectively. If the language is upset in some way, then the player is left without any guidance.

Sometimes the language is part of a genre. Puzzle/adventure games have implicit rules about what information matters and what doesn’t. If something can be read, there’s a fair chance that the information on it will be relevant to something you need to do. So you need to remember that information. If the information is not only irrelevant but distracting, then the player’s trust has been violated twice over.

If you want a player to pay attention to details, then the player needs to be rewarded for paying attention to details and the details need to in turn be relevant. If you want the player to explore, you need to reward the player for exploring and make the process of exploring fun. And so on and so on. A game communicates what it wants the player to do, but the player still needs a reason to do it.

Small mishaps happen here and there. Sometimes a game can ask so much of its player that the language gets jumbled up and confused, and a player may miss out on some things because they had too much to keep in mind. This may be a good reason for the game’s design to help the player compartmentalize information, or practice interacting with different systems, or train on different kinds of challenges. But the game needs to communicate something clearly to the player. It needs to make sure the player is comfortable striking out on their own.

Concluding Remarks

I could probably spend hours talking about The Painscreek Killings and how it was such a…well…unfortunately not great game. There is a great deal of potential there, but that potential feels wasted in several ways. I distinctly recall enjoying the game for the first two-thirds, even when the actual gameplay was running around and collecting keys to use on locks. As simplistic as it was, it still held all of this promise.

But that promise gave way to poor communication. The puzzles I covered in this essay really only scratch the surface of the game’s problems. And in a sense, it’s tough to really grasp those problems without dealing with them directly. Text can only communicate so much.

This game, though, stood out as a wonderful example of “what not to do.” The bad puzzles in here felt like the clearest examples of the problems of design that not only plague similar games in the genre, but also with puzzle design more generally. These are not problems of laziness. These are problems of language. These are problems that exist because true communication demands that we be able to see the world through someone else’s eyes.