Words: 3544 Approximate Reading Time: 25-35 minutes
I love puzzle games. I like puzzles in general, since a good deal of what I do for a living essentially revolves around solving puzzles. And puzzle games come in so many different varieties, often with so many interesting mechanics, that you can find who knows how many hours of enjoyment just solving puzzles.
There are a lot of amazing puzzle games on the market right now, but my object here isn’t to list the games I really like. Instead, I want to focus on a very particular problem: obtuse puzzles.
An obtuse puzzle is a puzzle that has a solution, but the solution is so convoluted that it feels almost impossible for the player to figure out the solution without the help of a guide, or else by brute forcing things.
An obtuse puzzle is not the same as a tough puzzle. Sometimes you can run into puzzles that are hard, but you understand that everything you need is there, but something just isn’t clicking. You might even end up using a guide because it’s too tough. But that doesn’t make the puzzle obtuse.
To help explain, let’s imagine you run into a puzzle that’s really hard. You struggle and struggle, trying a lot of different possible solutions, but nothing seems to work. Eventually you get frustrated and consult a guide. In this hypothetical scenario, you look at the solution and you have one of two reactions:
- “Oh that’s the solution! Of course, it all makes sense now!”
- “Wait, that’s the solution? How was I supposed to figure that out?”
If you have reaction #1, then you’re likely just encountering a tough puzzle. But if it’s reaction #2, then it’s very likely an obtuse puzzle.
It’s not a perfect system. If you’re not used to solving puzzles in general, you might be more inclined to reaction #2 even if the solution makes sense. But as you get more used to the process, the system gets better and better at highlighting good versus bad puzzle design.
Why are obtuse puzzles bad design? Ultimately, a puzzle should feel solvable. There’s a very important distinction to be made between a puzzle being solvable and a puzzle feeling solvable. You’re unlikely to come across puzzles in games that are quite literally not solvable. But just because it’s possible to solve a puzzle does not mean that solution makes sense. And “making sense” is the key to making a puzzle feel solvable.
If you’ve played puzzle games before, you might have run into these obtuse puzzles. In a sense, it can be tough to design a puzzle game without at least one obtuse puzzle. Not because the designer is trying to make things obtuse. But there’s a fine line between tough and obtuse puzzles, and it’s easier to miss the mark than you may think. Why? Because a common mistake is substituting the psychological for the logical. By which I mean that things can make sense in your head, but may not make as much sense from an “objective” standpoint. This trap is easy to fall into, especially as things get more and more complex.
So in this essay I will try to explain a few problems that can lead to obtuse puzzles. Think of these as potential traps. It would be easy to say “avoid these traps,” but some of the things I’m going to point out can be part of making an interesting puzzle game. So instead, I’ll focus on explaining how they can be dangerous. I’ll provide some examples to help illustrate my points throughout.
Puzzle games start with a set of mechanics and rules. When you do Action X, you get Y result. You need to perform so many actions in such-and-such an order to reach the solution. The particular actions, the order, and any other elements can be varied in an infinite number of ways, but at the end of the day, that’s what a puzzle looks like.
Of course, since there are a lot of puzzles and puzzle games on the market, the number of unique and interesting ways to pull off puzzle design dwindles over time. There are definitely a lot of possibilities, and sometimes puzzle games package old designs in new presentations that can be fun and interesting. So we’re not exactly going to be starved for new good puzzle games.
But a problem can creep in when trying to design the mechanics, as it’s possible to try to make the mechanics so complex that they don’t feel “normal.” That is, you’re not used to the mechanics, so when a puzzle requires using a particular mechanic you’re unfamiliar with – or even worse, haven’t really been introduced to – the result is that you feel lost.
To give an example, I’ll point to one of my favorite games, La-Mulana. La-Mulana is an adventure Metroidvania with a heavy emphasis on puzzle solving, kind of like playing as Indiana Jones exploring a temple. The game is big and tough to get through for several reasons, and while a lot of the puzzles are good and I love the game as a whole, there are definitely a handful of genuinely obtuse puzzles.
I’ll pick out one example. In one area around the middle of the game, you run into a confusing level that has teleporters that take you to various different parts of the level. However, as you go through, you realize that the teleporters all loop back on themselves: you’re basically stuck.
There is a slight hint, if you’re paying attention: as you go through, you run across some pots you can break, and one of them always drops an item. Since pots usually only drop items once, this should be odd. But what are you supposed to do?
It turns out you’re supposed to stand on that spot and press down to warp into the floor, not unlike a pipe from a Mario game.
In fact, there is a small picture on the side of the screen in a completely different area that shows what essentially looks like Mario going down a pipe (the same puzzle mechanic is used in that room, but since the game is non-linear, you don’t necessarily run into them in that order).
Here’s the extra kicker: that mechanic is used a grand total of three times throughout the entire game, and that includes one use in a bonus area. So if you were to play only the main game – a process that can take tens of hours depending on a variety of factors – you’d still only have to do this twice in your entire playthrough.
But how are you supposed to connect the dots between the weird pot and pressing down? What if you’re not used to paying attention to the little markings on the walls, since there are only a small handful of cases where they provide important information? It’s these kinds of questions that make the puzzle obtuse.
You can see the underlying logic. The area itself is supposed to be confusing, so a solution that’s out of the ordinary would definitely be called for. And the game as a whole requires you to pay attention to little details, so it might make sense to require the player to look at everything for clues.
But there’s a limit to this logic. It makes sense if you’re making the puzzle, and you can even go backwards and trace out the logic if you put in the effort. But if you’re in the middle of playing the game, especially if you’ve either A) not run into the solution yet, or B) not encountered the solution for a while, then the player is unlikely to connect the dots.
And so that’s why it ends up being an obtuse puzzle. The mechanic is used so insanely infrequently, and is introduced so late in such an obscure way, that there are so many ways for the player to not even be able to figure out what they’re supposed to do. It’s an almost perfect setup for the “How was I supposed to know that?” reaction.
So mechanics should be kept simple as much as they can. You want the player to know and understand the rules, and the puzzle is about working with those rules to reach a solution. As opposed to the puzzle being about figuring out what the rules are. Those latter kinds of puzzles need to be reserved for introductory concepts, where the game is teaching the player how the mechanic works. But if there’s no introduction, the player is left completely ignorant not only about what they’re supposed to do, but even about how to figure out what they’re supposed to do.
Broadly speaking, puzzle games fall into one of two camps. They can be split into isolated levels that are essentially self-contained puzzles. Or they can throw the player into a large world that is supposed to be explored, where puzzles are scattered around the world, and sometimes the solution to a puzzle in one part of the world requires finding a clue in another part of the world.
A classic example of the latter game is Myst, which initially puts the player on a strange island. The island is essentially sectioned into a somewhat isolated “zones” (for lack of a better term) that have some kind of puzzle that needs to be solved to go into another world, which then has its own puzzle elements. The structures of the puzzles are often fairly complex, and some of the puzzles, particularly in the hub island, require solving some initial steps to figure out what to do next.
More recent entries in this genre, often spiritual successors to Myst, include Quern and Obduction. La-Mulana, which I mentioned above, operates as an exploratory game as well, being modeled more as an adventure-puzzle game (though importantly, Myst and its successors are solely puzzle-oriented, and thus don’t feature any combat or real danger of any kind).
The fun of the exploratory games is that they feel somewhat organic. Isolated puzzles can be fun, but they usually have to exist merely as puzzles (there are some games that break this mold, an example that comes to mind being Stephen’s Sausage Roll, but mostly there’s no overarching connection between the levels). Exploratory games allow the designer to tell a story and put the player into a world (often one that is uninhabited) where the puzzles are there for a reason. Games don’t have to do this, but they can add a nice layer to the fun.
But this same exploration can cause problems. Namely, the puzzles put into these worlds essentially rely on the player wandering around and locating clues, and then putting those clues together to help determine a solution. It can be a lot of fun, but there’s a very fine line between “you need to just explore a bit more to find the clue” and “where the hell am I supposed to find the next clue?”
In other words, exploratory puzzle games can end up leaving you completely stranded, without any indication about what you’re supposed to do next. And the problem is that you don’t even know what to do in order to locate the next clue to help you out. So you not only feel lost, but as you keep running around the world you get more and more frustrated trying to find something.
A secondary problem is that as information gets split up over time, it can be harder and harder for players to keep the necessary clues straight in their head. Especially if information is “permanent,” that is, it doesn’t disappear when it is “used up” (or there is not a clear indication about what information is relevant to a particular puzzle). So you can run around, collect a bunch of little pieces of information, and then run into a puzzle and be unsure what clue is supposed to help you solve it. And as a consequence, you might be stuck trying out a variety of possible solutions based on a particular incorrect clue, because you aren’t sure whether your solutions are failing because you’re going about everything in completely the wrong way, or you’re on the right track and just need to do things in just the right way.
I ended up running into each problem with Quern and Obduction. Obduction puts you into a rather large open area that felt somewhat dull to walk around, and figuring out what to actually do became so exasperating that I couldn’t even keep playing. There was so little information available that things just felt needlessly spaced apart.
Comparatively, Quern did much better by making its world much smaller in scope, but it ran into the second problem by having poor feedback about how puzzles were supposed to work. Usually this isn’t a problem at the start, as it’s fairly obvious what to do for the first handful of puzzles, and then things get more complicated from there. But as things get more complicated and you start collecting more potential clues, it becomes harder and harder to know what you’re trying to find. That is, what is the next puzzle that I’m supposed to be solving to move forward?
An example of a game that does exploration and puzzle-solving well is The Witness. The game is fairly open-ended in its exploration, but does a few things right in this respect. Firstly, the mechanics are very simple. Rather than a bunch of different mechanics that operate kind of like a point-and-click adventure game (i.e. open inventory, locate item, attempt to use item on object until it interacts with something), The Witness has a single mechanic that is iterated on in a variety of ways. The larger hub is split into separate zones, each with their own particular version of the basic puzzle. And then on top of that the world becomes a sort of puzzle on its own, by teaching the player to look for particular patterns in the environment. There are a few obtuse puzzles in the game, but in a game that includes over 600 individual puzzles – and when those obtuse puzzles aren’t necessary to actually get through the game – they don’t stick out nearly as badly.
So exploratory games shouldn’t be avoided at all. They’re actually some of the best puzzle games out there. But in making exploratory games, I think a good rule of thumb is to make the explored area compact. You want the player to be spending their time solving puzzles, not running around. At least, that’s true if the aim is to make a puzzle game. The bigger the area, the more (pointless) running around there is, the less of a puzzle game you end up with.
Another important rule is to think carefully about how the player is likely to explore the world and associate particular pieces of information. Obviously, some part of this process is done through testing. But testing on its own doesn’t necessarily eliminate obtuse puzzles. So as an extra precaution, try to step back and trace the steps backward. The solution is to do “A.” In order to figure that out, you’d have to read note “B,” and the player would also need to find item “C.” So when the player reads the note, are they going to see in their heads what you see in your head? If you want to make the note obscure, is it so obscure that it doesn’t necessarily make sense? Because while obscurity can be useful to making a puzzle tough, it can also lead to the puzzle just not making sense. Hence the importance of thinking about it as a logical, rather than psychological, exercise.
Puzzle games start out fairly simple, and then get progressively harder as the player moves through the game. The reason for this progression is obvious: you want to begin by introducing players to basic concepts, and then slowly start to build on those concepts in more complex forms. A game that was incredibly complex from the start would turn off players for being inaccessible. A game that was too easy would be dull. Certainly nothing controversial here.
As the game continues, there is a need for harder and harder puzzles, but as difficulty increases the odds of obtuse puzzles are going to rise. Because remember that there can often be a fine line between a puzzle being tough and being obtuse.
Games that rely on open exploration tend to run into the problem of making its more difficult puzzles obtuse. Mostly this is due to the factors I mentioned in the previous section. But even games that split their puzzles into distinct levels can run into this problems in a variety of ways. Sometimes it can occur by trying to introduce new mechanics (see the first section on complexity), or sometimes by trying to use existing mechanics in ways that a player might not realize is possible.
And sometimes it can happen by just doing things that make a puzzle really tough, without any good reason.
An example that sticks out in my mind occurs in The Talos Principle. Strictly speaking, The Talos Principle kind of merges the idea of discrete levels and open exploration. While the individual puzzles are sectioned off, they take place within hub worlds that often incorporate the mechanics from the levels. Occasionally this can require you to figure out how to take puzzle elements out of one level in order to make use of them in the hub areas, to then solve additional puzzles.
Usually these puzzles are supposed to be the hardest, because they then unlock additional secret areas with even tougher puzzles. So obviously, the puzzles to unlock the extra hard puzzles should be hard themselves, right? It makes sense.
The problem is that some of them are made hard through rather obtuse methods. The example I am thinking of involves a particular hub puzzle. It requires bouncing a laser off of various crystal stands that you have to position in the proper way to reach a particular spot.
One of those stands was hidden in a tree, meaning locating it required you to pay close attention to things you weren’t normally used to paying attention to. (I’ll note that a patch ended up relocating the stand to the top of a tower that made it more visible.) Nor had the game necessarily indicated that you should look into the trees, and it was fairly well hidden at that. So if you decided not to resort to a guide, you could potentially end up running around for hours (unless you just gave up entirely) trying to locate a core component that was literally hidden from you.
We occasionally talk about things being “artificially difficult,” and while in one sense that term has no meaning (since by definition the game is artificial, being created), this is a perfect example of what that term is supposed to mean. Something that is made more difficult than it really needs to be, or where the difficulty is not the result of the mere placement of puzzle elements, but of particular elements being obscured from the player.
So a good rule of thumb is that when a puzzle is supposed to be difficult, it should be asked what is making the puzzle difficult. A good puzzle should essentially provide a player with all of the information they need to solve it, and allow them to figure out how to arrange the pieces given to arrive at the solution. If an integral piece of information is missing, the player should be led to understand that something is missing. But if the difficult relies on something that is outside the player’s “reasonable” comprehension, then the puzzle isn’t really difficult, so much as it’s tricking the player. And tricks aren’t puzzling.
I’ve only described a few basic elements of obtuse puzzle design, and more importantly, the factors that can contribute to particular puzzles being obtuse. I think it’s almost impossible to make a relatively long puzzle game that doesn’t have at least one obtuse puzzle. Shorter games will probably be able to get around the problem, but if the objective is to make a bigger puzzle game (or to churn out multiple games based on the same basic premise), then the likelihood of making an obtuse puzzle is going to increase over time.
So the key here isn’t so much to worry about whether a game has any obtuse puzzles. Obviously, the ideal number of obtuse puzzles in a game is zero. But failing that, it’s important to avoid having too many, having them be too obtuse, and having them be too important to progression.
The idea is that these three factors contribute to the obtuse puzzles sticking out in the minds of players. They can ultimately detract from the overall quality of the experience, because the player is left focusing not on the positive feeling of solving the good puzzles, but on the negative feelings of running into these obtuse puzzles. So strangely enough, finding these obtuse puzzles and getting rid of or fixing them can potentially end up being more important than making additional puzzles.