Tuesday 30 July 2019

An Existential Statement of the Form ∃x φ(x) such that φ(t) is True

This post is a bit unusual as I will talk about my experience with The Witness, a video game released in 2016. I'm writing about this game because it is a puzzle game which does not give explicit, verbal, explanations about its own rules and as such I feel its existence is relevant to a view of art in general.

The Witness is a game that has a very large scope. Starting from a simple idea of a puzzle where a single line runs from a beginning to an endpoint, the full game contains an entire three-dimensional island to explore with various themes and settings.
The core gameplay of The Witness, as it's presented to the player, consists of solving line puzzles. The player has to draw a line from a circle to a rounded-off end of a line.

Throughout the game there are multiple symbols added to the puzzles that define ever-more complex rules for the route the line can take between these two points. These rules are never explicitly explained to the player, however, and the main hook of the game consists of the player figuring out what the rules are that govern these puzzles and their symbols.
The clearest example of this is the swamp area. Before you enter the swamp, you have traversed through at least two other areas. Then when you enter the swamp you see this set of puzzle panels:


These panels, starting on the right, gradually teach you how to define a shape within the panel with a line. It starts out extremely simple and by the sixth panel you have a decent idea on how to define a shape. When finishing this puzzle, you look to the left and you see the following:

In this view going straight seems like the most logical step. The walkway leads to an open area and the yellow path is brightly marked. This yellow marked path actually contains a puzzle that you haven't been shown the rules to yet. By this point you've nevertheless realised that part of the game is figuring out these rules, so you fumble until you get it to work and bring you further along, only to encounter two other groups of puzzles ten seconds after:

These puzzles require definite knowledge about how you're allowed and able to combine the various shapes within a single line segment, knowledge these first six introductory panels haven't touched upon. So you get frustrated and you move backwards to where you came from. At the earlier junction where you went straight, you now turn right and you find these:

A total of eight panels that very explicitly serve as a reference on how you can combine the shapes by a single line. Panels four to six even share a single solution to make it very clear what matters and what does not matter when defining these shapes. After you have found these, the rest of the puzzles in the area are quite straightforward, increasing the difficulty largely by introducing a higher density of shapes and adding a few smaller rules.

Acquiring this kind of knowledge is how most players will spend a large part of the game. Knowledge is introduced and then tested in a number of puzzles, which tend to be somewhat trivial once you have formulated the rules right. After you solve all the puzzles in each area, you activate a laser that points to the top of a mountain, clearly indicating that this mountain is somehow a point of interest.

Once you have gone through a number of areas like this, you have become acquainted with the game's mechanics and you enter into its last areas: the town and the mountain. In these areas it is no longer about gathering knowledge about the mechanics of the various puzzles, but about applying this knowledge and understanding it more deeply to solve a wide range of complex and creative puzzles, without ever being able to do anything other than moving around and drawing lines.

This latter part of the game, once you have seen most of the island, was for me by far the most enjoyable part and I found it ended rather quickly once I got there.

So far, so good, you might think. Designer Jonathan Blow has made a silent game about solving line puzzles on an island. Fun to explore if you're into that sort of thing and if you don't like drawing mazes then maybe this game isn't for you.
The story doesn't quite end there though and in order to delve more deeper into other aspects I first would like to point out some flaws I find with the part of the game I have just described.
As you can see from the description above, the first part of the game is more or less about learning, while the second part is about applying the knowledge gained in the first part.

If the first part of the game is about learning the ruleset of the island, it is important that this introduction is done clearly and, more importantly, without contradictory information. For the most part, Blow and his team succeed in doing so and have demonstrated well that it is perfectly possible for a game to let players figure out what they need to do without actually telling them. The earlier example of the swamp puzzles show that is very possible to provide a clear introduction to the rules for a puzzle using non-verbal communication.
Even the existence of environmental puzzles, considered one of the game's biggest secrets, is at one point communicated very, very, clearly. One of the areas closest to the beginning also has you use elements from the environment to find its solutions, which should be a huge clue that you should pay attention to your surroundings.

Before I delve further into those, however, I would first like to bring back the focus on what has frustrated me the most while playing The Witness. This is the fact that it at times deliberately obfuscates the kind of information that is necessary to understand the rules to the game. Blow has stated in interviews that this kind of obtuseness is something he has tried to avoid and while I believe he has genuinely tried to do so, I can't say that he has always succeeded, at least in the way I experienced this as a player.
For example, let's take a look at the introduction of a symbol with one of the more complex rulesets. The first panel and a solution look like this:

Which is fairly straightforward, even if you have no idea what's going on. The puzzle only has twelve possible routes and three of those are correct solutions. Even when merely fumbling around at random, I still have a one-in-four chance of finding a valid outcome. Like in the introductory panels of the swamp, this is more than alright, as any grounded deductions about the functioning of a symbol can't reasonably be made from any single panel. Which is exactly why the following panel is so problematic:

This is the next panel you encounter in a set of five introductory puzzles. The other panels are only activated once the previous panels have been solved, so there is no other way to continue except solving this panel. Because we only have the knowledge gained from the previous panel, the player can't reasonably be expected to have a well-formed hypothesis of the function of the symbol yet. At the same time, making a lucky guess isn't realistic either. Where the previous puzzle had twelve possible routes, this one has 184 possible routes. Only six of those routes are valid solutions and all of those solutions are longer than the average route. So while the number of solutions has doubled from the previous puzzle, the number of possibilities has increased fifteen-fold. You are therefore asked to extract an enormous amount of knowledge from a tiny amount of information. While in retrospect it is technically possible to come to the right conclusion given what you already know about the puzzle, I don't believe it is an effective, nor refined, method for teaching a new set of rules. From my own experience I can only think that these particular panels were designed in this way as to provide an artificial difficulty level that is perceived to be necessary when solving any kind of 'puzzle'. In my own playthrough, I got stuck trying to figure out this rule for about thirty minutes, before obtaining a hint for it outside the game. This quickly made me understand what the rule was and I consequently solved all other puzzles in that particular area within the next half hour.
The difficulty therefore doesn't lie in comprehending the system, but in simply knowing its rules. If you know the rules, it is easy, if you don't, it's extremely difficult.

The Witness is described as a game where you solve puzzles and I believe many of its problems stem from this description. In my opinion, the game mostly presents its panels as logic puzzles. In a logic puzzle you are given a set of rules and you have to understand how they interact with each other to see what their possible outcomes are. Figuring out the rules themselves isn't a challenging puzzle, or at least it shouldn't be. Shoots and Ladders, for example, does not automatically become a puzzle game by simply refusing to explain the rules.
Where Blow has gone astray in some parts of the game is that sections that should have been there to non-verbally explain the rules of a certain puzzle and have instead been treated as if they were puzzles themselves.
This has made some parts unnecessarily difficult, while at the same time partly undermining Blow's own vision of creating a game that allows a player to grasp the concepts of said game by simply observing well and 'being treated as an intelligent person'.

There are also other parts of the game where things are made more difficult than they need to be in a more arbitrary matter. After finishing the puzzles in the jungle, for example, the laser you have to activate is located within a small maze. This is already somewhat annoying on its own account, but it is made worse by the fact that the actual entrance inside the maze to the room that contains the laser looks like this:

It's right there in the center and I couldn't see it either. There is a little open area with a distinctive tree next to the entrance to help recognise its significance, but this tree simultaneously blocks the clearest view of the entrance. In the end I just used the knowledge that every maze has a solution by simply following a wall to its inevitable conclusion, which is tedious at the best of times. I can understand the importance of mazes for this game and have no problem with the other instances, but this particular one was a needlessly burdensome implementation for what in every other area is a fairly straightforward path.

Another component of the game that has been somewhat of a double-edged sword for me are the aforementioned puzzles which incorporate the environment, most notably the sunshine. I wholly applaud their inclusion and I think it's one of the main reasons The Witness feels like every part of it is somehow connected to all the other parts. At the same time I can't help but feel that the gameplay implications are occasionally extremely frustrating for the player. The vast majority of these puzzles require the player to stand precisely in a specific spot to either find the solution or see the puzzle at all. Even when you know where to stand, it is still sometimes awkward to get it right, with many tiny button presses required to obtain the correct viewpoint. It is thus once more the case that if you posses a certain kind of knowledge, the puzzle is hardly a puzzle at all, but if you don't have this knowledge the solution is literally out of sight.
At times these puzzles are made even more difficult by introducing a changing element on a timer. In that case moving around haphazardly to find the right viewpoint and at the same time trying to keep an eye on where you are going is cumbersome and adds little to your understanding of the puzzle's mechanics. I believe that as soon as you understand how a puzzle ought to be solved, executing that solution should be trivial. Unfortunately this isn't always the case in The Witness. Blow's previous game, Braid, also had these kind of moments where once you understood the solution to a level, there was still some finicky platforming maneuvers necessary to execute that solution.
It must also be mentioned that I had to play the game on a low graphic quality setting. This admittedly kept the game playable, but it certainly didn't help with of some of these puzzles that are so dependent on the way things appear. Especially in puzzles that had to do with shadows it was at times unclear whether I had the wrong solution or if I had simply mistaken one pixelated line for another.

Then there are also some decisions made during the design process that I can't really understand or account for. In some puzzles, I would estimate about one tenth of them, the puzzle panel turns black after you input a wrong solution. Then you have to go back to the previous panel and redraw the solution for that panel to light up the other panel again. This already tedious endeavor is made worse by the fact that the solution to the previous panel is still highlighted there. You just need to mindlessly trace a line that is clearly visible before you can go back to the panel you want to be working on. 

I have yet to find a good reason for those particular panels shutting off after inputting a wrong solution. As I said before consistency is very important if you want to teach somebody about anything, so having some panels shut off while others do not without any particular reason can't really be justified in my opinion. Perhaps some playtesting would have shown the errors of my ways, but even their necessity at parts of the endgame isn't enough for them to be included earlier in a more forgiving setting.
The wrong turn taken in the swamp described at the beginning is also an example of this kind of problem. It's understandable that having all the panels in a single line is monotonous and therefore unwanted, but simply luring the player away from something that is required to progress further isn't a very elegant design solution either. There surely must have been better and more consistent solutions for these kind of issues.

As a whole, I would argue that one of the more general problems in the Witness are those kind of diminutive lapses in consistency.
The interface of the game is kept very simple, for example. A great effort has been made to make sure that the only interaction the player has with the world is done so by drawing a line from point a to point b. It is one of my favourite features of the game and is wonderfully implemented, both mechanically, visually and conceptually.

That being said, there is one common item in the game that doesn't require you to draw a line. These are small voice recorders, found at various places all around the island. You have to click on these and then they play a soundbite. Why an exception has been made for these voice recorders is never made clear. There are other ways to play media on the island, including audio, and all of those use a line-drawing interface. The only reason I can think of for housing these soundbites in tiny, clickable, voice recorders is that in that way they are more difficult to find, or perhaps easier to hide. I'm not sure if that alone is worth losing a certain sense of cohesion in a game that otherwise goes through extreme lengths not to use buttons or any other forms of interaction.

Elsewhere in the puzzles there is at times a similar inconsistency. This doesn't hurt the puzzles in isolation but are detrimental to an overall view of the puzzles, and therefore the game, as a cohesive whole. One of the earliest puzzles you will encounter shows that its possible to have two different outcomes from a single line.

Yet in most of the puzzles found in the rest of the game, even those with multiple beginning- and endpoints, there is only one valid beginning and ending for any single line. There are some important puzzles much later in the game where knowledge of this kind of possibility is required to progress at all, but I think that showing these kind of multiple solutions as one of the few distinct possibilities at the beginning of the game is similar to the 'I before E, except after C' rule. Even if it's true on some occasions, the rule has just as many exceptions, so perhaps its better not to teach it at all.
In another instance the rules are more complicated than they need to be. In the game there are two different symbols which indicate two different kind of exclusion of other symbols. Their rules aren't readily compatible if they were to be used in the same panel. As far as I can remember they are never used together in the game, but at the same time that begs the question why there are two different types at all. I can understand why one of them was introduced in a particular area and then barely used in the rest of the game, as it only applies to one other set of symbols. Yet given the view that one part of the game is about learning mechanics and the other is about implementing those mechanics, the inclusion of this symbol in one area seems like an ad-hoc solution to flesh out a section that would otherwise be rather short.

It is easy to confuse these kind of inconsistencies with an expanded ruleset that keeps elaborating on the players knowledge and possibilities. I would therefore say that there is an important distinction between having a player expand their understanding and simply negating or ignoring something that was previously explicitly shown as an (im)possibility.

These criticisms aside, The Witness is definitely an unique game. It is vaguely reminiscent of the nearly uncrackable adventure games of the 1990's and as such it often gets called an adventure game, as well as a puzzle game.
The thing is, I don't think The Witness fits into either of those monikers. In my opinion it's an understanding-what-this-thing-is game. This works on many different levels, but is best exemplified by the fact that the very last part of the game, far beyond the ending as described in the beginning of this text, is something called 'The Challenge'. The Challenge is a tightly timed sequence of puzzles that are randomised each time you attempt it, while also requiring you to memorise the solutions to puzzles early in the sequence to use in a later part of the sequence. To be able to complete this challenge, you have to fully internalise the mechanics of the game, as nothing short of skill and understanding will allow you to complete it.
So while the game seemingly is about solving puzzles, or discovering the so-called secrets of the island, in my experience the very core of the game is concerned with not only exploring, but also communicating, the fullest of consequences of what can be done with a single concept. It's about taking a simple idea, drawing a line between two points, and expanding that idea as far as it can go while still maintaining a cohesive, playable game. Making that game cohesive and enjoyable is trying to let the player venture on this journey as well and let them think about this concept as if they thought of it themselves. In other words, The Witness is an understanding-what-this-thing-is game.

It is precisely in this point that The Witness is different from any other game I've ever played and I think it therefore has gotten a somewhat odd reception. The game has received critical acclaim for the most part, but neither the positive nor the negative reviews manage to make a meaningful distinction about this point. For some the lack of a singular narrative and meaning has been a point of criticism, but this almost becomes irrelevant if the game is seen as a far-reaching exploration of a single mechanic. The little narrative that is available in the game then mostly serves as commentary on the exploration itself, not its outcome. This is very clear at the endgame, where a secretly recorded casual conversation about the inclusion of the voice recorders is re-enacted by voice actors and played from a voice recorder in the game.
Likewise and oddly enough has The Challenge at the end been negatively perceived by some critics as a strong break with the 'calm and contemplative' gameplay of the rest of the game. I don't think this is the case at all, with The Witness cramming in a great deal of implications in a relatively short amount of time, with The Challenge merely being a possible, but logical, outcome of those presented facts.
A very common comment I have encountered are the 'incredibly difficult puzzles', but that really isn't true either. If The Witness came with an exhaustive manual detailing all the rules and mechanics of the different puzzle types, I believe arriving at the first ending in The Witness will be a somewhat trivial matter for most players. Solving those puzzles therefore hasn't been interesting at all in my view.
What however is very interesting and peculiar to me is that at the same time I also believe if a player was somehow in possession of this manual, it will take them much longer to complete The Challenge at the very end. While I at times have my problems with the way The Witness teaches its mechanics, I am in full agreement with the idea that this kind of self-reliant teaching ultimately leads you to a deeper understanding of what the game is.

One of the concerns that Blow had while making The Witness has been investigating what makes video games unique as a medium. The obvious answer is interactivity and it is precisely this kind of deep, elongated, investigation into the workings of an otherwise lifeless object that interactivity allows for. In the physical world it's never sufficient to merely analyse the components of a game in order to figure out how it is played. Chess pieces don't tell you anything about even the basic rules of chess and a football field doesn't tell you anything about the offside rule. Software in general, and video games in particular, are the only medium that essentially consist of nothing other than rules. To explore the full extent of a single rule therefore seems like a good way to investigate the 'true' possibilities of the medium.
The Witness is commendable as possibly the first game that attempts to discover what can be done if a simple rule is truly taken to its furthest consequences, without taking the common real world limitations of finances, time or company structure into account. It has come short at this attempt in some places, but being the first of its kind these shortcomings are expected and can be overlooked when the game is taken as a new beginning, rather than an end.