## Friday, January 26, 2024

### The authority of game rules

I just set myself this task: Without moving my middle and index fingers, I would wiggle the ring finger of my right hand twenty times in ten seconds. I then fulfilled this task (holding the middle and index fingers still with my left hand). What was I doing by fulfilling my intention? I think I was playing and winning a game, albeit not a very fun one.

Here are four ways of describing a game fitting my intentions:

1. Victory condition: Wiggle ring finger 20 times in 10 seconds. Rules: Don’t move middle and index fingers.

2. Victory condition: Wiggle ring finger 20 times in 10 seconds without moving middle finger. Rules: Don’t move index finger.

3. Victory condition: Wiggle ring finger 20 times in 10 seconds without moving index finger. Rules: Don’t move middle finger.

4. Victory condition: Wiggle ring finger 20 times in 10 seconds without moving middle and index fingers. Rules: None.

Each of these generates the same gameplay: exactly the same things count as victory, since cheaters never win, and so you win only if you follow the rules. Ockham’s Razor suggests that they are all the same game. In other words, whether we put some constraints as rules or build them into the victory condition is just a matter of descriptive convenience.

If this is right, then here is a first attempt at a simple account of what we are doing when we play a game:

1. To play a game is to try to achieve the game’s victory condition without breaking the rules.

This gives a neat, simple and reductive account of the authority of the rules: Their authority comes from the fact that according with them is a necessary condition for achieving an end that one has adopted. It is simply the authority of instrumental rationality.

Of course, all sorts of complications come up. One is that games sometimes have score instead of a victory condition. In that case, what you are doing is aiming at higher scores rather than trying to achieve a specific victory condition, with some sort of an understanding of how breaking the rules affects the score (maybe it sets the score to zero, or maybe it you get whatever score you had just before you broke the rules). This tricky, and I explored these kinds of directional aiming in my ACPA talk last fall.

A bigger problem is this. My story started with a single player game. But things are more complicated in a two player game.

Problem 1: According to (5), if you plan to cheat, then you aren’t trying to achieve the victory condition without breaking the rules, and hence you aren’t playing the game. But to cheat at a game you have to play it! So in fact you never cheated!

Response: I think this is the right result for a single player game you play on your own. By planning to cheat in some particular way, I am simply the changing what game I am playing—and I can do that mid-game if I so choose. For instance, speedrunners of video games sometimes set themselves rules for what kind of “cheating” they are allowed to do: Can one use emulation and save states? Can one use glitches? Can one use automation, and where? If they are playing solely on their own, none of that is really cheating, because it is allowed by the rules they set themselves: their goal is to complete the game faster within such-and-such parameters.

But if you are playing against another or there is an audience to the game, things are different. We could just say this: You are cheating in the sense that you are deceiving the audience and the other player into thinking you are playing the game. Or we could say that there are two senses of playing a game: the first is simply to try to achieve the victory condition without breaking the rules, and the second is an implicit or explicit agreement with other persons to be playing a game in the first sense (e.g., indicated by signing up for the game).

Problem 2: Suppose I plan to beat you at chess while following the rules. To that end I drive to your house. Then my driving to your house is an attempt to achieve chess victory without breaking the rules, and hence by (5) the driving appears to be a part of the gameplay.

Response: This seems to me to be a pretty serious problem for the account actually. There seems to be an intuitive distinction between an action according with the rules that promotes victory and a move in the game. Another example is taking a drink of water while playing chess and doing so with the intention of improving one’s mental functioning and hence chances at victory. In taking the drink, one isn’t playing.

Or is one? When I think about it, the distinction seems kind of arbitrary and without normative significance. Take an athlete who is training for the big game. We want to say that they aren’t playing yet. But notice that in most modern sports the training itself falls under rules—specifically, rules about performance enhancing drugs. We could easily say that the training is basically a part of the game, a part that is much more loosely regulated than the rest. Similarly, many sports regulate the breaks that players can take, and deciding how to apportion the allowed break time (e.g., to take a drink of water, to relax, to stretch, or to refrain from taking it in order to make the other player think one isn’t tired) seems like a part of the game. Or take bodybuilding. It seems quite a distortion of the game to think only of the time in front of the judges as the gameplay.

What about the drive? That sure doesn’t seem to be a part of the game. But is that so clear? First, in a number of settings not showing up yields a forfit, a kind of loss. So you can lose by failing to drive! Second, you can choose how to drive—in restful or a stressful way, for instance—based on how this will affect the more formal gameplay.

Note that in chess, thinking is clearly a part of the gameplay. But you can start to think however early you like! You can be planning your first move should you end up winning the toss and playing white, for instance, while driving. Or not. The decision whether to engage in such planning is a part of your competency as a player.

Problem 3: It seems possible to play games with small children without trying to win, hoping that they will win.

Response: Maybe in such a case, one is pretending to play the game the child is playing, while one is playing a different game, say one whose victory condition is: “My child will checkmate me after I have made it moderately difficult for them to do so.” This is deceitful (the child will typically be unhappy if they figure out they were allowed to win), and deceit is defeasibly wrong. Is there a defeater here? I have my doubts.

Problem 4: I can try to run to my office in five minutes as a game or in order to be in time for office hours. There is a difference between the two: only in the first case am I playing. But in both cases the definition of (5) is satisfied.

Response: Maybe not. In (5), we have an ambiguity: it is not clear whether the fact that the victory condition is the victory condition of a game is a part of the content of one’s intention. If it is a part of the content of one’s intention, then the problem disappears: when I run in order to be on time, I am not aiming to get there in time as a game. But if we require the ludic component to be itself a part of the intention, however, then we lose some of the reductive appeal of (5) absent an account of games.

Should we require a thought of a game to be a part of the intention? Maybe. Suppose you find an odd game machine that dispenses twenty dollar bills if you tap a button 16 times in a second (quite an achievement). You try to do this just to get the money. Are you playing a game? I suspect not. For suppose a friend on a lark took your driver’s license and loaded it into the machine, and the only way to get it back is to tap the button 16 times in a second. You aren’t playing a game!

Perhaps we could say this. Instead of requiring the game part of (5) to be in the content of the intentions, we can require that something about the victory part be a part of the intentions. And perhaps what makes something a victory for you in the relevant sense is that in addition to whatever instrumental and intrinsic value it has, it is in part being pursued simply to achieve a goal one has set oneself as such. That’s what makes it not entirely serious. A game is a kind of whim: you just decided to pursue a goal, and off you go, because you decided to do so. (Of course, you might have good reason to have whims!)

Problem 5: Doesn’t the story violate the guise of the good thesis, since the victory need have no good in it apart from your setting it to yourself as an end, and hence your setting it to yourself as an end can’t be justified by its good.

Response: The case here is similar to one of my favorite family of edge cases for action theory, such as when you get a prize if you can induce electrical activity in the nerves from your brain to your arm, so you raise your arm. The raising of your arm has no value to you. But you have reason to set the rising of your arm as an end, since setting the raising of your arm as an end is a means to inducing the electrical activity in the nerves. In this case, the rising of the arm is worthless, either as an end or as a means, but aiming at it has value, since it causes the electrical activity in the nerves. (Depending on how one understands the response to Problem 4, it may be that one is then playing a simple little game with oneself.)

Similarly, if something is purely played as a game, the end has no value prior to its being set. But there are two values that you can aim at in setting victory-by-the-rules as your end: the value of achieving ends and the value of striving for achievable ends, to both of which the setting of achievable ends is a means. And maybe there is the good of play (which may or may not be subsumed under the values of striving for and achieving ends). So you have the guise of the good in an extended way: the end itself is not an independent good, but the adoption of the end is a good. That’s how it is in the electrical activity case.

Final remarks: The difficulties do not, I think, affect the basic account that the fundamental normative force of the rules of a game simply come from instrumental rationality: following the rules is necessary for the achievement of your goal. And there is a secondary normative force in multiplayer games, coming from general moral rules about compacts and deceit.

But perhaps we shouldn’t even say that there is normative force in the rules. They simply yield necessary conditions for achievement of one’s end. Perhaps they have no more, but no less, normative force than the fact that I need to exert energy to get to my office is.