Topic: games
Thesis: one can use “games” to talk about many things, but first one needs a good definition. Here’s a useful one: an activity where one trades freedom for security.
The concept of a “game” is the site of a fascinating, twisting-and-turning philosophical discourse. One philosophically important thread begins with Ludwig Wittgenstein, who had two noteworthy claims about games [Magne@Medium]:
The word “game” does not have a strict definition; rather, it refers to a set of activities that all share a family resemblance. Despite the lack of a strict definition, one can truthfully describe games as usually having rules and a point—that is, a goal to be worked toward within the confines of the rules.
A certain class of games, called language games, are embedded in everyday speech.
Language games won’t be my focus here. What interests me is this twin tendency in Wittgenstein’s thought: both to see games as a hazily defined family of things, in fact even to take games as his canonical example of something defined only by family resemblance; and to see the concept as somehow very important, so much so that it goes to the core of his accounting of language itself.
This twin tendency recurs elsewhere. A prominent case is the book Finite and Infinite Games [wiki] by the religious scholar James P. Carse. Carse uses the concept of games in an even more ambitious way than Wittgenstein. In his category of finite games he places all of what we generally think of as a “game”: baseball, chess, rap battles; and also plenty that we would not think of as a game but nevertheless is conducted according to rules: elections, wars, hearings, trials. Against all this he constructs the idea of the infinite game, which is unbounded, whose rules are free to shift, and whose only purpose is to continue the game. Per Carse, we are closest to a truthful and well-lived life when we are immersed in play of the infinite sort.
I will admit to a little discomfort, and even a pang of moral disapproval, at Carse’s idea that—to uncharitably paraphrase—existence is but a game. It is this very discomfort that motivates me now to write about games.
The cause for my discomfort, the root cause, is the issue of definition. By games Carse seems to refer to something quite different than what comes to mind for me with the word, and so I take a meaning from his phrases that he likely did not intend. Games, for him, have rules, boundaries, and an ultimate victor; the only exception is the infinite game, which has none of these. I am led to ask: what, then, makes it a game?
Carse’s answer seems to be that the infinite game defines itself in contradistinction to finite games: the goal of a finite game is to end, the goal of the infinite game is to continue; finite games are bounded, the infinite game plays with all boundaries; finite games have fixed rules, the rules of the infinite game are provisional, not only able but required to shift—the rules may not and do not stabilize. The only thing finite and infinite games have in common is the element of agency: players can choose whether they are players.
Indeed, the concept of a “game” in Carse’s thought may in its intent be exactly playful. His definition even of finite games is strikingly liberal: in his book he studiously avoids making any distinction between games played for fun and the rule-bounded struggles of war and society which he also calls “games”. Ironically, my difference with Carse might be that I take the notion of games seriously, and so I have difficulty meeting him in this device.
I want to explore further the tension here. I think the notion of “seriousness” itself ends up being quite important to my difference with Carse on the idea of a game. I also think we can do a lot better than Wittgenstein suggests as to defining what exactly a game is, and that this definition will go a long way in clarifying how one might align or differ with the concept of the infinite game.
Defining games
Let’s start with the definition itself. It is true, as Wittgenstein says, that definitions are never perfect substitutes what they try to define. Still, I believe my definition here does come reasonably close to what I or Wittgenstein or Carse mean by a “game”; and, where it fails, the failure is at least instructive. Let’s take a look:
A game is a restricted way of acting and thinking that we choose to impose on ourselves, where we exchange freedom (unboundedness of action) for security (boundedness of stakes).
Per our definition, if someone is playing a game, that implies they have chosen to abide by its restrictions. These restrictions are not just restrictions of action, but also of thought. The choice to play a game is subjective. No-one can, per this definition, be forced to play a game, any more than they can be forced to believe that there are five lights when there are four. Someone may be physically coerced into miming the actions necessary to play a game, as some of us will no doubt remember experiencing in phys ed; but ultimately the final decision, to internalize their role in the game and to live in it as they play, is up to them.
How does this definition fare against real-life examples of games? How well does it separate games from non-games? Let’s consider a few examples.
Chess restrains how one can interact with pieces on the board. All other actions are not restricted. In exchange for accepting these restrictions, the stakes of competition with the opponent are likewise bounded: if it’s a competition, there may be previously defined stakes, but otherwise one knows that the worst possible outcome is to have the experience of being defeated in chess.
Baseball restricts the actions of players on the field according to their role. In a sense, the stakes are unlimited: there is nothing in the rules that prevents a player from taking a ball to the head and suffering fatal trauma. However, players are restricted from acting with the intention of harming each other. By taking on this restriction, one benefits from a limitation of stakes: one can trust that other players will not act intentionally to harm them, so any injury a player sustains will most likely be the result of their own action (e.g. pulling a muscle in a bad slide). Apart from this, the stakes are clearly defined and not substantially different from the stakes of chess.
Charades, for the performing player, does not directly restrict one’s actions; instead, it restrains one’s attention to the task of embodying or indicating the prompt. The performer agrees to the restriction of focusing their attention on the task; in return, the other players agree to interpret whatever they do in that light. Therefore, the performing player makes an exchange: in service of their goal, they are allowed to do embarrassing things with their body without the fear of social judgment they might otherwise expect. The exception that proves the rule here is that, if a player on the bench should mock a performer’s behaviour, they ruin the game for everyone by violating the game’s implicit contract.
Minecraft, as a computer game, imposes the restrictions on behaviour implicitly and automatically: for example, a player does not directly choose to accept the inability to mine obsidian with their hands. They implicitly take on all the game’s restrictions when they start it up; in exchange, the stakes are limited to what they can see on-screen. (C. Thi Nguyen writes about games as agency sculpting because of how they introduce constraints on the player that allow them to discover, through adaptation, new ways of engaging with the challenges the game presents [twitter, Oxford Press].)
Spin the bottle commits players to an action: specifically, kissing whomever the bottle should point to when it comes to rest. The stakes are therefore exactly the experience, enjoyable or no, of having that randomly selected kiss. Because the outcome is random, players have the freedom to decide for themselves whether the kiss really “mattered” beyond the game, and to conceal whether or not they enjoyed it.
Play, in the very broad sense that describes the behaviour of children or pets, constrains behaviour inasmuch as the players may not intentionally act to injure one another. We can see very clearly, in both dogs and children, how quickly a player ceases play when they perceive the other player to have broken this rule.
So far, we’ve looked at many quite different games and seen that they match the definition. Some have a point, like baseball or chess; others do not. All of them show some restrictions on the actions of players, as well as some restrictions on the good or bad outcomes of those actions. They also, I suggest, each imply a way of thinking that complements the game’s restrictions, both making it perfectly natural to comply with those restrictions and allowing the player to fully inhabit their engagement in play. This way of thinking may not be easy to put into words: sometimes, indeed, the way of thinking is the main reward for learning the game.
What about edge cases of the definition: things that seem like they could be games, but in the end are not? Let’s consider some of those examples.
A court case is like a game inasmuch as the players agree to certain norms of behaviour in exchange for boundedness of stakes: the limits of the law, further limited by the judgement of the judge and jury. However, it is not a game insofar as its agents are coerced into participation. In addition, in society we treat legal rulings as real: to many people, a defendant convicted of a crime is really a criminal in a way that an acquitted defendant can never be. The legal system’s way of thinking imposes itself on social life at large—and to the extent this is true, it does not meet our definition of a game. (Carse, by contrast, would consider a court case an example of a finite game, because even a defendant who is dragged into the courtroom can still choose not to play the role of defendant. More on this in the sequel!)
A duel is like a game inasmuch as the agents agree to constrain their violent actions in exchange for boundedness of stakes. A participant may take a fatal bullet, but it’s understood that if this should happen it will only be according to the rules; and, whatever the outcome, no further violence may arise from the dispute that prompted the duel. However, when duels were common practice, to refuse a duel was an act that brought shame and lost status. The threat of lost status qualifies as coercion, and so a duel typically can’t be considered a game.
“Game” can also be used to describe animals hunted for sport. Is sport hunting a game? From the perspective of a hunter, it is: the sport hunter tries to kill the prey with some restrictions, albeit hazily defined, that are meant to keep the game “sporting”. To illustrate the point: “fishing” with sticks of dynamite does not qualify as sport fishing. The reward for following these restrictions is a sense of personal pride and victory if the hunt should succeed.
From the perspective of the animal being hunted, however, the pursuit is (usually) not a game—perhaps with the occasional exception of some legendary lion or bear. Therefore, the status of sport hunting as a game is relative. It is a game from the perspective of one agent, but not from the perspective of the other. We can call this a conflicted game: i.e., a game where not all agents agree on the game.
Another example of a conflicted game is the “game” of the 2021 Netflix series Squid Game (spoilers for Episode 1). In this case, players compete for a prize where failure means death. Significantly, the contract they sign at the beginning of the competition uses the euphemism “elimination” to describe death; accordingly, many players die before they understand the true stakes.
This is one of two things that makes the organisation’s game a conflicted one. The other is that the people competing are only allowed to quit if they make this decision collectively by majority vote. It is a condition that serves a narrative purpose: the game is metaphorically linked to competition in a capitalist society. In both cases, people embedded in the current order wish to preserve it if they think it will ultimately work to their personal benefit. Clinging to this wish, they resist efforts to change the current order—forcing others, and sometimes themselves, to pay with their lives.
Let’s close with an especially interesting case: the act of reading a novel. Can one argue that reading a novel is equivalent to playing a game? A version of that argument might go like this:
“When I read a novel, I am accepting a constraint on my behaviour—to keep my eyes on the page and read what’s on them—in return for certainty of the outcome. I can know that I won’t be directly affected by what happens in the novel because it is not real.”
This argument helps to clarify what the trade-off of a game really is, because it is not what this argument assumes. Reading a book is not “constraining oneself” to look at the page any more than going for a swim is “constraining oneself” to go into the water. What separates a game from a non-game is not the choice to engage. Rather, what separates a game from a non-game is that a player, having entered into the game, accepts the further constraints that the game implies. To move a piece on a chessboard is not to play at chess; to move a piece with knowledge of the rules, and with the intention of following them, is to play at chess.
Yet there is a stronger version of the argument that goes like this:
“When I read a novel, I accept a constraint on my thinking. I agree to treat the characters and events as if they were plausible and real. In return, I know that the story cannot affect me to a greater degree than the degree to which I care.”
This suggests an interesting framing for novels: one can think of novel-reading like emotional gambling. I can, to some extent, choose whether to “play along” with the author of a novel as I read it, treating it like it’s real and absorbing its impacts. The more invested I become in this story, the more trust I place in the author and the greater the emotional stakes. If I care, the author can lift me to great heights, or wound me so deeply that I lie in bed for days.
As a player in a game, I can withhold this cooperation if I don’t think the author is up to the challenge of bringing things to a conclusion in a satisfying way. I can continue reading, but stop caring. This is, perhaps, the game embedded within reading a novel: the challenge of getting the most out of the story by controlling whether, how much, and when to invest one’s care (or, indeed, when to abort reading and deposit the book in the nearest Little Free Library). In a linear medium like writing, this is all an audience member has; but if one chooses it, it is enough to enable a certain kind of game.
Now that we have this definition, in all its fuzzy usefulness, the next step is to show how it applies beyond its boundaries: that is, how we might use games as a metaphor to better understand the flows of society, power, and coercion. That discussion is one I’ll reserve for a follow-up post. What we have so far is a definition that reveals something essential about games and gives an instruction manual for finding games in unexpected places.
Recap of terms
Game: a restricted way of acting and thinking that agents choose to impose on themselves, where they exchange freedom (unboundedness of action) for security (boundedness of stakes).
Conflicted game: an activity of multiple agents that some treat as a game, others not; or that some treat as one game, others as an entirely different game.