top of page

The Philosophy of Game, Part 1: Basic Definition of Game

There’s a lot of examination of the mechanics of game design and much of it is quite excellent. At the same time there are essays, lectures and discussions on the semantics and philosophical definition of gaming. I’ve never been entirely happy with any of them, eventually it has occurred to me that as a published game designer and qualified philosopher (if that term has any meaning, MA in modern continental and analytic philosophy) then if not me, who?

This series of essays will take on quite close readings of existing positions and will go into some discursive depth on a range of issues, as such they will tend to be quite long and detailed. I will attempt to avoid too much philosophical jargon but when I do use it I will try to take the time to explain concepts in layman’s terms. This may make such breaks tiresome to those familiar with philosophical terms, but that is a risk we’ll have to take.

In this first essay I will attempt to define “game”, beginning by running through some popular existing definitions and picking out some failings before advancing my own version. Chief among those existing definitions is that of Jesper Juul who has suggested that to be defined as a game something must be: 1) Rule based, 2) Have variable, quantifiable outcomes, 3) Have a value attached to the outcomes (some outcomes are better than others), 4) Involve player effort, 5) Have an attachment of players to the outcome (winners ‘feel good’ and losers ‘feel bad’), 6) Have negotiable consequences (the game can be played with or without real life consequences). This definition has a couple of interesting points but ultimately it suffers from significant flaws and issues.

For point one, “Rules based”, almost every activity has rules, without further definition of “Rules” this element is largely meaningless. In his essay Juul suggests that “Free Form” play is not a game and is not a game because it does not have rules. Firstly, it should be mentioned that totally free form play, with not one rule or definition of who or what the participants or objects portray is generally engaged in by children so young as to blur the lines between play and basic participation with the world. Most free form play has rules, if only those contained within the larger society in which the game exists. The first form of free form play rules are simple descriptors of what is happening and what people or objects represent but these definitions, which are usually separate from reality, are intended as rules. The free form player defines a pencil as a screw driver not simply as part of the play, but to set rules pertaining to how the pencil may be used within the play. The second form are that the rules of society still persist over the free form play, no matter how free form our play is, assault and battery during play remains generally unacceptable unless specified as part of the rules of certain game sports. This may seem an unhelpful point but it is vital to acknowledge that this element of definition actually means “Is ‘Game Rules’ Based”, which is to say that games are things that have rules that persist only and specifically during the game (which is still true for many activities) and which are arbitrary, but I will go on to examine this element in my own definition. Suffice to say, “Rule Based” does not achieve its intended goal of definition.

In relation to having variable, quantifiable outcomes Juul specifies Ring-a-ring-a-roses, presumably because participants go through a set of pre-defined steps that they can perform in their own personal fashion, but that the final outcome is always the same. The problem of this definition is that it is, oddly, both too broad and too narrow. Any computer game with a single ending is ruled out by this definition. Personally, I have played computer games on high difficulty levels and after re-loading multiple times had the definite sense that there is only one possible outcome, eventually I will perform the necessary steps to achieve victory, with sufficient time even by luck alone I will succeed inevitably. This definition only works if we do not see the process of dying and re-starting in a computer game as part of the game, which even casually we do, we refer to loading a single game back up, starting a new game only when we regress to the beginning and start over. Once the new game is begun though there are, in fact only two actual outcomes, we proceed to the end or we abandon the game and walk away. Between we will simply go through a set of pre-defined steps that we can perform in our own personal fashion. Single end scene computer games and Ring-a-ring-a-roses are actually identical under this definition, both ruled out, or in if walking away is an outcome we count as quantifiable within the game. Because, once again, all activities have variable, quantifiable outcomes. Every activity has at the very least the variability of being completed or not, and most day to day activities have a huge array of quantifiable outcomes.

Within the need for variable outcomes Juul suggests the necessity for a level of player agency in variability, giving specifically the game of “noughts and crosses” (Tic-tac-toe in the states) as an example, since with two adequately aware players the result is not variable, it will always be a draw, so to achieve practical variability it requires players to proverbially “make their own fun” by altering the game in some way, as is true for most games where skill is vastly uneven. It could be suggested that the inevitably of certain mismatched or perfectly matched conflict games are why some people play them, what even defines them, since a pre-determined result is one over which no hard feelings can be suffered. At the very least it brings them into line with a whole sub-set of patience games where the outcome is pleasing but largely inevitable.

Player agency in shaping a definition is also arguably a problematic position, making the Valorization of the outcome (some outcomes are better than others) a difficult point. Valorization is again either largely dependent on the players, or is simply an extension of definition one. Rules generally define the aims of the game and suggest that by participating in the game players are striving to achieve them, but the decision to value that outcome is entirely personal. Many gamers, particularly small children, will claim to value their losing position over their opponent’s victory in a so called “moral victory”. One of my own games allows players to sacrifice their chance of victory for the good of the group, since they may value their nobility beyond the winning of a game. Games have a rule defined end state and define play as a path towards that end state but this is simply a necessity for the production of consistent rules mechanics and should not be separated from the definition of the existence of rules. Games are incapable of insisting on a valuation of that end state. It may be the definition of a good game as opposed to a bad one that it feels harder to not value the outcomes in a good one, but it is more than possible not to value the outcome of a bad game at all. The nature of Juul’s counter example here, “watching a fireplace” is indicative of how odd a definition step this is. The question of valuing the outcome of “watching a fireplace” is mainly the question of what the outcome of watching a fireplace is? I would suggest that the outcome of watching a fireplace is a sense of cosy well-being and security, at which point the idea that participants don’t value that sense of well-being seems difficult to posit. Valuing the outcome of fire watching only excludes it from Juul’s definition if it has no outcome, but the definition is not that the activity should simply have an outcome, or that it defines its outcome, simply that if there is one it should be ranked and valued. Presumably the argument would be that there is not a variety of endings to watching a fire, which there is since we can walk away before achieving our end goal of well-being, but even if that were the position it is not one relevant to this specific definition.

Juul uses Player effort to place games of chance as borderline game cases. This is problematic since games of chance are a substantial section of played games, in fact it is probable that including gambling machines and national lotteries games of chance are the most played form of game. A definition of games that makes the most played of games a borderline sub section has to be questionable, since at that point one would define games as games of chance and games of skill as a sub-set of the whole. That aside, participant effort is needed for almost any activity, watching a movie is Juul’s example of an activity not requiring participant effort, but watching all but the most plot light movie does require effort in the form of focus and attention. The act of watching a fire is again included, though the existence of pokers and other fire tending paraphernalia does imply a certain interactivity available to those who might choose to watch a fire. In this instance Juul’s choice to define the activity of “watching” movies, fires or games of chess is used to suggest that things watched are not games, but they can be, the act of watching a thing is different from participating in its processes, generation or maintenance. Tending a fire is an interactive activity and watching a game of chess is passive, but chess is a game and fires are not. The point seemingly being made is that games are interactive in that participants can affect their outcome, rather than their personal experience of the outcome. But not every participant can actually affect the outcome of every game, seriously overmatched opponents or those ruled out of activity by poor design can still be playing a game but have no agency.

Re-formulated this definition might be that at least one participant can affect the outcome of the game for participants other than themselves, or the ‘actual’, outcome. This doesn’t fully rule out activities such as watching movies, since vocal participation in viewing is more than able to affect the nature of the experience for others, but it will presumably not affect the ‘actual’ outcome of the movie separate from the participant’s perception of it. This layer of definition would not form an effective defense though since it applies only to acts that are watching rather than participating in an activity, in fact the first definition, that of an activity having rules (specific to the activity itself) will rule out every activity this definition rules out, since it is largely illogical to have specific rules for a non-interactive process, while simultaneously ruling out additional activities this definition allows.

Player attachment to the outcome of the activity is the next element of definition and it is probably the least helpful. It is possible for participants to become attached to any given outcome of any given activity and equally possible for them to avoid being attached to outcomes in activities. Juul’s own paper gives as an example the spoilsport who refuses to become attached to the winning or losing of a game, but the spoilsport does exist and that which they play remains a game irrespective of their opinions, and Juul again returns to the activity of watching a log fire as an example of an activity impossible to become attached to. In actuality it is the easiest thing in the world to imagine the builder of the fire deeply attached to the result of watching it, (if said fire has been built on a cold night in a harsh environment it’s hard to conceive of a watcher not attached to the outcome of their watching) as they wonder when a certain log will burn through, which will burn through first, if it will collapse into a shower of sparks. The fact of human psychology is that it is capable of attaching to anything, no activity is inherently passive and few are inherently active.

The most significant of Juul’s definitions is the last, Negotiable Consequences. Within a game the consequences are, apparently, negotiable, we can decide and control their extent and limitations. Juul acknowledges that every one of the previous definitions would not rule out driving a car from being a game, but this final definition is intended to solve that issue since “the consequences of traffic are not optional - moving in traffic always has real-life consequences”. However, I would argue that this element also can be shown as flawed from both sides, because it depends on a pre-accepted definition of consequences.

Games have some negotiable consequences, but they have some non-negotiable consequences. The playing of a game has the non-negotiable consequence of not having spent time doing something more important and useful and the acknowledgement of gaming addiction suggests this consequence as being significant. All human activity costs time and therefore has a base level of consequence. But additionally, though the cost of losing a game can be negotiated and partly that cost is what is being referred to, the consequences of losing are not the consequences of the game, the consequence of the game is losing. The fact that once you have begun a game with a win/loss end state you will eventually be defined as winning or losing is not negotiable. Juul gives war as an example of an event without negotiable consequences, but wars and games have non-negotiable consequences, and the consequences for losing them are in both cases fully negotiable. The consequences of participating in the game, that dice will be rolled and figures moved and the consequences of war, that men will be shot and land taken, are in both cases non-negotiable given that the game or the war is being participated in. That the consequences of winning or losing either is negotiable is largely inherent to the processes of gambling and peace talks.

So, approaching from the other side, the concept that moving in traffic always has real life consequences. Generally moving through traffic competently always has a set of consequences, the reaching a set destination, but again, this is no more of an actual consequence than the consequence of playing a game. ‘I am in Basildon’ is no more of a consequence for actions than ‘I have completed a game of “This War of Mine”’. The implication seems to be that serious unforeseen consequences that one would wish to have avoided can occur due to driving, but they cannot die to playing a game. Clearly this is untrue, due to my playing a game and taking my attention from cooking my dinner I can burn down my house and kill all my loved ones just as easily as if I drive them off a cliff.

In short each of Juul’s definitions seem flawed. Juul was selected here because the definitions above are part of a Meta project taking into account several other definitions, but I will now further address some of these other definitions of game.

Johan Huizinga (1950) “a free activity standing quite consciously ‘outside’ ordinary life as being ‘not serious’, but at the same time absorbing the player intensely and utterly. It is an activity connected with no material interest, and no profit can be gained by it. It proceeds within its own proper boundaries of time and space according to fixed rules and in an orderly manner.”

This has two primary issues as a definition of games, firstly in that games are able to be played professionally with a range of profits and material interests attached. Secondly, games are more than able to spill out of their set and bounded areas. Chess players can think about their moves and problems constantly and far beyond the physical playing of the game.

Roger Caillois (1961) ”an activity which is essentially: Free (voluntary), separate [in time and space], uncertain, unproductive, governed by rules, make-believe.”

Sadly, much of life is uncertain, unproductive, voluntary and governed by rules, making these terms too broad to be useful. Games are not generally actually dependent on make believe, chess does not actually attempt to recreate the movement of bishops nor requires you to imagine high ranking men of the cloth assaulting siege engines. Again, the claim of spatial separation for games is dubious as a necessary definition since games can spill through time and space, particularly with the growing popularity of public “event” gaming.

Chris Crawford (1981) “four common factors: representation ["a closed formal system that subjectively represents a subset of reality"], interaction, conflict, and safety ["the results of a game are always less harsh than the situations the game models"].”

While the last definition holds given the first, it is hard to explain next to abstract games. The claim that ‘Blokus’ represents or models a subset of reality is not strictly viable, given that not all games represent subsets of reality (in fact very few honestly attempt to) they cannot be defined by the lesser nature of their results in comparison. Interactivity and conflict are present outside of games and are not inherent to them and so fail to uniquely define games.

There are a range of other definitions that make reference to rules within games as defining features, but two are of particular interest to my suggested definition: Bernard Suits (1978) “using only means permitted by rules, where the rules prohibit more efficient in favor of less efficient means, and where such rules are accepted just because they make possible such activity.” and David Kelley (1988) “a game is a form of recreation constituted by a set of rules that specify an object to be attained and the permissible means of attaining it.”. These definitions suggest not only that the essence of games lies within the nature of their rules but that their rules exist only in order to sustain the game state (just because they make possible such activity). They also suggest that the purpose of the game state can be seen by their rules defining their aims (specify an object to be attained).

I would suggest that the essence of the definition of a game is two-fold, firstly to define games with reference to their nature as activities that are in whole defined by their being subject to a specific sub-set of rules and secondly by defining the purpose to which games are put which can be found by virtue of their goals being outlined within those rules.

Games are defined by rules in a manner not present in most activity, there is a clear difference between game rules and the rules of general life. Life has various categories of rule, rules of nature, rules of society and rules of intent, all of which are significantly different from Game Rules, but all of which persist during games. While playing a game the “rules of nature” cannot be ignored, in the case of game playing this means that to play a game the player must obtain the game in some form and cannot play it without first becoming aware at least in part of how to play it, they will have to set-up the game state in some fashion before participating in it. The rules of society splits into laws and conventions, the game player will need to abide by the laws of society during play but the game may allow them to contravene its conventions. Interestingly while games allow the breaching of certain conventions (usually that it is bad to wish injury and death on others) they introduce a whole other set of unspoken conventions based on fair usage of dice and cards, among other elements. While these rules persist over general life, and so make defining games as being under the auspices of rules an unhelpful position ‘game rules’ have a separate and significant existence. Clearly, defining games as activities defined by game rules is unhelpful, so what makes a rule a ‘game rule’ should be clarified.

Primarily, “game rules” may be dropped or ignored at any time with the only consequence of so doing that the individual who does so is no longer playing the game. Furthermore, they will be adopted only when the game is begun by its players and will be dropped once the game is officially ended. This separates them from societal or natural rules or laws which will persist