First published on Chris Bateman’s blog, Only A Game on 14th March 2006

So central to the modern videogames industry is agon (competition) that many people consider ‘game’ to be almost synonymous with the notion of competitive play. We play to win, the presumption states, and this indeed describes a great many of our modern videogames. I hope that, by having looked at alea, mimicry, and ilinx before agon, I do not need to put forward the counter-argument (that there can be more to games than agon) and can instead focus on exploring the myriad complexities that this category of play entails.

Agon is one of four cross-cultural patterns of play identified by the noted intellectual Roger Caillois in 1958. Caillios was writing at a time before the hobbygames explosion of the 1970’s, before the proliferation of arcade game novelty in the 1980’s and long before the modern videogames market. Consequently, his descriptions of agon are focused on the nature of sports and contests of skill. He described agon as follows:

A whole group of games would seem to be competitive, that is to say, like a combat in which equality of chances is artificially created, in order that the adversaries should confront each other under ideal conditions, susceptible of giving precise and incontestable value to the winner’s triumph. It is therefore always a question of rivalry which hinges on a single quality (speed, endurance, strength, memory, skill, ingenuity, etc.), exercised, within defined limits and without outside assistance, in such a way that the winner appears to be better than the loser in a certain category of exploits.

Whereas Caillois’ other categories of play have not required significant revision to bring into clear focus with respect of modern games, agon requires some slight expansion. When Caillois was writing, the only reasonable form of agon was between two people in a specific challenge or sport. (One could include, perhaps, the agon between matador and bull in the bloodsport of bullfighting, but since Caillois was insistent in seeing agon as a fair contest between participants, I believe this would not qualify: the bull’s chance of survival is considerably less than that of the matador).

A computer game allows the player to engage in a game of agon versus a virtual opponent. This seems to evoke the same behaviour as facing a living opponent. Facing the game, the player first sizes up their capacity to compete. People who cannot throw do not, for instance, enjoy participating in field events based around throwing. Similarly, people who cannot operate a first person shooter (a style of game with a particularly codified form) do not play FPS games in an agonistic style. But those that have capabilities in a particular type of game frequently then enter the state of wishing to test themselves against a degree of challenge. This could be against other opponents - which is quite obviously a fit to Caillois’ agon - or it could be against pre-set challenges, such as the main gameplay of a single player game.

It is worth noting that contests of agon present themselves in certain distinct forms which are worth identifying briefly, if only to provide a wider foundation. Firstly, there are games of one versus one agon, such as a fighting game. Such games are the most intensely agonistic form, and most recognisably fit Caillois’ description of agon. Then, there are one versus many games, such as the FPS which (squad variants not withstanding) place a lone player in contest versus many opponents. Still, the underlying assumption is that the challenge has been balanced fairly (although pragmatically, few games are so well tweaked for this to emerge). Finally, many versus many games, such as strategy games when whole armies fight, which are akin to team-based sports which offer agon between equally matched sides. These distinctions, while notable, do not fundamentally alter the nature of agonistic play, although they may alter the appeal. Fighting games, FPS games and strategy games imply different audiences, with an affinity for challenge in the forms of fast reaction control skills, pathfinding & aiming or thoughtful complexity respectively being key distinctions.

The asymmetric case of the one versus many scenario is something that we are highly familiar with in videogames, but which might have seemed strange to Caillois. Even in the case of one versus one and many versus many we see asymmetry in games: playing a fighting game, it is not possible for the computer opponents to be a balanced challenge for all players. Instead, the game presents a variable degree of challenge; the player climbs the curve of difficulty inside the game. Similarly, a strategy game often has scenarios of increasing difficulty. Instead of presenting a perfectly balanced challenge, the game starts easily (in principle, at least - few games balance themselves to be sufficiently easy for all comers), and then increases the degree of challenge until the player encounters the biting point where they know they will have to perform at the best of their ability to achieve victory. This is when Caillois’ agon really takes hold: prior to this point, one might imagine the player’s actions have merely been training. There is a parallel in sport: an Olympic athlete must still rise through many qualifying rounds against opponents of differing capabilities until they find they are competing against competitors of their own calibre.

This rising challenge versus the player is intended, arguably, to ensure that at some point the player will face a challenge worthy of their abilities. I suggest, therefore, that we can consider there to be some relationship between fiero and agon. Fiero, the emotion Ekman identifies with triumph over adversity, is associated with the physical gesture of holding arms aloft in victory; it is something we can readily observe whenever an athlete wins a difficult game, or in sporting fans whenever their team scores. It is, I attest, intimately connected to agon.

Out of respect for Nicole Lazzaro, from whom I first learned of fiero, I am considering games which tap into this fiero-motivated reservoir of challenge to be games of hard agon (after Lazzaro’s Hard Fun emotional key). These games seem most popular with Hardcore gamers, although this is not to suggest that hard agon is not enjoyed by many Casual players (possibly younger males in particular, although this is far from proven). Fiero can be a tremendously rewarding emotion, and therefore can be a tremendously addictive emotion. Why else would certain players pit themselves against challenges of such depth of adversity that they must endure nearly constant frustration as they pursue with dogged determination the repetitious play intrinsic in most games of this kind? When victory is achieved, the fiero 'pays for' all the pain experienced on the way. Indeed, the greater the depths of pain endured, the larger the fiero for some players. (For others, and especially for more mature players, the residue of frustration reduces the fiero to less engaging relief).

Games of hard agon dare players to beat them. Some are so punishing in the demands they ask of their players, that they practically reach the boundaries of becoming an ordeal. But, provided the fiero payoff is there, ultimately, it becomes worthwhile. It is a drama which we are very familiar with - the Rocky films alone made millions out of their gradual descent into formulaic fiero fantasy.

I argue that many players often do not notice the (debatably) ramshackle design of many games because when a game is offering hard agon, shoddy game design (bad control mechanisms, poor mechanics etc.) are just friction adding to the frustrations which are endured on the path to victory. This places a much greater burden on games which do not use agon as part of their core play to have superior game design, delivering a smoother play to their audience.

At the opposite extreme to the trials of hard agon are those games which offer trivial victory over opponents. Some of these games are diluting the agonistic play with aleatory (random) elements. Consider for one example, a party game like Mario Party or Bishi Bashi. These are ostensibly games of agon - but are really a heady mix of agon and alea. It is not the chance to prove superiority versus one’s foes in these games which provides all the fun, but the random chaos of the mini-games. It is also possible to combine agon and mimicry. I presume part of the fun of a game like The Incredible Hulk: Ultimate Destruction (putting aside the ilinx inherent in mindless destruction briefly) comes partly from the illusion of power being supplied. The player is invited to enjoy being strong relative to their opponents - practically the opposite of the fiero-related enjoyment delivered by hard agon. [Note: I haven't actually played this game, so I may be incorrect in this assumption, and it's a safe bet it rapidly devolves into hard agon; first hand opinions and alternative examples welcomed.]

I propose to call these styles of play easy agon (again, in respectful reference to Nicole Lazzaro’s influential Four Keys model). These games are almost invariably drawing upon a mix of play elements, diluting agon with either alea, mimicry or ilinx in varying degrees. Some players may consider there to be little of interest in play when it has been made so unchallenging (perhaps reflecting the importance of fiero to such a person). I personally find nothing problematic in exploring such child-like escapism against virtual opponents. Such play is amusing and entertaining, and far more suitable for stress release than the tension of fiero.

The space that the player ends up within in respect of any given game of agon is determined almost entirely by the strength of the player (determined in part by their own abilities, in part by the game parametrics) relative to the strength of the opposition. Games of hard agon are at the very least evenly matched, and more commonly are biased against the player, so that the player must work even harder to win, and thus achieves an even greater payoff in fiero. Conversely, games of easy agon begin when the player’s strength is weighted higher than the opposition - indeed, these games are arguably at their most fun (and by fun in this case we mean the fun of amusement, not the fun of fiero) when the player is ludicrously overpowered with respect to their opponents. This was surely what made Rampage fun to play when it first came to the arcades, and I assume the recent Hulk game shares something of this feel in its early play.

Because it is the relative difficulty which determines which space the player is in (hard agon, through the "true agon" of equally matched opponents, and finally to easy agon), games can deliver both kinds of play by simply having a large enough range of difficulties. The Dynasty Warriors games in particular demonstrate this facet. On their high difficulty settings, these provide hordes of tough troops and even tougher officers which lure the player to overcome the odds stacked against them and earn fiero. On their lowest difficulty settings (and perhaps after some judicious powering up), the player can enjoy the feeling of effortlessly cutting down hundreds of enemy troops, and knocking enemy officers about the battlefield like they were mere stuffed toys. Both can be intoxicatingly entertaining, but in general terms, these two different settings appeal to different players. (My love of these games is perhaps rooted in the balance between the hard agon I used to enjoy as a teenager, and still have a certain nostalgia for, and the easy agon I find more suited to relaxation in my rapidly approaching middle age).

The implications for game design are that for games wishing to court a wide audience, there may be much value in beginning by balancing the game for easy agon. This is the harder task for almost all games; it is very easy to conceive of means to make a game harder, but generally difficult to conceive of means to make a given game easier without subtracting game mechanics (and thus changing the nature of the gameplay). However, I suspect that game designers who work on highly competitive games of hard agon might argue that balancing for hard agon is an equally challenging task, if for no other reason than the Hardcore audience for hard agon is extremely demanding and particular in their tastes. I welcome insights from people with practical experience in this area, as my own disenchantment with designing games of hard agon limits my knowledge of this particular field.

The nebulous notion of game difficulty seems to be important when considering agon, whereas this was largely unimportant to the other categories of play identified by Caillois. I used to believe that it was possible to build self-calibrating games that could adjust to match the level of difficulty desired by the player. However, it is becoming readily apparent that the level of difficulty desired by the player is not something we can presuppose. Some players want to be met at about their level of challenge (they desire "true agon"), some want to be met with very little level of challenge (desiring easy agon) and some secretly yearn to have the game beat them into a bloody pulp that they may later emerge victorious and aglow with fiero (desiring hard agon).

Perhaps it will be necessary for games with dynamic difficulty mechanisms to identify the player’s desire with respect to agonistic play from the onset. Sorter questions may be the easiest option; something like: "Do you want to: (1) Play for amusement (2) Face a reasonable degree of challenge (3) Triumph against overwhelming odds." Wrangling the language so that this can be presented to players without confusion is a difficult task, but I remain confident that sorter questions are a viable means by which we can establish the nature of play that the player desires. I am hopeful of a more elegant solution, but perhaps not any time soon.

The fact that different players desire to be in a different place with respect to the degree of challenge connects notably with the model of Flow. The flow channel (depicted), where abilities and challenge are equal, is an area, not a line. When ability outstrips challenge, boredom results, when challenge outstrips ability, anxiety results. Hard agon therefore lies near the top of the flow channel - where the player is close to their limits. Easy agon lies near the bottom of the flow channel - where the player faces just sufficient challenge to entertain. "True agon" lies in the centre of the flow channel. Perhaps finding means to chart a player's position in the flow channel will allow for an automated game balancing system in the future.

Agon is a noble pursuit; the desire to improve one’s abilities and to face others in competition in order to see who will emerge victorious is a motivating force behind all sports and many games. Videogames have expanded the remit of agon to include a wider array of challenges, beyond the central ground of equally matched opponents, and into asymmetries of ability that manipulate the emotions of their players, either through the addictive challenge of fiero, or the trivial escapism of amusement. As tools for entertainment, this blurring of the honourable challenge of agon is essentially harmless in videogames. Furthermore, it provides new opportunities to think about play - even the well-worn play of competition - and perhaps to explore new ways to bring that play to an even wider audience.