Game designer Greg Costikyan is most famous for his work on the award-winning Paranoia tabletop role-playing game (1984), pictured left, and the valiant-but-doomed indie publisher Manifesto Games. In 1994 he published an article in the British role-playing journal Interactive Fantasy that was expressly intended as a rebuttal to Crawford's earlier presentation of his terminology in The Art of Computer Game Design (1984). Entitled “I Have No Words & I Must Design” (and available online at Costikyan's website), Costikyan's express goal is to move towards a common critical language – a project as urgently needed today as it was almost two decades ago. He presents the following definition:

A game is a form of art in which participants, termed players, make decisions in order to manage resources through game tokens in the pursuit of a goal.

Notice immediately that Costikyan wants to assert games as art, contra Crawford's distinction – this is an aesthetic value judgement, but about art rather than games, per se, and thus lies somewhat outside the scope of this discussion. He also adds the following remark:

What's key here? Goals. Opposition. Resource management. Information.

Despite the goal to push against Crawford's approach, Costikyan's definition is ultimately quite similar in its component aesthetic value judgements – although it does not create a hierarchical assemblage of priority between them, and in his subsidiary remarks Costikyan suggests a great deal of additional elements that "strengthen games" (we'll return to these points shortly, as they are highly relevant). Nonetheless, Costikyan's key constitutive elements of games embody two of the same value judgements as Crawford: the victory aesthetic (goals) and the conflict aesthetic (opposition). What Costikyan adds is "resource management and information", which are presented in the context of the player's need to make decisions. This corresponds to Sid Meier's aesthetic assertion that "a good game is a series of interesting choices", and can be understood as a decision aesthetic. There is some potential overlap with the previously mentioned agency aesthetic, but it may be prudent to treat these as different – in particular because agency tends to concern the fiction of the game, whereas decisions tend to concern the rules of the game (although the blurring between these two is unavoidable).

However, decision-making is a subset of problem-solving tasks that certain players relate to the concept of a game, and thus the decision aesthetic is part of a wider aesthetic of play. As Costikyan remarks in his countering of Crawford's suggestion that Zork (1977-79) is a puzzle and not a game:

Almost every game has some degree of puzzle-solving; even a pure military strategy game requires players to, e.g., solve the puzzle of making an optimum attack at this point with these units. In fact, if a game involves any kind of decision making, or trade-offs between different kinds of resources, people will treat these as "puzzle elements," trying to devise optimal solutions. Even in deathmatch play of a first-person shooter, players will seek to use cover and terrain for advantage – 'solving the puzzle' posed by the current positions of opponents and the nature of the surrounding environment, if you will. You can't extract puzzle from game entirely.

Thus even though Costikyan himself seems to favour putting emphasis on the decision process over the wider activity of puzzle-solving, there is a more general problem aesthetic that can be recognised – one that views all play as puzzles or problems to be resolved. The overlap with agency is somewhat obscured when the focus becomes problem-solving rather than decision-making: we can make a decision about what colour car to drive but this is not a problem that can be solved, except in the most contrived use of the word 'problem'.

Should Costikyan's flagging of resource management be taken as part of problem(or decision) aesthetics, or as a separate assertion? This is unclear. Focussing on resources invites the kinds of decisions that correspond generally to the problem aesthetic (i.e. the activation of the orbital-frontal cortex and the consequent release of the reward chemical dopamine). Yet decisions are also possible in other wide senses, and these rapidly converge on the aforementioned agency aesthetic. Perhaps these two are in some kind of tension – between the formal, rules-focussed slant of the resource decision or problem-solving process at one end and the informal, fiction-focussed bent that is often entailed in attributions of agency to a game. These problems will have to pursued at other times using different methods. However, it is quite plausible that what is being alluded to here is close to a systems aesthetic that we will see later with Raph Koster's and Dan Cook's approaches to defining games.

Crawford, responding to the places where Costikyan's definition diverges from his own suggests that the reference to tokens sounds like "a throwback to the days of boardgames" and the reference to resource management reflects "a strategy gamer's approach to the problem", although he does not ultimately find either problematic. It is interesting, however, that Crawford uses the term 'throwback', which involves a value judgement – as if Crawford believes that now videogames exist there is no place for boardgames. A great many boardgamers – including Costikyan and myself – would vehemently disagree with this assertion. Thus if Crawford is making a value judgement against a boardgame-inspired term, there seems to be a case that Costikyan is also advocating a tabletop aesthetic. This might amount to nothing more than the belief that the fundamental systems used in tabletop games are compatible with their counterparts in videogames, and this may in fact be a subtle variation on the decision aesthetic. Still, a tension between a videogame aesthetic and a tabletop aesthetic that prioritises one form of play over the other may be present in this kind of discourse. I will not dwell on these media-specific aesthetics, but it should be recognised that both these and genre-specific aesthetic stances can be found in most discussions about games.

Costikyan's approach becomes far wider when he looks at those things that "strengthen games". Although these do not form part of his definition of 'game', they still embed aesthetic value judgements. Costikyan draws attention to diplomacy, colour (meaning the fictional gloss layered over the raw game rules), simulation, variety of encounter, position identification, role-playing, socialisation, and narrative tension. These broadly divide into general aesthetic positions. Firstly, those game-strengthening possibilities that specifically concern the way that players interact with each other beyond conflict – diplomacy and socialisation in Costikyan's terms. This could be termed a social aesthetic (diplomacy, socialisation), and it is worth noting that Costikyan (like Crawford) wishes to make the conflict requirement necessary and the social aesthetic tangential. It would not be difficult to defend the social aesthetic over the conflict aesthetic – co-operative games can often be expressed in terms of conflict, but such arguments are generally sophistic. Like Crawford, Costikyan still has some hierarchical judgements on games embedded in his approach.

Secondly, those game-strengthening possibilities that rely on the fiction or the player's imaginative faculties – specifically colour, position identification and role-playing. This imaginative aesthetic can be found elsewhere: in Imaginary Games I go some way towards defending an aesthetic value judgement that places this at the root of all of our experiences of play, for instance, which reflects a long and slow transformation of my beliefs about games over the last decade. Costikyan's simulation probably also belongs in this category – he notes that "simulation is a way of providing color" and also that "it improves character identification", which seems consistent with this approach. Lastly, Costikyan asserts that games can be strengthened by variety of encounter and narrative tension, which are forms of an uncertainty aesthetic that we'll examine more closely later.

What is striking about Costikyan's approach, even allowing for the apparent precedence given to the kinds of aesthetic value judgements also found in Crawford, is that it is so wide reaching. Definition aside, Costikyan allows for an incredibly diverse range of aesthetics of play within his framework. He still wishes to mount his priorities in similar ways to Crawford – namely the prioritisation of the victory and the conflict aesthetic. There may be a psychological or neurobiological explanation for this in terms of gender, or more specifically, in terms of testosterone. This chemical is associated informally with the male gender, since it is the male sexual hormone and governs the development of male sexual traits, but it affects both genders in the same essential ways and women with high testosterone levels display the same psychological traits as high testosterone men. It is a hypothesis I am currently investigating as to whether the psychological implications of higher testosterone levels in terms of persistence, tolerance to frustration and consequent enjoyment of conflict will serve as an explanation for why certain game designers wish to single out victory and conflict as key to games, but it is my strong intuition that these two aesthetic positions are related to testosterone in some way.

Next week: Koster's Theory of Fun

Whatever we consider games to be, people have incredibly strong opinions about them. I previously claimed that a great many attempts to define games could be interpreted as value judgements asserting a particular aesthetic stance, and in the series of pieces that follow I'll examine a number of definitions of the term 'game' and excavate the underlying aesthetic judgements. This is clearest wherever a definite boundary is being erected, and will become difficult when more general claims are made. There is ample room for many objections to my conclusions, not least of which being the need for further research, but it is not a plausible retort that one of the aesthetics that follows is the correct way of interpreting the term 'game' – the very nature of this approach voids any attempt to establish a single 'correct' definition. As Wittgenstein remarked, the meaning of a word is how it is used, and all of these aesthetics relate to ways the word 'game' is used by different people. We may prefer some of these uses to others, but this simply confirms that definitions of games involve value judgements.

At the start I shall take as essentially unproblematic only two conditions used within game definitions: that games entail rules, and that games entail fiction (or, in the terms discussed in my introduction to game aesthetics, that games have both functional and representational aspects). Both these apparently foundational assertions are sufficiently general that when taken singularly they exclude very few things people would consider as games, and indeed, include many things that a great many people wouldn't consider a game. Note that even play activities that have no stated rules can still be formulated in terms of rules – when I play Fetch with my dog, I consider this a game, but my Labrador probably does not conceive of what's happening in terms of the rules I can justifiably use to describe our game. Similarly, the fiction in (say) Chess is very thin indeed, but I don't think it's helpful to claim there is none at all. At the very least, Salen and Zimmerman's Huizinga-inspired "Magic Circle" sets apart the activity of the game from other activities, however porous this condition might be, and in this sense allows us to recognise a fictional world. Besides, we will later kick these blocks away in order to be left with a complete (but rough!) model of the aesthetics of play.

Because their stances are more strident and therefore more interesting, I shall start with the definitions of 'game' provided by professional game designers, and later consider the more moderate positions advanced by academics. It is the nature of this project to approach definitions of games on the assumption that they embed value judgements, but it is not necessarily the case that a proponent of a particular definition endorses the corresponding aesthetic judgement (although for the most part I believe they would, albeit with some considerable caveats). Nonetheless, in terms of the process of uncovering implicit aesthetic positions on play, each definition is archetypal of a wider viewpoint that is been reflected in my player studies over many thousands of players. The individual proponents of definitions, therefore, are akin to spokespeople for a particular aesthetic archetype and I apologise in advance to everyone involved for any violence I may have done to your specific views through the process of excavation.

To begin with, there is no better example of my claim that definitions of game conceal aesthetic value judgements than the famous taxonomy of creative expression given by Chris Crawford (2003), pictured above. Crawford is keen to note that his terminology cannot be considered definitive, but the nature of his definitions are so rigorous that they bring most out most clearly the kind of aesthetic judgements this exploration is aimed at uncovering. His taxonomy is constructed as a series of questions, each of which divides activities into two classes (sometimes with a subjective judgement aspect that Crawford is forthright in recognising). 'Game' is an accolade attained only by passing through each of the "gates" Crawford has set up, which implies value judgements that at other times in the same book he appears keen to disavow.

The sequence proceeds according to the following outline. Firstly, the motive of the creator is considered. If it is beauty, Crawford terms it 'art'; if it is money, he terms it 'entertainment'. Since Crawford places 'games' under 'entertainment', games seem to be immediately excluded from art, despite Crawford elsewhere (1984) yearning for games as an art-form. However, he also allows that this first step could easily be omitted. The rest of the steps are as follows:

• Within entertainment: "is it interactive?" – if no, it is in the same (unnamed) class as movies, books and films, else it qualifies as a 'plaything'.
• Within playthings: "is there a defined goal?" – if no, a 'toy'; if yes, a 'challenge'.
• Within challenges: "is there an agent to compete against (or the illusion of one)?" – if no, a 'puzzle'; if yes, a 'conflict'.
• Within conflicts: "can you impede your opponents?" – if no, a 'competition'; if yes, a 'game'.

Thus, only those things that are interactive, have goals, include opponents and allow attacks against those opponents qualify as a 'game' in Crawford's terms. However, simply listing these conditions doesn't do justice to the value judgements implicit in Crawford's taxonomy, since by placing 'game' at the apex of this sequence there is an unavoidable sense of priority to this order. Note that if the result of the final gate had been a different term – say, 'battle' – the entire taxonomy could be seen to refer to the space of play and games. By choosing to give 'game' this highly restricted meaning, Crawford appears to not only assert a number of aesthetic value judgements concerning games but also a hierarchy of aesthetics of play. The steps correspond broadly to an agency aesthetic (interactivity), a victory aesthetic (goals and challenges), and a conflict aesthetic (opponents and direct attacks) – all three of which can be found expressed by other game designers, although rarely in so rigid a sequence. Furthermore, other game designers do not tend to give so much weight to conflict: Crawford devotes two entire steps of his process to distinctions within conflicts, making direct competition a requirement of games rather than a class of games. This oddly groups games such as Snakes and Ladders, golf and all golf-simulations, Ubongo, Race for the Galaxy, and Knizia's Fits and its descendents, as being competitions rather than games, despite most people's strong intuitions that at least some of these are indeed games.

Over the next few weeks, we'll contrast the aesthetic value judgements found in Crawford's taxonomy with those found in other game designers (namely Greg Costikyan, Raph Koster, Dan Cook and Jane McGonigal) as well as those found among academics with an interest in play and games (Bernard Suits, Roger Caillois, and Thomas Malaby) in the hope of creating a very rough map to the landscape of the aesthetics of play.

Next week: Costikyan's Critical Language

I will be submitting a revised and edited version of the Implicit Game Aesthetics serial that starts today as a paper to an as-yet-undetermined journal or conference. Please help me decide where to publish it as I am really quite unsure!

Sorry, no new material this week as I’ve been too busy. I’m working on something big, though, so please be patient. In the meantime, let me draw attention to a discussion in the comments instead. Responding to last month’s post on Journey, Jose Zagal remarked:

"Something is seriously wrong with any creative medium that manages to so radically block its own potential."

For some reason I had a double take when I got to this part. It looks like you're equating creative medium with the industry, and in particular with a certain portion of the industry. While I'm grateful that Sony has supported ThatGameCompany, I think it's a bit unfair to argue that "the medium" is the one that's blocking any potential. I think at this point in time, we have more to blame ourselves (the consumers) than anything else (e.g. the recent "we don't like the ending of Mass Effect 3" campaign). The platform holders have lost a lot of that power, the market has grown and changed (allowing for much more diversity), we now have more reasoned/critical/informed discourse about games than ever before.

To which I replied:

Jose: I appreciate your counter-argument here, but I'm not at all convinced that the industry can be absolved of any responsibility in this matter. Other creative industries - including books, films, television, music - all support lively niche markets for artistically-motivated works. In books and films, best-selling content is produced without having to step outside of the commercial sector.

In games, however, the industrial sector has been mono-maniacally obsessed with farming teenage boys to the extent that the large media corporations with a stake in games were so far removed from a fair understanding of the audience that they allowed the mass market sector (currently serviced by social gaming companies, Nintendo and Apple) to lie fallow for decades. This was pure commercial incompetence, and I would argue that the failure to invest in arthouse games is exactly the same kind of manifest incompetence.

We could have a lively, artistically-motivated game development scene at a commercially relevant scale if each of the big publishers invested 1% of their marketing budgets into creating such a space. But they have no interest in doing so, and it can't simply be claimed to be the audience's responsibility that the publisher's are too cowardly to invest in the potential of the medium. That many such investments would fail to turn a profit is no excuse since this is true of all videogame projects - including the mountains of dross that are pursued in the often mistaken belief that the publisher is funding a commercially valid project. And that's not to mention some perfectly reasonable game projects that fail simply because they were allowed to acquire excessive budgets (LA Noire springs to mind).

So yes, I do believe it's the industry's fault that the creative potential of games is under-developed. The responsibility doesn't end there, but I certainly believe that's where it starts.

Click here to add your own comment into this discussion.

The puzzle game Clusterpuck, designed by International Hobo and developed by Codename, arrived on PlayStation Home yesterday! The concept and design was developed by Joel Atkinson, under my ever-watchful eye – but neither of us have actually played it yet! During development, Codename sent us videos of player trials and that was our only point of reference for the game.

We’d love to know what people make of Clusterpuck, so please leave a comment if you’ve played it.

Talk about games in recent years has been dominated by the distinctions between the business models for games-as-products and games-as-services. But it may be helpful to also consider the purchase model that players have in their head when the approach a game: is it buy-to-own, or is it pay-to-play?

When I first met with Capcom, back in 2000 (not long after founding International Hobo), they gave a presentation on their business strategy: franchises. Capcom did not develop and release any game if they did not consider it had the potential to produce sequels and spin-offs. At the same time, Ernest Adams was working for EA, and he later informed me that EA had more or less the same approach: “We don’t make games, we make brands” was the stated policy. Even before the formal games-as-service business model was in place, publishers were already operating on the basis of repeat business – a kind of product-as-service model that relied on getting players to buy into a brand, and to keep playing and buying in the future.

Videogames arrived into an entertainment landscape that already recognised that if the market was to bear more than one company it would be necessary to get players to buy more than just one game box. Dungeons & Dragons, first published in 1974, had two expansions the following year, and although it was not the first game to sell expansions it was far and away the most successful at doing so. This extensible approach to games had no precedent prior to the wargames of the 1970s: the purchaser of a copy of Parker Brother’s Monopoly in 1935 was expecting to get a box containing everything they would need to play, and the idea that they might spend more money later on additional components would have been very difficult to accept. After D&D, however, all tabletop hobby games aimed for expansions: Eon games’ Cosmic Encounter, for instance, which had been turned down by Parker Brothers, published nine expansion sets over the five years following its publication in 1977.

Cosmic Encounter (pictured above) was ultimately to achieve immortality by being the inspiration for the original trading card game, Magic: The Gathering, which arrived in 1993. Again, the line between game and service is blurred here. Players bought a starter deck (compare buying the base game for World of Warcraft), then needed to keep buying booster packs until they had all the cards they wanted, or all the cards in the set. A few months later a new expansion set would appear, and players would continue being farmed for money. This made Wizards of the Coast a lot of money, and created a lot of competitors – almost all of which failed miserably, often simply because they lacked the property that was originally the design specification for M:TG: a short play time. The revenue model this and other trading card games used was not dissimilar to a lot of contemporary digital game services – it was essentially a microtransaction model – although most of those give away a lot more free content than any trading card game ever considered.

In many respects, the microtransaction model for games originates even earlier, in the arcade. Early videogames like Space Invaders (1978) and Pac-man (1980) made their money own quarter at a time, and indeed did very well out of this. However, the business model was different in an important way to contemporary game services. Arcades bought cabinets for a fixed price, then recouped costs from coin drops until they ultimately turned a profit. Excess coin drops beyond what was needed to break even made more money for the arcade owner, but not for the manufacturer of the cabinet whose only benefit was in the increased demand for cabinets successful titles would create. A contemporary game service cuts out the middle man and puts the revenue straight into the pocket of the developer – with a significant cut for whichever social service provider (e.g. Facebook) is hosting the service.

The product model of selling a game in a box (whether a tabletop game or a videogame) was necessary prior to the incredible transformation of the information infrastructure represented by the internet. It necessitated getting the game as good as it could be prior to release, and marketing heavily when the title hit the shelves. Success was measured not just by sales, but by the ability to sell subsequent content with the same branding (whether as expansion packs or as sequels). Although delivered and priced as products, there is still a sense that these games were offered as a kind of service in so much as the players they were attracting were always intended to be repeat consumers. The service charges may have been spaced out – and left to the players to determine when to buy – but repeat revenue was always the objective. Many tabletop players, myself included, poured quite a lot of money into Steve Jackson’s Car Wars (1980) over the decades following its release, because it was well supported by expansions and we wanted this extra content.

In videogames, resistance to moving over to an overt service model often rests on objections about ownership. A player who buys a game in a box owns a game. A player who plays a game service often owns nothing but data on a server, and there are questions as to whether they even own this. There is no possibility of resale (which in itself should be a reason for lower price points) and there is no choice of price point, as there always is in boxed product, which comes down in value the longer it has been in  low demand and remains unsold. Players of games-as-service simply accept the pricing model offered to them – much as players of arcade games had to do previously (although different arcades often priced differently in that case). Players of games-as-product always had more options in terms of what to pay and when to pay it.

The resistance to the service model has an echo in tabletop. Successful hobby game companies such as Fantasy Flight have business models that are essentially the same as those described above for Capcom and EA. Both Arkham Horror and Descent: Journeys in the Dark are boardgames designed to operate as sources of long term revenue generation. My wife and I have already bought seven boxes of Arkham Horror material, with two more still to purchase sometime in the future, and this is not an uncommon situation with the players of these kinds of games. Problems frequently arise among the community of players when the original product line is to be replaced with a new edition – Descent seems to have produced this issue for some of its fans who are unhappy about a new edition, feeling more explicitly that they are being raked for cash. Role-playing game publisher White Wolf faced similar fanbase ire when they entirely rebooted their cult World of Darkness product line.

The difficulty some players have in accepting games-as-service is rooted in part on responses to game ownership that originate with tabletop games. “I buy the game”, the player feels, “I own the game, my game”. In the service model, this becomes “I play the game, I pay for the game” – there’s never a sense of it being their game, a situation directly parallel to the arcades. As well as business models operated by the providers of games, we can therefore recognise purchase models believed by their players: the buy-to-own model is psychologically very different from the pay-to-play model, even though some games blur the lines. Magic: The Gathering may feel like buy-to-own, but in fact the economics end up being pay-to-play – it’s a game crypto-service, a covert game-as-service masquerading as a product. A similar argument could be made for Modern Warfare, with its regular new edition being an essential purchase for its players who are effectively paying an annual fee to continue playing this particular crypto-service.

Player resistance is strongest where the buyer model clashes with the business model most overtly – hence, paying for DLC that uses material that ships on the retail disc produces considerable gnashing of teeth. The boxed game is thought about on a buy-to-own model – the idea that you don’t own everything you buy is a source of anger. However, from the developer’s perspective they need money to keep making games and DLC is simply an extension of the expansion model that dates back to D&D. Nobody complained when Aliens: The Boardgame shipped with extra pieces that were needed for its expansion since it was clear this was beneficial for the player: if it had not, the expansion would have also needed a box and would have been more expensive. The example of shipping some data resources for a DLC on the disk is parallel in most cases.

As I implied in the opening, videogame publishers traditionally associated with boxed products were always interested in repeat custom, because no-one is in business to make less money when they can make more. Everyone needs money to live, after all, and game developers (despite what some players seem to believe!) are no different. The question facing many game publishers and developers now is: will we benefit from moving over to a service model, and can we make the transition? In some cases, the service model will make more sense – sports games like Madden with an annual update no longer make sense as boxed products, as Rik Newman has noted in the comments here. Players will be resistant to the change whenever they are still thinking in terms of buy-to-own, and especially when what the publisher offers seems to be a move towards pay-to-play. It remains to be seen which kinds of games can successfully make this transition, which is less about switching from product to service than it is about changing how particular game franchises are monetised.