Skip to content

Four Critical Modes on Games

Where is ‘games criticism’?

This is a bigger question than it seems at first, with a few different ways it could be posed. With the closing of Electronic Gaming Monthly and the inevitable decline of its print brethren, the question could be about whether authoritative opinion on games is moving permanently from the analog to the digital world; or, as it becomes more and more obvious that the audience for video games is much larger and diverse than previously envisioned, the question becomes much more plaintive: “where is a voice to tell me what games I’ll like?” The proliferation of game-focused bloggers with a specialty in serving specific readerships rather than large ones turns this question into something a little more existential: games criticism is now everywhere, and as a result is it nowhere?

Still another way of asking this question is from the perspective that game criticism is always happening, and as a subject is much larger than the various forms it has taken, is taking, and will take in the future. The question could pre-suppose that all the work that’s being done right now exists within a history that stretches back to maybe not the first game ever made, but certainly the second. After all, what is creation but something of a critique of the past? This implies that perhaps the question posed is really a shortening of the real question at hand:

Where is games criticism right now?

I’ve tried to summarize here what I think are the dominant approaches to game criticism at the present moment. I refer to these approaches as ‘modes’ because we tend to shift between them, often without realizing. This effort is in no way meant to trivialize the work of any individual critic, but simply to give context. The easier it is for us as a critical community to see where we’ve been and identify where we are now, the easier it is to think about where we are going.


This is by far the oldest method of critiquing games and at this point is probably only really discussed aloud among game designers of an old-school variety. While in general Formalism is simply a focus on the inherent elements of games, their rules and play, this approach is most thoroughly articulated in the single concept of ‘elegance’. Elegance in a game has a simple definition: “A small ruleset which gives rise to a very large possibility space”. The single aesthetic imperative here is one of player choice. By this measure the more choices that a player has at any point in the game and the fewer basic rules they have to juggle in their head at a time, the better.

It’s likely that this aesthetic didn’t rise out of any sort of intellectual milieu. In the early part of human society, before the advent of the printing press and widespread literacy, one might propose that the only way a game could spread or pass from one generation to another was if its rules were easy to remember. Likewise, it was unlikely that a person would be able to remember that many games at time, so the game they did know needed to be able to keep them entertained even after many, many plays. These factors add up to the reality that a game was more likely to survive the easier it was to remember and the more times it could be played before becoming boring (or being solved).

How often is this appraoch actually applied to games? It’s hard to say. Its premise is so ingrained in the very mindset of game designers and game critics that it’s basically taken for granted. We take it for granted that a game should present the player with lots of interesting choices. Video games change the rules situation a bit, it’s not necessary to remember them all because the code simply won’t allow you to break them. However it’s still the case that simply operating the game shouldn’t be the most difficult part.

A real boon in using Formalism as a critical standard is that it’s basically universal. Its ideal can be applied to the underlying rules and mechanics of even the most cutscene-laden game.


This is by far the most widespread mode of examining games and is associated primarily with video games. In this view games are products that contain a variety of different kinds of media, such as sound effects, thematic graphics, and story sequences in text and cinematic form, with each element as valuable as every other, including the actual play of the game.

This angle on games is typified by the feature lists used to advertise games, where “stunning cinematic adventure” is a bullet along with “customizable new weapons”. Enthusiast press outlets such as and Game Informer also practice this approach, writing reviews where storylines and game mechanics can be assessed in complete isolation from one another. This mode of critique is present even in more independent voices such as Michael Abbot’s analysis of the character of Jade from Beyond Good and Evil, which eschews any deep discussion of rules or play for an interpretation of the game’s presentation of its heroine.

While Productism definitely reflects the most popular assumptions about games, and to some extent the reality (a ‘game’ in the sense of a product does in fact denote more than simply a collection of rules), there is a drawback.

The primary problem with critiquing games as media products is that it’s difficult to turn around and critique other games by the same standards. Because they must often give equal attention to the typically more glamorous elements of a game, it can be hard for practitioners of this method to appreciate a game where those elements aren’t present. For instance, it’s difficult to assess an abstract puzzle game like Tetris by the same standards you would use to examine a game with 30 minute cut-scenes like Metal Gear Solid 4.

What ends up happening is that you have one set of standards by which you judge media extravaganzas, and another which you apply to more abstract and/or traditional games. Inevitably, coherency suffers.


Unlike the critical modes of Formalism, which we’ve probably inherited from the long history of pre-digital games, or Productism, which is really an attempt to deal with the modern situation and has been practiced for a couple of decades, the critical mode which sees the most important part of a game as the expression of their mechanics has only recently been articulated. Expressed most clearly in an essay by game designer Rod Humble entitled ‘Game Rules as Art’, this mode holds that the real ‘meaning’ in a game should be derived from examination of its rules.

This position takes a critical stance when it maintains that “a game needs nothing else apart from its rules to succeed as a work of art”, according to Humble. In this respect ‘Mechanism’ holds that other elements of a game, not directly affecting the play, are at best ancillary and at worst superfluous; this is in direct opposition to Productism’s claim that all elements are equally valuable. Humble’s essay also implies that it is the ‘message’ of the game’s rules that should be fore-grounded in critical assessments rather than the possibility space that those rules create, which side-steps the rule-to-choice value proposition favored by Formalism.

While Mechanism has been most vocally adopted by the makers of ‘art-games’ (though they’re now gravitating towards a more basic avant-garde aesthetic) it has also been echoed by at least one more mainstream developer. According to his GDC talk ‘I-Fi: Immersive Fidelity in Games’, one of Clint Hocking’s goals with his game Far Cry 2 was to bring the mechanical motivations of the game more in line with its thematic motivations.

Interestingly, if you accept its premise, Mechanism is as universal as Formalism. Supposedly, the mechanics of any game project some kind of message, and the message of any game can be critiqued. This is as true of ancient, abstract games like Go as it is of modern first-person shooters like Far Cry 2.


In some sense this last way of looking at games isn’t about games at all, as they’re traditionally conceived at least. This way of looking at games puts aside one of their most fundamental features: structured goals. Instead what’s emphasized is the extent to which the player can freely explore the game, making and discarding their own goals as they proceed. I’ve included it here mostly because it’s had large impact on both how games are received and how they are designed.

Though we have always had games like Will Wright’s classic, SimCity, which emphasize the exploration of dynamics over the accomplishment of goals, a new crop of games is pushing this premise even further. Part of the draw of the Grand Theft Auto series is the disruption of virtual worlds that actually aren’t that dynamic. What connects the tinkering of ‘God’ games and the trifling of ‘Open-world’ games is an emphasis on the player’s actions that are not connected to specific goals, or sometimes even a set reward schedule.

One could perhaps also connect this to the rise in ‘player-made’ content exhibited mostly recently in LittleBigPlanet and in Halo 3‘s ‘Forge’. This is another avenue in which players can manipulate a game’s structure outside of predetermined goals.

It might be the fact that there is no way to critique these games on their own merit. Instead one applies a version of Formalism to a game that has an enormous possibility space; even if by lacking a goal there’s no way to judge one choice as better than any other. Productism’s belief that even the most trivial visual flare is important could be adapted to find value in virtual mayhem and destruction.

It’s very likely that Playism, as a design aesthetic, will become more prevalent. Perhaps eventually the products that embrace this aesthetic will be identified as something other than games. For now, however, they are recognized as games and their implications must be grappled with.


Conversations about games usually succumb to what might be called the ‘Engineering Bias’.

The Engineering Bias is the often unspoken idea that if we were just smart enough we would figure out how to make the ‘perfect game’.  Writing that comes from this position, consciously or unconsciously, typically features sentences that begin with “Games are supposed to be…” and “Games are really about…”.

The truth is a little fuzzier than that. It may, at some point, turn out that the things we like about our games are in fact fundamental and long-lasting. It’s far more likely though that they’re simply a reflection of our present values; values that will change with time.

A good example of this is pixel art. Once a pillar of game development, pixel art has fallen mostly out of use in the production of mainstream games, and is at this point relegated almost exclusively to handhelds. Yet pixel art has become more beloved as it has passed out of the spotlight. People like Koji Igarashi and Anna Anthropy continue to use pixel art not because it’s their only choice, but for its particular beauty and potential. A similar phenomenon can be observed in the community that has grown up around the creation of new works of Interactive Fiction, the successor genre of the once popular text-based adventure games.

These forms have passed through what seemed like, and sometimes was, necessity and have become aesthetic choices. Perhaps someday when all our games are super-immersive, HUD-less crime dramas there will be a certain charm to a fantasy RPG with 20 minute cutscenes.

With that in mind I want to stress again that the four modes I’ve briefly talked about here are simply what I’ve identified as the dominant critical rhetorics of the moment. They are clearly not the only positions that are possible. For instance, members of the New Games Journalism movement sought to critique games by putting them in the very personal contexts of the writers’ own lives. Richard Terrell over at Critical-Gaming has slowly built up his own critical vocabulary and is using it as a guide for his own experiments in game design.

The hope is that by pointing out some of the lenses through which we examine games we can get a better sense of the still developing critical community. Once we have a sense of where we stand in relation to each other, then it will be easier to refine our approaches and disagree more constructively, which in the end is the real job of critics!


  1. Joe Osborn wrote:

    I really like this — I’m actually about to give a talk on the four categories of game criticism, and I hit upon a similar taxonomy: Commercial Review, Mechanical Analysis, Game Criticism, and Embedded Narrative — in your terms, Productism, Formalism, Mechanism, and Playism. I think your terms are snazzier.

    It also seems like these forms are related and writing can overlap. I considered two axes, something like this diagram (with some examples):

    This way, we can account for stuff like New Games Journalism by situating it at the intersection of Embedded Narrative and Game Criticism, or at the border between Mechanism and Playism.

    Monday, March 2, 2009 at 9:33 pm | Permalink
  2. Joe Osborn wrote:

    Er, that diagram is at

    Monday, March 2, 2009 at 9:33 pm | Permalink
  3. Charles J Pratt wrote:

    Hey Joe, thanks for dropping by!

    It’s nice to know that someone else is mulling things in a similar direction. Now I know there must be something to it!

    I think that the biggest difference between our two approaches is that you’re far more generous than I am. To me there are serious conceptual problems with Productism and Mechanism. Playism, like I said, is barely about games at all.

    The truth is that I’m a Formalist. While I recognize that there’s no single legitimate way to examine games, I think that Formalism is ultimately the most coherent and the widest ranging.

    By the way, I wandered over to your site and read your essay, ‘Parametric Criticism’. I think it’s a great argument for a new way of looking at narrative in games. It made me wonder what you thought of Brenda Brathwaite’s recent post on stories and databases?

    Tuesday, March 3, 2009 at 5:13 pm | Permalink
  4. Noah wrote:

    hey, nice post, charles.

    i think another useful angle on ‘productism’ is that of the value proposition .. the idea that, say, the orange box is a better buy than halo 3 because it contains FOUR games rather than one, or that flower is better somehow for costing $10 on psn rather than $50 in a box.

    Wednesday, March 4, 2009 at 5:23 pm | Permalink
  5. Joe Osborn wrote:

    Wow, I completely missed that Brathwaite post, thanks for mentioning it — the closest I’ve read lately is Steve Gaynor’s “Storymaking” at .

    Thanks for reading my paper — as a quick summary to the rest of the commenters, I believe that the narrative of games is something produced in the player’s head, a synthesis of game art, game music, the player’s mental model of the game mechanics and their inputs, and all the rest. In other words, all these disparate elements are fundamentally just different kinds of inputs to the player’s brain, a perfect machine for authoring stories and filling in blanks.

    I would look at something like the division cited in Brathwaite’s post as an artifact of the coincidence that the game writer and the game designer are the most obvious authors of explicitly meaningful tokens in a game, and if they’re working towards different artistic goals then a divide of some sort is inevitable. However, a game-story divide is probably no different than a game-art divide or a story-music divide — sometimes it can be intentionally used to artistic effect, sometimes it’s an unintentional break in the player’s constructed narrative. It’s a question of cognitive dissonance, right?

    I share your concern about productism, but broad classes of mass-market movies and books are also treated in this way. It may be that we’ll see a Hollywood model of big budget mass-market hits funding smaller indie-style experiments; in such a case, productism is probably an important style of writing, even if it generally sucks to read. If anything, our culture of productism is too informed by the world of software and hardware and not enough by the world of everyday experience. That’s why I feel that you give playism short shrift; if players make stories using games, then those stories are the rewards and memorable aspects of the experience. I might be misunderstanding your definition of playism, but if it amounts to “game travelogue”, I think it could be a way for the enthusiast press to garner legitimacy and “soul” at the expense of shader-worship and polygon fetishism.

    I’m thrilled to chat about this — do you mind if I incorporate these terminology and this conversation into my talk?

    Wednesday, March 4, 2009 at 9:45 pm | Permalink
  6. Joe Osborn wrote:

    Oh, and I forgot to mention Mechanism — I see Mechanism as one of the really interesting schools of game writing, and for better or for worse one that I fall into easily. But I would have some reservations at agreeing that Mechanism ignores the ancillary aspects of a game’s design (presumably by this you mean the artwork, sound design, and story) — if you consider these ancillary meanings as “game mechanics” having a greater or lesser degree of “playability”, then they fit nicely into a mechanistic critique of games. Consider Raph Koster’s example of Tetris re-envisioned as a game about dropping bodies into a mass grave — the art, story, and sound influence the produced narrative, so they are, in some sense, game mechanics. If Far Cry 2 were set in Antarctica with frostbite instead of malaria, it would tell a different story and have a different feel; the political significance of ethically grey African civil war is deeply intertwingled with the core mechanics.

    As an elaboration, every game has an art style; this style is actually a group of rules dictating how things in the game should look and move and animate, and even though it was an artist, not the player, who “played” these rules, they’re still a vital part of the milieu of the game’s narrative events in the player’s mind.

    I would suggest that Formalism, not Mechanism, tends to ignore these less-playable rules.

    Wednesday, March 4, 2009 at 9:52 pm | Permalink
  7. Hi charles,

    I like this post quite a bit.

    I had one question, though: I wonder if you’re overplaying the form-content distinction in your treatment of productionism and mechanism. Maybe it’s the practitioners of these schools that are really at fault for this, but I don’t think even the most doctrinaire mechanist could deny that the contents of the ruleset matter. (For example, it’s important that the squares in Humble’s the marriage are pink and blue squares. the reference to the husband-and-wife relationship is what invests the mechanics with meaning.) I think a sane version of mechanism would just insist that the mediation of that content by mechanics is paramount in interactive art.

    Wednesday, March 4, 2009 at 10:22 pm | Permalink
  8. Joe Osborn wrote:

    Also, when I think of the intersection of Playism and Formalism, I come up with in part because of their peculiar use of metaphor (“If you’re anything like me, you will find yourself — figuratively — filled, many times during each battle, with the strong desire to rip the earbuds out of your ears and lick them, expecting honey to be dripping out.”, ) and occasional pure insight ( in a criticism of the transparency of Fable 2′s Dog as a game mechanic, “We wouldn’t call dogs ‘man’s best friend’ if they weren’t a little bit annoying, or at least mysterious.”).

    Wednesday, March 4, 2009 at 10:58 pm | Permalink
  9. Charles J Pratt wrote:

    Noah – Game pricing hadn’t occurred to me, but I don’t think it would be inappropriate. The basic gist of Productism is that it privileges the media of a game as much as any other aspect, but there’s certainly room for expanding it’s scope to talking about the whole ‘package’, price included of course.

    Joe – Thanks so much for the detailed feedback, you’ve definitely struck on some things that I’m going to have to consider.

    While I think that we agree on some things, I think there are places where we miss each other as well. Like I said I’m a bit of a Formalist and so I’m not entirely comfortable referring to the ‘ancillary elements’ of a game as ‘rules’. To me ‘rules’ are those things which affect the possibility space of a game and ‘play’ is the interaction with that possibility space. On the other side we have code in video games that determines how it looks, sounds, etc., but code and rules are not the same thing. Following that I would say that we ‘consume’ the ancillary elements in games, rather than play them.

    While you’re right that the player’s mind is constantly forming a story and filling blanks, I would say that this is analogous to reader response theory in literature and film, that’s still something that is different from how I’m using the word ‘play’.

    I’d be really flattered if you wanted to incorporate some of this into your talk. Like I said, you’ve brought up some good points that I’m going to mull over, and if you’re interested I would very much like to continue this conversation in the future!

    Iroquois – Always good to have you drop in!

    My impression after reading Humble’s piece, and additional words from Jason Rohrer and Jonathan Blow, is that the Mechanists were pretty strict at one point. I think they’re drifting away from this now though. Hell, Jason Rohrer just released an abstract puzzle game! So you’re right that there isn’t anyone around quite so doctrinaire anymore.

    However, I would point out that my primary complaint about Mechanism still stands, and even a less ridged belief in the expressive power of mechanics is still incorrect. In my opinion mechanics don’t mediate anything; instead of expressing something, they usually are the thing itself!

    Here’s an example: spikes in Mega Man. How do we know that they’re spikes? Not because of the mechanics of the game, but because of the image that is associated with them. It’s the image that’s doing the mediation. What are the mechanics mediating? Danger? Well, no, because they actually are dangerous. Sure, they’re not threatening to player’s actual life, but they’re still deleterious to their progress in the game. It’s a different level of danger, but it’s still danger. It’s not a representation of danger.

    Perhaps, though, I’ve misunderstood how you’re using the word ‘mediation’.

    Anyway, this comment has gotten way too long, but you guys have made some great points and I just wanted to give them their due!

    Thursday, March 5, 2009 at 6:25 am | Permalink
  10. Joe Osborn wrote:

    I think the key in the spike example is that it’s the intersection of two rules: First, that “terrain can instantly kill you”, and second, that “terrain that kills you will look spiky”. The second rule is used by the player to determine the appropriate action given the first rule. So, I think the line is much fuzzier than a pure formalist might suspect — rules that only concern the visual elements _do_ influence player behavior, whether that’s a glowing aura around an item pickup or an especially spiky-looking floor tile.

    The key difference between reader response theory and interactive media is that, while the player is integral in both cases, the player’s role in the combination of elements in a game is much more direct and total — UConn’s Roger Travis ( ) calls games and the activities of Homeric bards “performative play practices”. Most recently, he described Bioshock’s climax by saying that “the moment of having to kill Andrew Ryan makes sense only in contrast to the interactivity the player has been allowed to enjoy elsewhere in the game”. This doesn’t have a clear analog in reader response theory, right? Even if a film was very open to interpretation in some places and closed in others for contrast, it wouldn’t have the same impact.

    Friday, March 6, 2009 at 1:24 am | Permalink
  11. Charles J Pratt wrote:

    I definitely agree with you about reader response theory. I think that the interaction that one has with a game is a very different type of participation. Games traffic in interactivity, and just like silence is an important part of music so non-interactive segments of games gain power only in relation to those parts the player had control over. I’m familiar with Mr. Travis’ work and I like it a lot, it reminds me of Markku Eskelinen’s idea of configurative performance, only with much more comfortable with the narrative implications.

    As far as the spike example goes, I think you’re definitely right that the visual element of a game can have an effect on the way it’s played, and it is important to signal players in different ways the behavior of the system (if only to prevent them from having to learn by trial and error). However, I’m afraid I don’t agree that this means that the line between a rule and the image associated with it is fuzzy. While it’s true that the picture of a spike communicates danger, it doesn’t actually have any effect on whether or not the terrain is dangerous. You could replace the spikes with apples and that terrain would still kill you instantly. Conversely, you could step on spikes and they could do no harm whatsoever, and then you would know that you didn’t have to worry about them.

    The point I was trying to make is that the image of the spikes represents danger, while the rule that determines which terrain instantly kills the player is creating actual danger. This is what makes the latter a rule and the former simply a sign.

    Now, I don’t want to give the impression that I think that the representational layer of a game isn’t important, I very much think it is. I derive a great deal of pleasure from beautiful games and I would be just as pissed as the next guy if I touched an apple in a game and suddenly exploded. I just think that it’s important not to confuse the content of a game with its mechanics, because to me they are very separate things.

    Friday, March 6, 2009 at 5:21 am | Permalink

Post a Comment

Your email is never published nor shared.

Comment spam protected by SpamBam