Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 31, 2014
arrowPress Releases
October 31, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Designing Mechanical Narratives (Part 3)
by Taekwan Kim on 04/12/13 09:13:00 am   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Note to those that have been following this series:

I’ve taken quite some time with this final entry. The truth is that pretty much all of it had already been written some time ago. But in the writing of it, I came to the conclusion that the shoehorning in as examples of some games I had made was exactly that—it was distracting from the central topic I had set out to examine. And once I began to question my motives, it became increasingly difficult to get myself to even post the dang thing.

I’ve rather failed thus far to establish what a mechanical narrative even is, why it should matter that we have such a specific and contained concept in our toolbox of ideas—the whole point of this series. And trying to define it by using such obscure examples just wasn’t working. The examples themselves needed too much explaining, significantly muddling the idea of the mechanical narrative.

To be fair to myself, some of the topics I wanted to address by bringing up those examples were and are genuinely useful here—particularly the observation that the win conditions of a game, and whether they even exist or matter to the player, greatly impacts the player’s mechanical experience and mechanical narrative. (Without trying to sound too ridiculous, the question, “What is winning?” is in fact quite practical and informative when considering mechanical narratives.) At any rate, though I may at some point return and talk about those games, I’ve decided to conclude the series with the following post, without further mention of them (as originally promised).

My apologies if I have kept anyone waiting.

The Difference between Mechanical and Procedural Narratives

Commenter D F mentioned on the last post that the types of narratives we are talking about should not be composed of narrative “events”, that they should be systemic and generated through process. I am actually in agreement with this (or rather, would say that they simply are, as opposed to should be), and am particularly thankful for the comment bringing up the term “procedural narrative”.

The distinct entity I have been trying to describe in this series (the mechanical narrative) is in fact different from procedural narratives. Traditionally, when we say procedural narrative, we are referring to narratives that have randomized elements, where the elements themselves (and usually the general plot parts of the story) have already been prefabricated. The “procedure” part is the algorithm used to link these elements together.

As such, procedural narratives are still about diegetic actors (e.g. non-player characters, the avatar/player character, etc.) and plot elements, whereas in mechanical narratives, the actors are the actual player(s) and the game itself. In a procedural narrative, the algorithm or “rules” are used to generate a story. In a mechanical narrative, the rules themselves essentially are the story.

Let’s back up a bit. What the heck is a “diegetic actor”? Briefly, the difference between a diegetic and non-diegetic actor is more or less the same as the difference between the player character and the player. We can dig into this a bit further with Wikipedia’s article on diegesis in film:

The classical distinction between the diegetic mode and the mimetic mode relate to the difference between the epos (or epic poetry) and drama. The "epos" relates stories by telling them through narration, while drama enacts stories through direct embodiment (showing) [my emphases].

In terms of classical poetics, the cinema is an epic form that utilizes dramatic elements; this is determined by the technologies of the camera and editing. Even in a spatially and temporally continuous scene (mimicking the theatrical situation, as it were), the camera chooses where to look for us. In a similar way, editing causes us to jump from one place (and/or time) to another, whether it be somewhere else in the room, or across town. This jump is a form of narration; it is as if a narrator whispers to us: "meanwhile, on the other side of the forest".

It is for this reason that the "story-world" in cinema is referred to as "diegetic"; elements that belong to the film's narrative world are diegetic elements. This is why, in the cinema, we may refer to the film's diegetic world.

To make a poor analogy via film, we can say that traditional and procedural narratives in games are like the stories contained within the content of movies—they exist as the game’s diegetic world. In this analogy, mechanical narratives, then, would describe the processes the viewer of a movie experiences while interpreting the film’s narration—the story told by examining how the various camera and editing techniques alter the viewer’s understanding of the narration.

Again, this is a rather terrible analogy, but the point is that a mechanical narrative’s story is not within the content, but within the dialectic between the player and the device.

Designing a mechanical narrative, then, is about crafting experiential possibilities that become understood as a “story”, as opposed to crafting a story that is then experienced. The confusion arises in that the experience of experiencing a dynamic or interactive story itself can produce a mechanical narrative. The story structure in games, after all, is also a set of rules to be followed.

Even more confusingly, rules can be both diegetic and non-diegetic at the same time. This, of course, is because the player character is the diegetic version of the player. And there are rules that both the player and the player character abide by, if ultimately for different reasons. (This is what I was partially referring to when I wrote in Part 2 that verbs have both narrative and mechanical definitions.)

Diegesis in Left 4 Dead

For instance, does Left 4 Dead produce procedural narratives or mechanical ones? Well, the answer is both, and simultaneously. But the difference between the experience of the player and the experience of the player character should be clear.

A low-health player character is not making a mechanical decision when it self-sacrificingly gives some pills to a fellow survivor, while the player could very well be making a purely mechanical decision based on his assessment of their respective skills or foreknowledge of upcoming health caches. Conversely, the player could be forgoing all mechanical considerations and likelihoods in order to play along with the procedural story—to assimilate with the player character’s motivations. A Versus game of L4D skews mechanical because the goal is no longer survival of the player characters, but point taking; the reality of the gameworld takes a far backseat.

To develop the example further, a player that decides to leave the extraction point in the finale to help out a downed player in the face of an oncoming tank and horde is either making a diegetic decision to “leave no man behind” or an extra-diegetic one to see if the player can succeed the challenge against the odds (or both). A player that extracts by himself when all other team members have been incapacitated is making an extra-diegetic decision based on the rule that the campaign will successfully finish if just one person gets on board (to avoid repeating the finale, and potentially failing entirely), or a non-diegetic decision that self-interested “winning” is more important than teamplay.

The labeling of these motivations is actually subject to change depending on the player. The decision to “leave no man behind” can also be non-diegetic—it can be an external principle and self-imposed rule the player follows anyway, in any situation, without any prompting from the game or its environment and reality. Similarly, one suspects that the diegetic characters of L4D would rather see that a fellow survivor escapes than remains behind to die for a hopeless cause, so the lone survivor scenario could conceivably be an act of extreme diegetic roleplay.

At any rate, a mechanical narrative, isn’t really about retelling of the player’s mechanical decisions per se. It is more about the player managing the tensions between the game’s diegetic reality and the player’s non- and/or extra- diegetic goals—of how the play was determined, and not so much the play itself. I would even go so far as to argue that mechanical narratives are only possible when the distinction between these motivated systems can be felt, through either involuntary difference or deliberate, conscious, voluntary assimilation.

Donald Rumsfeld and Squash

As an extreme (perhaps slightly too political, and rather insensitive) example of how the interaction between various levels of diegesis in gameplay can shape a narrative, let’s consider former Defense Secretary Donald Rumsfeld and the game of squash.

Mr. Rumsfeld has testified that “his ideas about transforming the military into a smaller, more agile force, like the one he pushed for in invading Iraq, were influenced by his squash playing.” The above linked article describes Mr. Rumsfeld’s play as follows:

Rather than tricky bank shots off the walls, a move that better-skilled players favor, Mr. Rumsfeld plays with power, hitting the ball hard and ending points quickly. And he relentlessly attacks his opponent’s confidence.

Mr. Rumsfeld himself noted, “When I pass [the opponent] in a shot and it’s a well-played hard shot, I saw speed kills. And it does. If you can do something very fast you can get your job done and save a lot of lives”. This was his mechanical narrative.

Unfortunately for this case, what we can see in this narrative is a need to create conditions that make power shots more favorable than bank shots. Which is to say, instead of bending his non-diegetic goals, Mr. Rumsfeld insisted that extra-diegetic structures be modified or imposed to suit his non-diegetic needs. A few examples (again, from the article):

- Mr. Zimmerman has never actually played his boss. But he says he has noticed that Mr. Rumsfeld, 74, often wins points because, after hitting a shot, he does not get out of the way so his opponent has a chance to return the ball, a practice known in squash as “clearing.”

- [Despite this, and given that interference/obstructions are inevitable in squash] Mr. Di Rita [a former regular partner] concedes that Mr. Rumsfeld rarely offers or asks for “lets,” a replay point when one player feels aggrieved.

- In squash, Mr. Rumsfeld’s main advantage over more capable and fitter players is that he refuses to play anything but “hardball,” a variation of the game that was once common in the United States but has largely died out over the last decade. Most competitors now play the international game, which uses a softer ball and a wider court, requiring more running and allowing more creative shot-making. The harder ball favored by Mr. Rumsfeld tends to come back to the center of the court, so players do not have to move as much to return it.

This history, then, led to a series of decisions which reflected the expectation that “diegetic” reality would meet a set of “extra-diegetic” rules and impositions that, in this case, ultimately only held true for the person dictating them, for the specific conditions under which they were conceived.

The United States had a “short-war strategy,” [retired Army general] Keane complained to Rumsfeld in September 2006. “Nowhere in it is there a plan to defeat the insurgency ourselves.” Changing course, he explained, would require as many as 28,000 more troops. Keane describes Rumsfeld as shocked at the news, “as if he was hearing it for the first time.” American forces, it turned out, were a bit too flexible [my emphasis]. They were configured in such a way that they could pack up and leave anytime, and that was not lost on the insurgents, or on bystanders picking sides.

As we now know, it was the “surge”, the vast increase of troops on the ground in Iraq, which eventually brought that war to a close. (This is a tremendously simplified narrative. It’s also debatable as to how much direct influence the Secretary of Defense even has over troop size and structure.)

From this example, we can see that it is the contingencies created by the entanglements of all three diegetic, non-, and extra-, which produce mechanical narratives. The diegetic actualities of the sport of squash (the need to move around the court, etc.) clash with Mr. Rumsfeld’s non-diegetic realities of being 74 years old but desirous of beating much younger, and in some cases better, opponents, which leads to extra-diegetic choices of hardballing and non-clearing. This produces the mechanical narrative that “speed kills”. The narrative is true because of the way the rules have been played.

Later on, the “non-diegetic” goals of speed and flexibility (also partially motivated by “diegetic” domestic political realities, though one must remember the strong support for the war by the American public at the time) lead to “extra-diegetic” dictations on troop structures and strategies which in the end fail take into consideration certain necessities of the “diegetic” real world. This produces the mechanical narrative that a military invasion that seeks to establish complete politico-cultural regime change can’t be “won” through “flexibility” and “speed”—the tragic narrative that the apparently truthful lessons of squash just did not apply in this case. This narrative is true because of the way the rules must be played.

In this extended analogy, the former narrative would be a “player-driven” one, whereas the latter would be “designer-driven”.

I do not mean, here, to trivialize war as an impersonal “game”; the last thing I would want to do is to make light of the death of so many, and I sincerely apologize if this analysis has caused any offense or hurt. What I am saying, however, is that the degree to which mechanical narratives can shape our expectations and thinking can be devastatingly [or gloriously] real.

The very thing that can make them feel so victorious can cause them to be blinding, particularly negatively so when we fail to realize that the “game”—any object upon which we are projecting our mechanical narratives—has changed, or was never what we thought it to be.

Designer Driven Mechanical Narratives

I had originally intended to title this series with the above header. The idea was to focus on the fact that mechanical narratives can be designer defined, and that this has nothing whatsoever to do with dictated “narration”. It is the degree to which diegetic or extra-diegetic rules shape, impose themselves upon, or eclipse the player’s non- or extra- diegetic goals that determines the amount of designer intent entering into the player’s experience. But, as I hope the above examples have made clear, these distinctions are actually quite fluid, and both involuntary and voluntary assimilation can generate strong narratives.

At any rate, the point once again is that we don’t have to leave the strength of these narratives up to chance, and being deliberate doesn’t necessarily mean “non-dynamic” or “non-linear”. We can be methodical in their construction simply by careful, considered crafting of the rules. By studying the implicit demands of the mechanics, and the range of possible motivations the player might approach the game with—and by expressly designing for, or designing in, the tensions between them—we can do much to insure that an engaging narrative arises, or to even specifically shape them.

The final thing I want to say about this topic is why the concept of a mechanical narrative even matters. Game apologists will often bring up the idea of the uniqueness of games. This almost always goes to the mainstay that games offer “interaction” and “choice.” But these things in themselves are rather meaningless—even DVD menus offer as much.

The true uniqueness of games as a medium is that a genuine dialog arises every time a game is played. A clash, or at least some sort of integration, between the rules of the game and the rules of the player happens every time, and that is what gives ludic interaction meaning. Interaction, therefore, is just a word—and an inadequate, non-specific one—describing the consumption of games, much as reading a book, or watching a movie describes the consumption of those forms of media.

The utility of the concept of “mechanical narrative”, then—or whatever we decide to call it—is to give name to that clash, that dialog. And it matters in that, through specificity and recognition, it helps us think and talk more directly about what the rules and design of a game do to the players that play it.

Related Jobs

Rensselaer Polytechnic Institute
Rensselaer Polytechnic Institute — Troy, New York, United States

Assistant Professor in Music and Media
The College of New Jersey
The College of New Jersey — Ewing, New Jersey, United States

Assistant Professor - Interactive Multi Media - Tenure Track
Next Games
Next Games — Helsinki, Finland

Senior Level Designer
Magic Leap, Inc.
Magic Leap, Inc. — Wellington, New Zealand

Level Designer


Kenneth Blaney
profile image
"The narrative is true because of the way the rules have been played."

It is a little off topic, but I think you've just given me vocabulary to describe the situation I've always found distasteful about economics as a science. In that same breath, I think these ideas might then relate in some way to the recently posted "How I used EVE to predict the economic downturn" article.

Taekwan Kim
profile image
I would say that line of thought is actually relatively on topic, if I am understanding it correctly (which I totally may not be). I might be going off completely on a tangent here, but I think it offers another good opportunity to examine diegesis and rules as presented in the post.

Let's talk simplistically about capitalism as a rule based system of interaction. For free marketers, regulatory rules are extra-diegetic impositions. The only fair rules are diegetic—that is, for them, the only fair thing to do is to allow "natural" competition as resulting from the random advantages and disadvantages distributed by nature and circumstance. For progressive types, the diegetic reality—this random distribution—is inherently unfair. There is nothing fair about being born into wealth or similarly advantageous conditions and being able to leverage that to edge out competition. Moreover, as noted in the EVE discussion, the longer a circumstantial advantage is leveraged, the more that advantage becomes disproportionate. Fair trade for progressives, then, simply *requires* extra-diegetic rules because, again, the “natural” state is already unfair.

This is where that question “What is ‘winning’” becomes useful. Is “winning” the pursuit of individual happiness or societal equality? Depending on how you answer that question, then, the rules you care about and follow are going to be significantly different, which of course means that the mechanical narrative changes depending on this worldview. The argument about regulation vs deregulation is therefore essentially an argument about which of these mechanical narratives is truer. Should economies be player-driven or designer-driven (or some sort of hybrid)? Etc. (I may have posed this whole thing in a rather biased manner, being the progressive that I am.) So there’s the “how the rules are played” bit.

The other part of the EVE discussion is how mechanical narratives are so readily transferable. My take on this is that life as a human is basically just a bunch of interactions with a bunch of rules, some self-imposed, some not. There are many, many times I find I am able to relate to and engage in intelligent discussion about activities that friends have experienced, but which I personally never have done, because the mechanical narrative is the same. And there is usually an immediate, mutual sense of shared experience. The rules play similarly enough that I can be knowledgeable and understanding of their psychological processes and thought patterns during those activities without having undertaken them myself, because I've experienced those same psychological processes as well, even if in an entirely different and unrelated context.

But that's also the potential danger, I think. Mr. Rumsfeld's mistake of false pattern recognition is one that is entirely too easy to make. I would say, however, that the more we play games (or more exactly, the more we learn by losing in them), the easier it becomes to recognize and differentiate between, in any situation, which rules actually matter, which rules “should” matter, and which are rules we only believe matter. So, once again, games are awesome, etc.

But uh... yeah, I've probably read into this way, way too much.

Kenneth Blaney
profile image
Yes. That is more or less the situation I am describing about economics. In this case, I'm thinking about how behaviors are altered based on research, but then new behaviors cause new research (and the cycle continues that way). So in a sense, the research impacts the research which shouldn't happen in a strictly natural science. Certain rules about economics are followed because "that is the way they are done" and only because that is the way they are done.

I wouldn't say you read into my comment way too much. I'd hopefully just suggest "great minds think alike" and hope that I can get away with complimenting myself.

Nick Harris
profile image
You may find these links to be of some interest:

Apologies for the screwy formatting on the last one...

The earliest post I can find on this topic is from August of 2009:

I also recommend these articles:

It appears to be an active area of research for next generation adventures (I prefer not to call them games because to be dramatic they may have to eschew competition and a balanced set of rules in preference for procedurally generated drama that conforms to an underlying theme) - i.e. interactivity means having to let go of a predetermined plot and just be satisfied that the theme was conveyed by the player's interactions with the simulation. I don't feel terminology, such as Ian Bogost's "Procedural Rhetoric", or Mechanical Narratives, or whatever I have described it as, matters all that much, so long as a pragmatic solution is sought to the problems of creating compelling synthetic drama.

Altug Isigan
profile image
In the semiotic sense we can distinguish between approaches that reduce narrative events to appearance of forms and approaches that produce narrative events through weaving an invisible semantic web. What produces an event is not simply that the ghost chases Pac-Man, but that the semantic web that is displayed through these forms generates a situation of running for your life that is "bigger" than it appears: The many details you need to tell to clarify the situation. That's the semantic web beneath the forms who communicate it.

In many narratives that we would label as "linear", you can feel how the deeper structure (the semantic web) fills the forms with more than their surface appearance. It's not simply that A needs to kill B if he wants to survive, but the complex structure of forces that make such goal inevitable and donate it with the many details that make the event itself so intriguing.

The problem is not one between systemic and procedural events per se, but one between shallowness and depth, for procedurally generated events may still be shallow and systemic ones quite deep. To get a better understanding on the issue, the attention needs to shift from the signifiers to the signifieds, from expressions to semantic webs.

Rich Boss
profile image
YES! I think I completely get what you are saying. If I am understanding you correctly, you are saying that the interactions between the player and the player character are just as important as the interactions between the game and the player character. Is this correct?

Taekwan Kim
profile image
Hey guys, I want to apologize for taking my time in responding. Some very interesting points were raised, and I felt the need to consider them thoroughly before trying to add to the discussion.

Mr. Boss, I think you’ve pretty much hit the heart of the matter. I agree with Mr. Harris that the labels applied here don't really matter. But I would say that I think these concepts might not actually be equivalent. My comprehension of Dr. Bogost’s procedural rhetoric is imperfect, but I believe the emphasis is ever so slightly different from the mechanical narrative. As I understand it (and again, this could be an incorrect understanding), in the experience of procedural rhetoric, there is no real distinction between the player and the player character, between the diegetic and non-diegetic rules. In mechanical narrative, the “narrative” or rhetoric forms because the diegetic and non-diegetic rules or goals are distinctly and experientially *different*; a mechanical narrative is always about the process of bridging that difference.

In this sense, the procedural rhetoric represents an *aspect* of the mechanical narrative—the aspect which is experienced when the player mostly decides to adhere to the diegetic rules and forgo non-diegetic ones (what I described in the post as “voluntary assimilation”). But there is another aspect to the mechanical narrative, which is the “narrative” or rhetoric that arises when the player refuses to care (or primarily care) about the diegetic rules. (This happens any time a player disregards a game’s win/loss conditions, or any time a player deliberately adopts the least efficient path towards a game’s win conditions—any time the player follows their own rules.)

For instance, the narrative of “An Illusionist in Skyrim” is only possible because Mr. Francis completely ignores—in fact, makes it a rule to prohibit—the implicit diegetic rule of reciprocal (or pre-emptive) force against an aggressor. This self-adopted, firmly non-diegetic rule (Mr. Francis even begins the series by establishing its non-diegetic origin, explaining, "In Skyrim, a mage is an unstoppable storm of destruction. In real life, a mage is just an illusionist: they can’t do much except trick you.”) is at complete odds with every diegetic rule in Skyrim, which forces Mr. Francis to take on a series of compromises and loop holes (to his own rules) to play the game.

(Mr. Francis, then, is more or less debating or ignoring the game’s rhetoric to form his own counter-rhetoric, and in this sense, you could also say that this example is simply the player-driven form of procedural rhetoric, as opposed to the designer-driven form that Dr. Bogost describes.)

This finally brings us to Mr. Isigan’s comments (hey! It’s been a while :D), which I think are completely spot on, and have much more precisely stated what I was trying to get at with regards to events and process. The mechanical narrative in the above example of “An Illusionist” isn’t about the specific stories that happen because of Mr. Francis’s rule decisions, it’s about the series of rule decisions and compromises he has to make to continue to pursue his non-diegetic rules, which just happen to cause those stories. (Of course, the stories that just happen to arise then further impact the rules which are at play, etc., as Mr. Blaney observed.)

Basically, the idea I am trying to emphasize here is that gameplay is more or less a result of the differences between the player and the game’s realities, rhetoric, or rulesets and goals—and that the dialog of *either* reconciling or ignoring those differences produces its own narrative. (There is always a mediated layer between the player’s goals and what is permitted/achievable by the rules of the game.) The important thing is that gameplay goes both ways; as much as a game dictates its own rules, it is the player that ultimately decides whether or not to follow or care about them. And that decision itself has meaning.