Sign up for news

Most people go on hiatus from blogging during the holidays, but that’s about the only time I have for it these days. The holidays are usually a time to avoid discussion of sensitive, agree-to-disagree topics, like politics and cat vs. dog preference. For me conceptual frameworks fall somewhere in the middle of that list, but I wanted to finally put down some of the problems I have with the Mechanics, Dynamics, and Aesthetics model of game design.

At a time when game designers were still too timid and uneducated as artists, it was a groundbreaking framework geared towards helping designers understand the artistic power they held. We have thankfully have come a long way in those 10 years, which have seen the rise of art-games and in general a wider spectrum of videogames.

A decade ago most designers still pined for the violent cinematic opulence of Arnold Schwarzenegger movies (if they pine for film techniques now, it’s more often those that convey empathy, not those that showcase explosions better). The notion that mechanics could be driven by aesthetics and affect them was novel to many, but 10 years later and MDA is still our prevailing theoretical model.

It’s not so much that there are problems with elements of the framework, it’s more that its incompleteness and the things it does not discuss for purposes of abstract simplification have become more and more apparent over time. Using MDA to teach videogame design is a bit like trying to teach someone to be a great conversationalist by explaining that words make up sentences, and you should choose your words based on what you want to say.

MDA will always have an important place in the historical context of game design, and in teaching introductory game design – after all, to people who don’t understand language, you do in fact have to first explain that words form sentences before you get anywhere.

MDA has become the poster-boy (to assign a gender to a theoretical model) of videogames rebound-relationship from film, board games. For all the criticism heaped upon designers who undiscernedly apply film’s techniques to games, there is little seemingly left for those who hoist board game design conventions onto videogames. Don’t get me wrong, I share the former disdain, I just have a problem with any uncritical need for any other medium’s fundamental strengths being forced on top of a videogame. And we all know whoever disdains the most WINS.

My chief criticism of MDA is that it completely fails to account for player feedback and modeling of the mechanics. This is trivial for board games where player state is reinforced through the board and tokens. In running a any sort of simulation at 30+ FPS, turns out there’s often a lot more to managing what the player knows. It considers how both players and designers interface through each direction through M, D, and A, but lacks critical abstractions as to how player modeling of a mechanic can cause it’s failure or success, and therefore the failure and success of any dynamics.

This is crucial from a critical standpoint and a creative one. Most games fail not because their mechanics were somehow unaligned with their aesthetics, they typically never even get the luxury of failing that way. When a game doesn’t work most often it’s because the player can not understand what they need to or can do, how they can do it, or why they need to. This information has an asethetic component in its presentation, and there’s some aesthetic choice in how much of this you reveal and when, but at its core it’s predominantly an information management problem.

Case in point: the original Black & White. A fantastic idea whose primary flaw was that the player rarely understood what and how they were teaching their creature. So achieving goals (explicit or self-directed) became far too difficult. To solve this problem, you could imagine cartoony thought bubbles over the creature’s head showing in animated form what it learned and which action triggered the learning.

This information is not mechanical at all – there are no rules directly associated with it. Although it is triggered by them, it is not a mechanic in and of itself. You could say it was part of the teaching mechanic, but mechanics in MDA are just rules, and have no informational component. There is no part of the framework that is concerned with how the player is modeling systems in their head as they learn them.

From a creative standpoint, working on any videogame that utilizes any kind of systemic depth, problems with how the player perceives those systems are how you spend most of your time. Certainly I do. Look at something like Skulls of the Shogun. If you can get the player to understand the systems  then you can address its flaws and keep them engaged, but if you can’t even achieve that it’s hopeless. Yes, you can change the system to reduce its complexity to the player, and we all now know how Sid Meier achieved this with the combat odds in Civilization, but that’s shirking away from any sort of depth in games, which I can’t abide by and there must be alternatives to.

My other complaints of MDA are related. Most troubling, dynamics become an entirely hand waving affair (which I directly relate to most videogames moving away from any sort of interesting systems design, well, that and a perverse fetishism of Super Mario 3 and it’s ilk). There is also no conception of time in the model. Sure, you could say using a rule for 30 seconds would be a different mechanic than using that rule for 15 seconds, but that’s missing the point. Even the slightest adjustments in timing can drastically affect how much the player engages with a game – that’s when it’s apparent it needs to be a first-order concept in your model.

Uncoincidentally, board games have little to no conception of time. Players always take turns, and time is always suspended. Even with the case of time limits, it’s not as if the passage of one second has any more or less meaning mechanically than the passage of another. Although as an aside, if anyone knows of any realtime board games, I’d be fascinated to hear about them.

Time is a crucial component to both the experience and system design of videogames. It is infused with them at every level, from the the smallest to the highest level mechanics. Fractions of a second of an animation can make the difference between responsiveness and sluggishness. They make the difference between an action you can see, understand, predict, and model, or one you can’t possibly notice.

Creating interesting dynamics through system design (in my mind) involves defining the types of events your systems create or the ones you want them to create, and applying positive and negative feedback systems until they happen at times that meet your aesthetic goals. MDA involves the notion of using positive and negative feedback to speed up/slow down systems, but includes no detailed notion for determining how fast or slow they are impacts the aesthetics.

One part of my own loosely defined conceptual framework for videogame design involves considering tiered player decisions over time. What kind of decisions will the player be making every 5-10 seconds? Every 15-20 seconds? 30-40 seconds? Every 1-2 minutes? And so on. Not every game must have meaningful decisions at every tier along the way (although if it doesn’t have any it’s probably pretentious). Crafting the player’s agency over time is crucial to the types of experiences I’d like to create anyway, a process which has been given little critical thought (although we at least have some terminology).

Any conceptual framework will be forced to remove elements from abstraction, and so will only ever be fit for analyzing/creating some subset of games. Those lacking abstractions, though, influence everyone who uses and studies the approach – they adapt those blindsides as well, to some extent. Sadly I think most people who would be at all interested in proposing a more complex conceptual model are still concerned with a dire pressing need of making more interesting games.

But will students be hampered by existing models such that we’re hurting ourselves in the future if we rely on it too heavily in curricula? Certainly, it has its place, but it is a small introductory place focused partially on the past, which is why it is so helpful a jumping off point. It’s just nobody’s really jumped.

Beyond all this, there’s a more worrying specter – the fact that even with any problems, our primary formal conceptual framework for game design still gets completely ignored by those claiming to be students of game design –’s wiki on Game Mechanics lists 24 things, 0 of which are actually game mechanics (if I’m kind maybe one or two are, eg. you could paraphrase “levels” as “if you perform x you will increase your level” which could loosely be seen as a rule).

We obviously need to be doing more to communicate these types of frameworks to those that pretend to be invested in game design, such that they don’t negatively impact our own work. That must include a deeper discussion on what are models are missing, how they could change, and if/why they need to, even if we need to agree to disagree.

7 Responses to On Conceptual Frameworks for Game Design & MDA

  • James Wallis says:

    Very thought-provoking piece. Thank you for it.

    There are a number of tabletop games that work asynchronously and where speed is important–from Racing Demon, which my mother was so good at that she had to wear gloves as a handicap, to Pit–but the only one I can think of that works in anything like realtime is James Ernest’s Falling, which you win by being the last person to hit the ground. I have a feeling there may be others but they’re not coming to mind.

  • Heh, technically Atmosfear is a realtime board game but, uh, it’s *terrible.* Friend found it at a second-hand store, still wrapped, for $3 and we played it in jest not long ago. Is there such a thing as a B-board game, akin to a B-movie? If so, Atmosfear is it.

    Anyway, good stuff, thanks for sharing. It seems that MDA supposes a player understands and has well modeled most of a system’s rules. If that’s the case, I think it’s a good framework and certainly a good post hoc analytic tool. But as a production methodology, you definitely can’t take it as read that if you create some rules, players will easily understand them. As you noted, a great many games fall down because of this.

  • Charles says:

    When he came through the Game Center Marc Leblanc said something that I thought was interesting, which was that MDA was really meant to help spur creativity, not be a substitution. It was meant to get game designers through certain blocks by helping them clarify the effect they were trying to create.

    That said, I agree that we need a framework for accessibility. Though from my perspective that’s less a game design problem and more of an information and graphic design problem. It’s great if game designers have some understanding of different solutions, especially if they’re independent, but it’s a different job to me.

  • Darius K. says:

    I say this to everyone, but you really ought to read Steve Swink’s Game Feel. It clearly addresses a lot of issues that MDA completely fails to hit (though it does have its healthy share of SMB fetishism).

  • Borut says:

    Thanks for the game reccomendations Nels & James!

    Charles – yeah, I mean no model will ever be the be-all-and-end-all, they’re all most useful as starting points. But looking at something like novel-writing, there’s tons of jumping off points of structural models that you can use starting out as a writer to get you going (whereas we have one primary one, although occasionally gamasutra will run pieces by different designers that kind of elaborate their own more formal approaches).

    Darius – yeah Game Feel crossed my mind when I was writing the post – I’ve only perused it at the bookstore, but it’s the only book I know of that takes a more videogame-y approach. I guess I’ll get off my ass and buy it I suppose. :)

  • elias says:

    I know there are some real-time board games out there. Tom Jolly seems to enjoy designing them since he’s done two: Light Speed and Camelot.

    This is probably a good list to start with. I haven’t really played any real-time board games, though.

  • Dan says:

    Not sure it adds to the solid points here, but Space Alert is most definitely a recent (and popular!) real time board game.

    I’d argue Spit (the card game) is too.