Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 24, 2014
arrowPress Releases
October 24, 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:

How I Analyze a Game
by Raph Koster on 01/13/14 06:26:00 pm   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.


The first thing I do is set aside my experience. It is only mildly useful, a single data point, when everyone’s experience is subjective. Oh, I’d like to think it is in some ways more valuable than that of a typical player. After all, I have a very specific set of experiences to bring to bear. But in practice, it probably makes my subjective experience well-informed, but therefore less than helpful.

Continental-continental_convergence_Fig21contcontLooking at the experience is like seeing the top of a mountain without knowing about tectonic plates. I use that analogy because the typical analogy is that of seeing only the tip of an iceberg. But an iceberg is substantially similar above and below the ground. Sure, there is a lot hidden under the waterline, but it’s not different in nature. When we look around the world, what we see, what we experience, is powerfully shaped by things that we do not see. Without understanding fault lines, volcanic activity, and all the rest, we won’t come to understand why a chain of mountains is where it is, and why it takes one form versus another.

That’s why I start with the stuff “under” the experience. Mechanics, inputs and processes, rules and tokens and actions. I strip away the surface until Gone Home is a game about flipping over cards on a desk to see what is underneath them. Papers, Please is a Spot-The-Difference game. The Stanley Parable is a choose-your-own-adventure where some of the options are written in invisible ink.

In all of these cases, I set aside a whole bunch of stuff you do in the game, for a moment. Moving in The Stanley Parable isn’t even a means to an end 99% of the time. It is empty decision cycles going by mostly for the sake of experience.

What are the systems?

Now I can examine that core mechanic. Odds are extremely high that I have seen it before. New mechanics are rare in game design. Press a button before a timer runs out. Move an asset from one pile to another pile. Make a choice to move in one direction in a simulated environment or another.

Sometimes you get pleasantly surprised; Papers, Please is the most robust version of Spot-The-Difference you’ve ever seen. It unfolds more ways of being different over time; it forces tension between mousing time for accuracy and speed for maximum efficiency; it forces games of memory on you, layering difficulty as you advance.

What am I looking to understand? I want to know what the possibility space is, extrapolating how the curves I see will develop. I don’t need to play all the way to the end of the game to see how every variation unfolds — I just need to see how broad the space for variations is. I want to know what choices I have to make, and how much consequence they will have in the system. Just the system, mind you, not the experience. I want to know if I can exploit the system I see. I want to know if my skill matters at all. I want to know if I am learning something as I go. I want to know how the systems build a mental model in the player, how they scaffold.

All of these things are elements on my tuning checklist for my own work. I look for them so I can look at the craft at this mechanical level. That’s because I analyze games in order to improve my own craft, above all. Sometimes I am doing it in order to help someone else improve their craft. There are many other reasons to analyze a game, but those are mine.

What are the systems about?

Now that I can see the skeleton, the tectonic plates, I look at how systems interact. I ask myself, “what was the designer trying to accomplish here?” And I make a guess. It is no more than a guess, unless there’s some explicit signal telling me the answer. That said, the guesses aren’t that hard, because usually it is telegraphed in a dozen ways. The commonest things that the designer is trying to do are

  • get the player to keep playing
  • get the player to pay money
  • get the player to pay attention to the experience rather than the system
  • get the player to feel good, usually by making them feel powerful
There are many exceptions, of course, but those four are the biggest ones. I get excited when I see a purpose at this systemic level that is more interesting. For example, when I see a systemic structure designed around
  • making the player want to help other players
  • making the player suspicious of the rules themselves
  • making the player understand a new “language” — a fresh way of thinking about problems
  • making the player feel good about things other than power: altruism, cooperation, creativity, their own intelligence, etc

Most systems are not like these. Most are the familiar first four. And not every game needs to do more than those first four, not if that was the intent. You can have an unambitious systemic design there, and execute it really well, and analyzing the game in this way will tell me that.

How do I touch the system, and how does it touch me?

Once I understand this, I can then assess two distinct aspects of interacting with it: the inputs I can perform, and the feedback I get back. We tread dangerously close to “experience” here, but with care we can consider these things in terms of “game feel,” responsiveness, reward signals, learning scaffolding, and so on, without actually embracing the particular fiction the game might offer up. In other words, I can look at things like whether the controls feel good, given the system that the game wants me to engage with. I can draw the conclusion that the best way to mark my own progress in Gone Home is to trash the house, because the game provides fairly minimal affordances as to what is clickable, and zero indication of what has been previously examined.

I now have enough to make a judgement, which is really just for my own craft purposes. I have decided to view the work through the lens of an intent, or an artistic goal. And I have looked at how the game’s systems allow me to relate to it. So I can decide, for myself, whether the game is Good At, or Bad At, meeting its own apparent intent.

For example, I can look at the opening bank robbery scene in Grand Theft Auto V and note that it acts as a tutorial. I can see that it gates advancement by getting me to perform specific actions correctly. I can note the fact that it doesn’t tell me which button performs a given action, but simply assumes I have read the manual or played many other games before that happen to use the same mapping. I can see that it tells me to “take cover” in an area where there are many plausible “hiding spots,” but that the game expects me to be in a specific particular one that is only indicated on a minimap via a dot that has not been previously explained. I can observe that in some areas of this tutorial, not performing the right action has no negative effect, and in others it results in an instant mission failure and reset.

All in all, this leads me to a judgement that there could have been more usability work done on this tutorial.

It is always possible that I have the intent wrong (though in general, assuming the opening to a game is meant to teach you basics is usually a safe bet, unless you’re playing 868-HACK or something). More on that in a moment, because now I can turn around, and look at the experience.

What is the game experience?

mountainsThe experience is an altogether different kettle of fish. I now try very hard to suppress everything that I now think I know about how the game works underground. Now I want to look at the majesty of mountains, feel the fresh air, smell the rock and the snow. Oh, not for enjoyment! No, I want to smell the snow to figure out how pure the water that went into is. I want to know how these mountains were made.

I have multiple things to look at. I have visual storytelling. I have graphics rendering. I have music. I have writing. None of these are specific to games. The game-specific bits of the experience were pretty much all covered already under game feel and feedback. No, here I am deploying the craftsmanship analysis of other media. Whether a texture has good color balance, and whether the overall screen picture has so much saturation (like, say, in the opening island of Assassin’s Creed IV: Black Flag) that telling what my next destination should be is hard. Does the level design provide “weenies?” How is the voice acting?

My standards here are not the standards of games. I am going to hold the story up to stories from other media. I am going to hold up the art direction to that of a film. I have many reference points here, and I am going to use them. This means, I will usually be disappointed. I mean, for all the really cool character design in Rayman Legends, the storytelling of the opening sequence is kind of abrupt.

What is the experience about?

But again, I need to ascertain what the various disciplines were shooting for. And here, intent is generally far more telegraphed than it is in the systems. In fact, seeing what the intent is here usually corrects any misapprehensions about what the system’s intent was.

It is pretty rare to see deep symbolism going on in games. When we do see it, it is pretty rare to see it be accessible. As it happens, I personally value the on-ramping quite a lot. When a game’s experience is intentionally obscure or obfuscated to make a point, that means that the intent is to speak to only a relatively small audience, one which is already clued in on aspects of the “language” that the work is using. I think of that as a form of “preaching to the choir.” To me, empathy and understanding lie at the core of art, so a work that demands all the work on the side of the player, rather than the game experience having empathy for the unaware player, for me falls down to some degree. That may have been the intent of the creator, but it’s not an intent I agree with, so I note it and move on.

Most experiences in games are what I would term “impositional.” The goal the developers had in mind is for the player to have exactly the emotional experiences they intended. There is a set of craft techniques that you can use for these. When they are good, these experiences prompt thought or reflection, leave the player changed. At their best, they do not leave you with pat answers. That’s what I expect from great stories in books, so it is what I expect from great stories in games.

There are experiences designed to be “expressive” instead. These involve a different set of techniques, and therefore have to be looked at on the basis of the strengths and weaknesses of those techniques instead. Here, I am more likely to use the great toys of history as a benchmark.

Either way, if I can’t tell what the intent is, I consider it a failing.

So now I look for the synergy between these different media of experience. Music and art and story pulling together, towards the same goals? And how lofty are those goals? Something like Tomb Raider has to reinvent a very familiar main character for today’s world, show a character arc (vanishingly rare in game stories, alas!), live up to a sort of cinematic vibe, and be beautiful to look at. A good understanding of the limitations of the platform help here. I can’t possibly ask for the same sort of experience on a big screen TV versus a phone.

Do these things all match up?

We’ll get a story that makes gestures towards great significance: big mountains! Sitting atop a generic power fantasy game. These are hollow mountains, built atop a foundation that cannot bear the weight.

And now comes the crux of it.
  • I know the intent of the systems.
  • I know what the systems actually teach.
  • I know the intent of the experience.
  • I know what the experience actually says.

Do these intents match? All too often the answer is no. We’ll get a story that makes gestures towards great significance: big mountains! Sitting atop a generic power fantasy game. These are hollow mountains, built atop a foundation that cannot bear the weight. We’ll see cases where the gameplay and the experience have literally opposite intents, or radical mismatches. These are a problem.

I try not to think about the ultimate “theme” or “moral” or “lessons” of a game until this moment, because it is so common for the answer to be muddled. But now is the time where questions arise about things like what the “politics” of the game are, what it implicit ethical opinions are, what it is saying about how the medium works, how it pushes at our understanding of form. It’s only here that I can form an overall opinion.

It’s not at all unusual for me to think very highly of a game that has put all its attention on experience, and very little on mechanics. The converse is also true; happens to me all the time. But I reserve my highest consideration for games that execute on every level. I am most interested in games that have ambition on both sides, where the intents on these very different levels of craftsmanship line up: in a word, like Papers, Please.

It is here where I say “oh, tedious mousing here is because it ties back to the narrative’s point. This system being out of balance is intentional. And the story doesn’t wrap up because the mechanics wrap it up for you.” That sort of thing.

What about fun?

And, in the end, what does this have to do with whether I personally enjoyed the game? Pretty damn little. Like I said, my own enjoyment is deeply suspect. It’s not only subjective, but it’s driven in huge part by having walked through the above process. Once I have looked at all the above, it is hard to unsee it. I am more eager to finish playing Tomb Raider than Papers, Please. I think I have more raw fun in Rayman or Forza. Because once I am done, I want to have fun too. And everyone’s fun is different.

In general, this analysis will tell me how likely a game is to be fun for the people it’s meant for.

All of this is how I analyze games because it is how I wish my work was analyzed. I want to know how I did on each part, and I want to know whether it works together. Criticism or commentary that only touches on a part of the above is useful to me, but only to a limited degree, precisely because these elements are all interdependent.

Most game criticism only looks at a fraction of these things in one given article. There is almost never a technical critique. The typical reviewer most readily touches on the basic subjective experience of fun; the more thoughtful may consider the experience layer. Very few think about the artfulness of the system design. These individual lenses are all fine, of course. It does mean that a creator must gather together a lot of separate analyses in order to get a picture of the whole.

A final thought

I recognize that this entire lens and process is deeply craft-centric. There are modes of criticism that are more about putting the work inside a cultural context, or focusing purely on the response a player has. I happen to think all of those are flawed for my purposes unless they engage deeply with the craft side; without an understanding of the creation process, the analysis is very likely to involve some big assumptions.

I also recognize that the language is largely missing for talking about all the systemic level stuff. It’s cast as reductionist formalism, or as less important than the player experience. Some of us are trying to change that. On the flip side, we have lots of language borrowed from other fields for talking about player experience. We typically give the experience a pass, though, by considering solely the parts that are obvious: the story, the graphics, the fun factor. We don’t look at the synergy enough.

One thing I will say is that this sort of analysis works equally well on an “art game” or a commercially-oriented blockbuster, because it is about assessing how the work is doing what the creator intended.

So, that’s how I analyze a game. I wish more people analyzed them that way, for purely selfish reasons. It would make my job as a designer much easier.

*Every game mentioned here is in my Top Ten Most Interesting for 2013 for one reason or another.

Related Jobs

Next Games
Next Games — Helsinki, Finland

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

Level Designer
DeNA Studios Canada
DeNA Studios Canada — Vancouver, British Columbia, Canada

Analytical Game Designer
University of Texas at Dallas
University of Texas at Dallas — Richardson, Texas, United States

Assistant/Associate Prof of Game Studies


Lars Doucet
profile image
Fascinating! To play devil's advocate, what's your thought on "The Death of the Author" as applied to games?

Raph Koster
profile image
It's a lens too. It's just not one that helps me improve my own games. So I can look at something and attempt to disregard authorial intent in it, but then I won't have an anchor to use to assess how well it is working at whatever it is doing. I'd end up falling back on "how well it works for me."

That sort of subjectivity is great for some purposes, but it's not great for giving me things to take to apply to my own work. After all, I can look at my own work anytime and say "how well does this work for me." That works well for self-expression, but doesn't really help me speak to anyone else. I personally want to speak to others through my work.

There is also the lens of "how well does it work for a large chunk of people." That populist lens tends to end up in a very consumerist place, though... so it can be super-useful for market-driven design, but not very useful for artistic expression.

Raph Koster
profile image
I should add, glancing over the Wikipedia article for "Death of the Author," that I do agree with Barthes' contention that we can't perfectly know the intent, that it can be muddled, that a text cannot be boiled down to a single meaning, and all the rest. :) I don't think it is impossible to attempt to guess at intent (particularly when so many games have relatively low ambitions!) and agree with him at the same time.

Lars Doucet
profile image

For the record I was really playing devil's advocate -- I'm not actually a "Death of the Author" fan (at least the maximalist version), though I do agree that criticism must be more than just interviewing the author about their intent (for the author doesn't always have the final word).

In the end I think criticism is similar to say, archaeology. Authorial context and origins matter.

Robert Leach
profile image
I think we're at an early enough stage in the lifetime of games where the intent of an overwhelming majority of titles can be assessed using your system or a similar one. For film, literature, sculpture/painting/photography, and music, many pieces are abstract enough and smart enough to evade easy assessment.

Games feel like they're finally starting to find their stride with regards to higher levels of meaning, and more complex levels of involvement by the players, and thereby are beginning to appeal to different people in different ways, sometimes vastly different than the initial intention. However, I think what you've said above is an excellent way to write intelligently with regards to the medium as it continues to mature.

Raph Koster
profile image
As I said elsewhere, even in other media, it's still VERY rare to find a piece that can't be assessed in this way... they exist, but it's really not very common at all. Abstract pieces, for example, are usually very susceptible to being thought of on the systemic side, because they are usually formal explorations of a space.

That said, I agree that games are appealing to people in different ways. This sort of analysis is not going to tell you anything about speedrunning, per se, just as an example.

Ben Fowler
profile image
A more objective way to analyze games is always welcome. If I reviewed games, I would generally just ask "what are the designer's intentions, and how successful are they" rather than how much I enjoyed the game. This seems like a more robust system of trying to do that, though for more perspective on how to design your own games. I don't think I could throw away my previous experiences, though. There's too much pressure to do comparisons, especially in trying to convey meaning to other people.

Shaz Yousaf
profile image
Very interesting article. Trying to make the 'message' of the game completely integrated into the mechanics and overall gameplay is something I try to think about more consciously now. But it certainly isn't easy.

One good way of approaching it I heard recently is "Whatever you are trying to communicate, make that be how you beat the game". That helps a lot to make the theme more easily understandable.

Eric Shofe
profile image
Great points to consider. Thanks for taking the time to write this.

Bradley OHearne
profile image
That was a good read. I'm a how-does-the-watch-work type, so I like dissecting things which I enjoy to see how they are designed, and achieve what it is that I enjoy. I frequently by extended and deluxe versions of DVD / Blu-ray movies just so I can watch the Documentaries in the Special Features (which I sometimes like more than the movies themselves). But all that being said, whether a game or a movie or a story I read in a book -- the formula doesn't dictate my enjoyment, but rather my enjoyment makes me interested in deconstructing the formula.

Robert Leach
profile image
As a viewer or player, I experience and react in my own way, but as a reviewer or student, this kind of dissection is essential. It allows the art to grow via awareness, and these lessons in theory that will be taught in schools will go on to produce results in the creative culture.

At least, those are my thoughts :)

George Blott
profile image
I'll add 'weenies' to my list of creepy things Disney was was responsible for. At our company we've always referred to them as "Tour D'Eiffel"s.

Great article!

Raph Koster
profile image
It was apparently a reference to a literal hot dog.

George Blott
profile image
Yeah! I read a bit into that forum thread you linked. I think when they defrost Mr.Disney he'll be pleased with all the hot dog innovations we've seen since his time. [redacted for space: a list of hot dog innovations]

Also, will be playing a fresh game tonight (Broken Age) in a different light after reading your piece. I imagine most game designers do these things in an instinctual manner when playing, I know I do, but not always with the kind of intent you describe. I'm curious to see the results of putting in practice your analysis "recipe".

Raph Koster
profile image
I really don't do it by checklist anymore... it is a habit, a reflex. You can often get most of the way through this analysis just by playing to the first boss, twenty minutes. It's actually pretty rare that a game unfolds real surprises past that, though it does happen.

So the output is usually just mental notes. If I stopped to write it all down, it would take hours. :)

Robert Hewson
profile image
This is why, to my mind, game design these days is sometimes undertaken in a back to front manor. Rather than games growing bottom up from interesting systems to an appropriate experience layer (to use your terminology) applied to complement the interesting system, designers come up with what sounds like an interesting experience layer and then engineers try to build the ground underneath.

The bedroom coders in the 80s and many of the small indie devs of today have to be engineers first, so perhaps this is why some of the more interesting games came/come from them. Bigger, more commercial developers are perhaps more prone to green light high-level concepts and then spend huge amounts of time and money constructing the systems beneath them.

Michael Wenk
profile image
I'm not sure how valuable an analytic analysis is on what likely is more intuitive. I guess it would depend on the playerbase and how they make decisions. While the more analytic personalities would probably model well, the less analytic people would not. In the end that would add a whole lot of uncertainty. I don't see how you can model that without asking those people what they want(surveys), or monitoring what those people say(datamining social networks, forums, etc)

Raph Koster
profile image
I am not sure what you mean by "modeling the personalities" to be honest. I really don't mean to imply I sit down and put together formal audience models or anything like that.

I don't "datamine forums." I hang out on them because that is where my players are.

TC Weidner
profile image
but what is really interesting in making art, movie making, music,game design etc, is that many of the masterpieces and classics come not by perfect planning and introspection, but by happen stance, quick thinking and adjustments, and just plain luck married with a creative mind and a solid foundation of an idea.

Its why I love behind the scenes looks into how certain movies, music, and games are made. There are always these crazy awkward turns.

example- My favorite movie of all time - Jaws, if the mechanical shark actually worked as planned, the movie would of probably not been very good at all. Instead, the shark was garbage, so a quick thinking team decided, well we wont show the shark til much later, we will replace it with music and shoot it form the sharks perspective... boom classic.

Funny thing about art.. such things can be found in all sorts of classics. So while I agree with much of what is said above, I would add embracing the chaos, listening to your gut, and being able to make lemonaid from lemons is just as important as any analytical introspection.

As Bob Ross always said, we dont make mistakes, just happy little accidents

Raph Koster
profile image
Very true! I think knowing all that background and history helps. But either way, if I used this method on Jaws, I'd notice the absence of showing the shark, I'd notice the suspense it drives, and maybe I'd notice the fact that the mechanical one, once shown, barely moves.

I wouldn't know the history of how it came to be, but the technique of "if you have a powerful nemesis, but it doesn't work very well, you can hide it and build suspense for its appearance via rapid movement from a different perspective... and then when you show it, show it in such a way that it's not moving."

That's a useful tool in the toolbox, you know? If I run into a similar situation later on with my own work, I can draw inspiration from that.

TC Weidner
profile image
true, but sometimes understanding where exactly the magic lies , since it is often more than the sum of its part, is impossible to pin down.

Raph Koster
profile image
I dunno. I come at this after having some formal training in music, studio art, and creative writing. It's pretty rare, in my experience, to say "it's impossible to pin down why this works.” Oh, it happens sometimes, but not that often...

TC Weidner
profile image
if it were possible and/or even common to understand the " magic" of art, it would be easily reproducible, it isnt. Thus the vast majority of movies, music and games are not good.. Proving my point.

Critics/experts are common, great artist and great art are not.

Raph Koster
profile image
Oh, I see the gap between what we're saying. :) I am assuming that most work doesn't reach quite that lofty a height. That's why I say it is rare to not be able to pin down why it works.

It's worth pointing out that there are quite a lot of classic "great" works that we do in fact have a pretty decent understanding of how they do what they do... But yeah, not all, not by a long shot.

That doesn't mean something that doesn't reach that height isn't good, or that we can't learn from it, of course.

It also doesn't mean that knowing how something achieves an effect automatically implies being able to reproduce that effect yourself! Art is hard, yo. ;)

Robert Leach
profile image
"It also doesn't mean that knowing how something achieves an effect automatically implies being able to reproduce that effect yourself! Art is hard, yo. ;)"

That, I believe, is the crux of the matter: good call!

Figuring out what works and what doesn't isn't that tough (formulaically successful games/movies/tv/books/music have been around for a long time). Figuring out how to do it perfectly is almost impossible, because that's where the difference between good and great is detail-work, extra effort, timing, audience, persistence, weather, personal mood/effectiveness, etc.

ganesh kotian
profile image
Keeping below points in mind always .....Thank u for the post.

get the player to keep playing
get the player to pay money
get the player to pay attention to the experience rather than the system
get the player to feel good, usually by making them feel powerful

Raph Koster
profile image
Ack, please don't make that the takeaway!

Rebecca Phoa
profile image
Thank you for posting this up! Always useful to read articles like this.

Matthew Thomas
profile image
This approach reminds me allot of the MDA framework (which is used more for research, but works just as well for analysis. I've used it for psychology research and system analysis).

A very nice read nonetheless. It's always useful to see how others approach analysis.