Game Studies: Research is Strange

First, a bit of meta-blog business. I’ve been toying with the second part to the Pillars of Eternity series, trying to craft a post as long and complex as the first. I’ve been getting some pretty severe writer’s block on that front, though, so I think I’ll put that on hold for a while. Instead, I’ll tease out some ideas about my approach to researching games, with (eventually,  promise I promise) a focus on slice of teenage life/time travel drama Life is Strange. Essentially, the question I’m looking at here is how you orient yourself to what you’re studying.

The idea that there’s a single method is, of course, ridiculous at best, and polemic at worst. Like any other object of study, what method of game studies you apply depends on what you want to get out of it. The problem, for me at least, is that the methodology models out there are either extremely tenuous and lead to overly vague wanderings or they’re overly formal and lead to totalizations that narrow your scope or even push towards confirmation bias.

One of the best compromises between those two extremes, I think, is Consalvo and Dutton’s “Game analysis.” To paraphrase, they propose a four-part method:

  • Object Inventory. “A useful way for researchers to understand the role that objects can play in a game is to create an object inventory that catalogues all known objects that can be found, bought, stolen, or created, and produce a detailed list or spreadsheet that lists various properties of each item.” Based on the results, you can come to various conclusions about the game; I could see a table of the power-ups for Binding of Isaac, for example, being extremely useful for coming to conclusions about how the game encourages varying playstyles based on random outcomes—though it’d be much easier to consult a wiki than actually play through the game to find the whole set.
  • Interface Study. “Examining the interface (and going beyond elegance of design or ease of use) lets researchers determine how free players are to experiment with options within a game. Alternatively, it can help us see what information is privileged… and what information is absent or difficult to find. Examining the interface also exposes the consequences of choices, such as developing a character along one skill path but not another; and also helps determine what the game developers have deemed essential (as well as non-essential) aspects of gameplay.” As an example, here’s a thought experiment–what would it be like to play the Dragon Age series if you stripped the “Character X approves” notifications, and had to determine what each character thought of you based purely on their in-game actions?
  • Interaction Map. You could also call this the “Bioware/Telltale” question–it involves a systematic examination of NPC interaction, player choices, how interactions change over time, in-game dialogue. Consalvo and Dutton’s examples involve how The Sims‘ characters respond differently as relationships develop; another direction would be mapping out how characters in-game respond to player good-evil choices in Fable and the like.
  • Gameplay Log. This element is meant to look at the overall gameworld, and emergent play. Essentially, it’s a record of how the players interact the game, and how the overall game world unfolds. Arguably, it’s also the most expansive of the four categories, along with being the most general. I like to think the quintessential version of this is Tom Francis’ Galactic Civilizations 2 war report, though Consalvo and Dutton would probably inteded something a little less informal and a little more rigourous. (And bless PC Gamer for republishing the diaries; it’s always nice to see older game pieces brought back to the public consciousness.)

There’s a lot to like about this approach. As I said, I think it strikes the right balance between letting researchers draw their own conclusions and providing some structure. It is perhaps a little tilted towards studying games over studying players, but as a starting point, I think that works well.

The problem with putting it into practice, in my experience, is that–barring a large supply of graduate research assistants to do it for you–it’s a damn long process. While not using this method directly, I attempted a similar form of close play for Planescape:Torment, concentrating on how the game treated belief and writing (gameplay log), how characters were characterized by dialogue (interaction map) and how it distinguished between different types of textual information (interface study).

It took months. I filled 3 notebooks with plot observations, character descriptions, and item sets. I opened up walkthroughs, designer interviews, and let’s plays. I took a 40+ hour game and extended it into hundreds. I tested the patience of my supervisor, my committee, and my loved ones.

Screenshot 2013-12-19 16.22.26
Planescape: Torment. Funnily enough, “an infinitely long stream of bilious green gas” closely matches my mood after I finished compiling my notes on the game.

Granted, the results were a very thorough, very complex understanding of Planescape: Torment. I wrote the forty paged dissertation chapter with ease, and I could (and probably will) spend the rest of academic career writing about the stuff I left out. Further, the systematic approach also greatly altered the way I played the game. It pretty much had to–when you’re stopping to jot down names every time you have a conversation in a game that’s famous for being 90% conversation, you’re going to have a very different experience from someone who’s just playing the game in front of them.

That doesn’t bother me so much with  Planescape: Torment, as I was pretty familiar with it at the start of the research session, enough so that I was already familiar with the feel of playing it. But if I had approached a game for the first time in such a manner, I think I’d be striking too far on the game side over player side, emphasizing game structure over player experience. Part of Bernard Suits’ definition of game in The Grasshopper is that it must be undertaken with a “lusory attitude,” which he defines as the state where “the rules are accepted just because they make possible such activity.” While I wouldn’t want to be too dogmatic about enforcing the idea, I like the concept that part of a game being defined as a game requires the player to enter into it with the attitude that it is a game; it enforces the role of the player, and the consent of the player. While I don’t think a lusory attitude is incompatible with critical analysis, I think the odds of the analysis overwhelming that lusory attitude increase as the analysis goes along.

To sum up, then, I came away from the Planescape: Torment analysis with two conclusions about my approach:

  1. It takes too long, and is too unproductive in the short term.
  2. It risks jeopardizing play in the name of structure.

At the same time, though, the other extreme, taking no notes on a first playthrough, is also too unproductive. If you’re tending towards studying 40+ hour games, then for expediency’s sake, you’ve got to something besides adopting a “I’ll come back to this later” attitude.

My immediate solution has been to replace endless notes with screenshots. It helps with the flow of play, but creates a backlog problem. Just taking a screenshot when anything interesting happens results in a lot of screenshots. I wound up with 150 for 11th Hour, and I hated that game; for a game I actually like, like Pillars of Eternity, I ended up with just under 2000.

Screenshot 2015-08-30 02.25.17
One of 11th Hour’s riddles. Solution: locomotive. There’s dozens of these. It’s a thin line between eldritch horror and the NYT crossword.

(Quick 11th Hour aside. The game’s best part is the idea that the two main female characters, both the one you’re trying to rescue and the one who gives you hints–“Do you want me to solve it for you?” is the game’s most merciful yet mocking sentence–are both more likable and more competent than the journalist-bro you actually play. The worst part is the rest of the game.)

All in all, though, it was an improvement, and the Pillars of Eternity post is the proof of concept; I had a much shorter playtime investment, yet still had enough evidence logged to go back and get what I needed. It still felt like I was missing a bit on the play/flow side of things, though, and the surplus of images created their own sorting problem.

I stumbled into a solution of sorts a few months ago, while searching for an answer to a different problem. The problem was (brace yourselves, tech-savvy readers) that my screenshot  method was no more sophisticated than pressing PRT SCRN, a function that, in Windows at least, has some well-documented problems when it comes to working properly in games using DirectX. While there are various workarounds, I went with the free version of FRAPS. It’s got a number of functions, but the one I was most interested in was that it allowed DirectX-screenshots to be mapped to any key your choice. At the time, I was frantically putting together DOOM images for a presentation, and it performed exactly as intended*.

zdoom 2015-05-25 17-00-12-65
*Well, almost. I accidentally mapped the screenshot to the same key that increased the light, creating screenshots of Hell’s brightest corridors.

Fast-forward a few months to last weekend when I binge-played my way through the four episodes of dontnod’s Life is Strange that have been released thus far. I’m sure I’ll talk about the game at greater length at some point; for now, I’ll just say that it stars a teenage photography student Maxine Caulfield who learns she can rewind time, and puts her powers to use solving a local mystery. It felt to me like Gone Home crossed with a Telltale game– that is, very detailed, evocative environments coupled with conversations and moral choices. If this is the future of the adventure game genre, then I am all for it.

The relevant point at hand, however, is that for about a third of my total playthrough, I went with taking screenshots at any moment that appealed to me, without checking whether the screenshots were actually working. The result was, well, this:

Screenshot 2015-09-25 01.13.48
Sadly, this is not a scene of a black bear in a coal mine blinking at midnight.

So I set up fraps. I was, however, feeling a little burnt out on the whole “manual screenshots” thing, as I had just taken four dozen screenshots over the course of four hours that had preserved nothing. Subsequently, I availed myself to another side of fraps I hadn’t noticed before (I realize this post is starting to sound like an advertisement for fraps. If it helps, substitute in your screenshot-taking program of choice.): the automatic timer. I set it up to take a screenshot once every 1000 seconds (16  minutes and 40 seconds), and went back into the game.

This solved a lot of my earlier complaints. I had a record of what I was playing that wasn’t so exhaustive that it took more effort to sort than actually replaying the game, and I could go with the game’s flow without feeling I was forcing an analytic frame on it. Theoretically, (I, uh, still haven’t tried this out), I could review those screenshots to remind myself what I found significant about the game, and if there was something that stuck with me that wasn’t encompassed in the shots, that was a good sign of what I should focus on during a replay. The big drawback was that 16 m 40 s was a bit too long a gap; after another third of the game, I went back and adjusted it to snap a shot every 500 seconds (8 minutes and 20 seconds), and that worked a lot better.

I’m not saying that any of these approaches are necessarily better than the others. Or that these are the only approaches–this is all very game-centric, despite my attempts to find methods that alter my playing style, and if I did want to focus player behavior, a different approach would probably be more helpful. Rather, this is just a description of my own evolving technique. But I really found describing the process I went through useful in putting into words what I want to get out playing and studying a game I’m interested in researching.

So what about you? What’s your method? Where could I improve?


I wanted to end this post with some of my Life is Strange screenshots–but that plan hit a snafu when I realized I set fraps to take the shots as bmp files, which are not supported by  Wordpress. Awkward. But one free mass converter later, we’re ready to roll. The shots are below. There are some spoilers, though I tried to remove the more overt ones.


LifeIsStrange 2015-09-26 01-08-49-14 LifeIsStrange 2015-09-26 01-58-49-09 LifeIsStrange 2015-09-26 03-38-49-08 LifeIsStrange 2015-09-26 05-02-09-10 LifeIsStrange 2015-09-26 05-52-09-12 LifeIsStrange 2015-09-26 06-08-49-15 LifeIsStrange 2015-09-26 21-54-05-36 LifeIsStrange 2015-09-26 22-27-25-34 LifeIsStrange 2015-09-26 23-59-05-36 LifeIsStrange 2015-09-26 23-15-46-07







Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s