Screencasts- a new method for game analysis?

By: Heleen
On: May 21, 2007
Print Friendly, PDF & Email
About Heleen


At the moment I am writing a paper for the tutorial “current themes in new media”. In this paper I am analyzing whether Fraps, a real-time video/ screen capture software program, is an appropriate tool to use to collect data for game research. In this paper it is my objective to see if new methods can be found in the field of game research that can replace old methods for data gathering in the humanities department.

As many game researchers have argued, for example Aarseth and Yee, that in the past games have been researched through various disciplines. Psychologists research if children behave aggressively in real life after playing a violent game and economists research the influence of ingame economies on real-life economies, etc.. Game studies are a very young discipline, so it needs its own methodologies, claims Espen Aarseth. Ian Bogost does a wonderful job in his book “Unit Operations. (A great review of this book can be found here). Bogost comes up with an excellent new way of researching games, by mixing computation, literary studies and philosophy.

I would like to step away from methodology and have a look at method. Since games are digital and interactive, other methods may be more suitable for data gathering where the interaction of a player with the selected game is being researched, than qualitative methods such as observation and interviews.

In Unit Operations Ian Bogost explains that:

In this process, the unit operations of a simulation embody themselves in a player’s understanding. This is the place where rules can be grasped, where instantiated code enters the material world via human players’ faculty of reason. In my mind, it is the most important in the study of a videogame…. If the experience of a game takes place in the player’s mental model of its unit operational rules, then game criticism would do well to give voice to these mental models and the ideology they communicate. (page 99 and 106)

I believe that in order to be able to grasp the mental model of a player a researcher needs to combine two methods: a real-time video (screencast) of the play session, immediately followed by an interview. In this interview the researcher plays the screencast of the play session together with the player and asks the player questions about the play session while looking at the screencast.

The method of real-time video capture is more accurate than observation, because all data is being saved. The problem with observation is that a researcher cannot write everything down, so he only makes notes of the things that are interesting to him at the moment of the observation. This way important information may get lost. Using cameras facing the screen or the face of the player may solve this. Only this method is very intrusive for the player. If the player is very aware of the cameras, it may influence that way he plays the game. A software program like Fraps, that records the play session by just recording what happens on the screen, is a much less intrusive method.

The problem with an interview is that is should immediately follow a play session and even than a player may not remember why a certain decision has been taken. It’s impossible for a human being to remember every situation that happened in an hour. By watching the video of the play session with the player immediately afterwards, helps the player a lot to recall the situation and explain why a certain decision had been taken. The researcher will be able to get more accurate and valuable answers from the player and this will help the researcher to make a mental model of the player’s choices.

I have used the Fraps software during the recordings of 8 play sessions of an hour. I found the software to be extremely useful when everything goes right. But when things go wrong, the data that has been collected is cannot be used at all. During the last two play sessions suddenly Fraps stopped recording and I could not figure out what made it stop. What a bigger problem was, that somehow the game crashed when the Fraps program was opened, so all information got lost of the game and the game had to be start from the start. Fraps would only record a couple minutes and than it would stop recording again. I was not able to get Fraps started again and I checked all reasons why Fraps would stop and I could not find the reason why it stopped. Therefore the last two play sessions where disturbed by the failure of Fraps and the collected data were useless.

My conclusion is that real-time video capture is an excellent method for game research. However, the program should be more stable, so you can really depend on it to work. 1 entire day of work got lost, because the program did not work and the game crashed. This cannot happen, that researcher should be able to control the program and know why the program stopped recording. If the program is more stable, it is an excellent method for game analysis, I believe.

Comments are closed.