
I considered titling this post "
Big Brother is Watching You...Game" but that seemed a bit too sensational. It also trivializes the achievements of the IT geniuses whose work I want to tell you about. These researchers are using
Tomb Raider: Underworld to study exactly how gamers play in order to improve the gaming experience.
Anders Drachen, Alessandro Canossa and Georgios Yannakakis of the IT University of Copenhagen will be presenting their paper, "
Player Modeling using Self-Organization in Tomb Raider: Underworld" at the 2009
IEEE Symposium on Computation Intelligence and Games in Milan next month.
They investigated the gaming behavior of more than 1300 players who completed
Underworld in November 2008. According to the paper,
"Data was collected via the EIDOS Metrics Suite (a game metrics logging system utilized by EIDOS). The data collection process is completely unobtrusive since data was gathered directly from the game engines of subjects playing TRU in their natural habitat (via the Xbox Live! web service) rather than in a laboratory [setting]."
Vast amounts of data were collected ("1 million recorded gameplay sessions of TRU which will form the basis for future research," according to the paper), but for this particular study, the researchers whittled that down to 1365
Underworld players' data in three particular areas:
- Number of in-game deaths and their causes (enemies, environment, falling)
- Time taken to complete the game
- How and when each player accessed the help-on-demand (HOD) feature
The paper details the methods used to crunch the numbers. Most of that sailed right over my head. (Hey, I was an English major.) But by analyzing all this information, the researchers were able to sort the players into four major clusters or types:
Veterans – Players who complete the game very quickly; who die very few times and whose deaths are mostly caused by the environment (traps, fire, drowning); and whose HOD requests vary from low to average.
Solvers – Players who take their time; who die quite often, mainly due to falling; and who rarely use HOD. According to the study, "Players of this cluster
are adept at solving the puzzles of TRU. Their long completion times, low number of deaths by enemies or environment effects indicate a slow-moving, careful style of play."
Pacifists – Players who die primarily in combat, whose completion times are below average, and who rarely use HOD. According to the researchers, pacifists form the largest group.
Runners – Players who die quite often, mainly by opponents and the environment, but who complete the game very quickly. Runners' reliance on HOD varies across the spectrum.
So what does this information tell us about game design and gameplay? According to the Danish researchers, "
the existence of four clusters of behavior, even in a fairly linear and restricted game like TRU, shows that players utilize the
space and flexibility offered by the design of the game, rather than simply using one specific strategy to get through the game." So basically the numbers confirm what we already knew intuitively: different types of gamers use different methods to get Lara from A to B.
In the past, smaller-scale player-modeling studies have been used to design more realistic AI for non-player characters (NPCs or "bots") in sports games and shooters, as well as creating interactive stories that change based on player interaction.
So how will game designers use this information to build a better Tomb Raider? I guess we'll have to wait and see, but these researchers envision on-the-fly player tailoring in which "information about the different player types can be used during play to dynamically alter in-game controllable parameters (e.g., help on demand accessibility, difficulty of jumps) to adjust to the needs and skills of the player type identified in real-time and ensure variation in gameplay."
Can't make a jump? Next time, Lara can go a little farther. Can't defeat that boss? Next time its health bar is a little shorter and its attacks do a little less damage. Spending too long on that puzzle? Suddenly the camera pans to show that elusive switch.
Now that would be cool indeed, though I suppose it could eventually put walkthrough writers like me out of a job.
Links: