I find it useful, when introducing our work to people who are as yet unfamiliar with biometry, to begin explaining the basic functioning of lie detector machines. When utilizing lie detector machines, the body of a test subject is hooked to several sensors capable of recording changes in a range of physiological processes such as her heartbeat, the electrical conductivity of her skin, the frequency of her respiration, and so forth.
Observing variations in such dimensions, the discipline of biometry is capable of approximating an objective account of the test subject's internal state variations. In other words, by measuring how one's body reacts to a certain experience (which could be a set of questions, or the screening of an advertisement, a video game session, etcetera), we can determine one's level of stress, concentration, anxiety, fear, etcetera.
Whereas traditional quality assurance procedures generate highly subjective answers to research questions, biometry offers a method that produces results that are objective and quantifiable.
By monitoring changes in heart rate, skin conductivity, respiration and the contraction of certain key facial muscles, the lie detector-like setup we utilized was capable of providing valuable insights about game design choices as well as materializing the psychophysiological effects the game has on the players.
Employing biometric experiments and methodologies to analyze a video game, you can obtain scientific answers to several questions that are crucial for its development and commercial success. Examples of such questions include:
The employment of biometric measurements is certainly not a new development in the field of video game design, tuning and testing: Triple-A titles such as Valve's Left 4 Dead and EA Sports' NBA Live 2010 have successfully demonstrated the viability and desirability of biometry as an analytical tool and as a factor of change for their products. Our research project and its benchmark video game Gua-Le-Ni pioneered and optimized the application of biometric technologies and methodologies with the objective of making them available for quicker iterations, exploring their viability as development tools for casual and independent video games.
To complement a wider quality assurance campaign based on questionnaires, interviews, blind-testing and hard-core performance tests, the Dutch research team at NHTV Breda University of Applied Sciences ran an initial series of biometric tests on Gua-Le-Ni. The aim of these initial tests was to structure a testing methodology incorporating the added perspective of biometry.
The first biometric analysis we ran on Gua-Le-Ni focused on its accessibility during the first few minutes of gameplay. The task that was assigned to the researchers was to determine biometrically the optimal speed of the game for the target audience indicated by the developers as soon as the player successfully completed the first tutorial. The game design goal in relation to the initial set of tests was that of achieving the feeling that the game was non-threatening and manageable at the most basic level of difficulty, hence likely resulting in an initially pleasant and positive experience for the casual audience we were developing for.
In terms of game logic, the initial speed of the game is determined by the initial walking speed of the beasts. In this sense, the results of the first test in terms of the walking speed of our bizarre creatures became a cornerstone for all the subsequent design decisions concerning the tuning of the speed and the complexity of the game.
The research team ran parallel tests on two slightly different versions of the game: In the first and harder version, the beasts crossed the screen rapidly (in 24 seconds), while in the second, easier version they would walk from one end of the page to the other more slowly (in 30 seconds). When the testers' stress patterns were analyzed and correlated with in-game questionnaires, we found that -- compared to the harder version -- the easier version showed fewer signs of stress in the participants.
Even in the slower version of the game, however, the recorded stress levels were much higher than expected in comparison with other successful casual games we tested biometrically with the same test subjects. The initial walking speed of the beasts was hence deemed still too high our intended players to simply enjoy the game. We could infer this outcome by combining the biometric data with the self-reported ones. As a result of this observation, the initial speed of the beta game was set -- for the tests that followed -- to 34 seconds. This value was further refined after the second set of biometric tests to the initial speed value of 36 seconds, with which the game was released.
A particular consistency in the Game Over pattern also highlighted difficulties for most testers in recognizing specific creature parts. Responding to the difficulties and the criticism of our test group, the graphics for the creature modules in question were redone.
In addition, we picked the lobster (the beast with the highest rate of stress and failures) as the creature to be used as an example in the tutorial, in order for players to familiarize themselves with its quirky appearance as early as possible in the learning curve of the game.