A More4 television series in the UK recently asked the question of what makes a masterpiece. It set out to explore if it was possible to use the latest scientific methods to help understand why some creative works are better than others. Focusing on art, music and movies, it featured experts who use various techniques to explore how the body reacts when exposed to creative content. Using methods such as fMRI (blood flow in the brain), EEG (electrical activity in the brain), GSR (electrical conductivity of the skin) and eye tracking, they demonstrated that it was possible to identify that people do indeed react physically differently depending on the quality of creative works, and in most cases, give a plausible explanation as to why.
Not all techniques use sensors, however. Software called Hit Song Science analyses the structure of a song and gives it a numerical score. When viewed graphically, songs which have been commercial successes fall into clusters, meaning they share some common properties. Songs in the same cluster are not necessarily in the same genre - for example, Beethoven and Coldplay could be in the same cluster - rather, they share similar structural elements such as chord progression. The creators claim that this software predicted the success of Norah Jones before she was famous, and which Maroon 5 song would first become a hit.
Hit factory
So if it's possible to analyse the structure of a song to predict a success, and it's also possible to analyse people's bodies to understand what they like, why isn't this done in the videogame industry to help produce better games?
Some studios do currently take some of these approaches, the most famous being Valve. An important part of its development process is to understand player behaviour and emotion during all playtests and use some of the physiological sensors mentioned above. Does this work? Well, Valve's games regularly score around 90 per cent or more on Metacritic and the studio has a very loyal fan base. Making the effort to understand and refine how your players are feeling when playing a game isn't the only factor in making the best games possible, but it certainly helps.
We also use physiological sensors during playtests, and often visualise players' emotional journeys using a method we call 'Biometric Storyboards'. Essentially, we show how the player feels second-by-second throughout the span of a game. Did the player really react to that set piece? The answer will be shown clearly, and with a follow-up post-interview, it's possible to also understand why. The results are sometimes incredibly surprising. What you think (or hope) players will react to most is often far from the obvious.
Designer challenge
It's not surprising that the best designers want to understand their players more and craft a better player experience. Peter Molyneux has said that Lionhead uses a new development method that resembles film storyboarding, allowing them to plan the player's entire journey. He suggests that there is an optimal path through the storyboard, a path he refers to as the 'golden line', which charts the 'ideal' player experience.
Other methods aren't so systematic - Ken Levine, the creative director of upcoming Bioshock Infinite, has recently spoken about how he tries to put himself into the mind of other players to understand how they would experience the game.
So here's the million dollar question facing game creatives: Knowing that there are available scientific playtest methods which could help validate the intended player experience, would they wish to see evidence that either verified or refuted their design?
As game players, we would almost certainly benefit from a more scientific approach to understanding and refining gameplay. Knowing that the game we were playing was what the designer intended could only add to the experience.
But would we be in danger of designing by science rather than creativity? I don't see it that way at all. Game design is indeed a creative endeavour, but one which can be enhanced by science.
Image taken from: Worth 1000