One of my favorite movies of the year is Moneyball, which has been nominated for multiple Oscars including Best Actor for Brad Pitt, who has done more with this movie to make statistics sexy than a million analytics wonks. The movie is based on the Michael Lewis book of the same name, and tells the story of how Oakland A's general manager Billy Beane employed analytics to assemble one of the most competitive teams in pro baseball even though he had a minuscule budget that was dwarfed by the spend of other teams. The sabermetric techniques used by Beane and his team went on to redefine the way players are measured and valued. His approach balanced the art of baseball with science.
Baseball is not the only game to undergo this transformation. In the last few years, the rise of social and free-to-play gaming, championed by companies like Zynga, has had a similar impact on the world of video games. Just as Beane's approach was controversial and met with resistance by Major League Baseball, so too is there much gnashing of teeth within the video games industry by those who see the encroachment of science into play as a threat to the art. Critics are quick to label free-to-play mechanics as “casino tactics.” And while some social gaming companies have undoubtedly gone too far in their use, it cannot be denied that the use of analytics is bringing a better understanding of how people play and how players stay engaged.
Analytics will never make a bad game great; they can only help improve on a fundamentally good idea. Analytics can also help us understand how people really behave within games—not how they say they behave, but what their actions are in a real world setting.
I'm a big believer in a wide variety of research methods to better understand the audience, from the totally qualitative to the unabashedly quantitative. However, I am also a big believer in the “House” rule as espoused by Hugh Laurie's character on the Fox TV show of the same name: people lie. This is consistently the problem with qualitative research because when people self report their behavior, they often register their “best selves” or how they want to be seen—not how they behave in the real world. Just as Beane's sabermetric approach was able to find the “hidden value” in baseball players, so too can analytics help us observe things about video game players that they may not even know about themselves.
Consider that the vast majority of games fail (this includes both console and online games). By “fail,” I mean that most players, even those who have paid $60 for a game, never get all the way through a game. When people who have paid a fairly expensive price (especially considering the times we live in) don't feel motivated to play the content they have already paid for, it suggests an incomplete understanding of the player. As games have moved online, we can now measure and test in real time various scenarios for improving or optimizing.
While some creatives within the games industry see the use of analytics as threatening, I think there is great opportunity for it to be empowering. This debate reminds me of a similar uproar in the industrial design business in the late '90s and early 2000's over the use of research in product design. Similar to today’s video game designers, old school product designers saw research as an affront to creativity, and an anti-innovative crutch used by fearful and risk averse marketers. But the evolution towards ethnographic and other generative research tools empowered a new generation of designers to get closer to the people they were designing for, and I don't think there is anyone in that industry today who wouldn't say that better products are being created as a result of that research.
Much the same will happen with the games industry. For sure there will be pitfalls for those who glorify statistics over all else (you can already see some who are producing games that are too clinical or too pushy, and have prioritized math over fun). There should always be a healthy, natural tension between art and science, and I believe the industry will find a middle ground. But as games make the transition from a predominantly product-driven model to real time, connected services, analytics will have to become as important as game design, and indeed an intrinsic part of it. Just as Billy Beane changed the game of baseball, so will the Billy Beanes of video gaming forever change the way we design for fun.