On an average day at Electronic Arts, players of the video game publisher's hit title "Battlefield 3" create over one terabyte (TB) of data. In the course of a month, the company will collect more than 50 TB from its respective titles.
For a long time, that information wasn't immediately harvested. Certainly, it would be sifted through as planning began on a sequel, but that was about it. Eighteen months ago, however, the company realized it was ignoring valuable data, and launched a program to change that.
Today, that program is quickly becoming one of EA's most valuable assets.
(Read more: The sexiest job of the 21st century—data analyst)
"The level of sophistication needed in the interactive gaming space is under-appreciated by anyone not in the industry," said Rajat Taneja, executive vice president and chief technology officer at the company. "The complexity of the data and the volume of the data...is more sophisticated than anything else out there in the industry. We are in the very early innings of our investment here. Looking 10 years ahead, data will be the core lubricant of most games in the online economy."
Data analysis in games is currently being used in a variety of fashions. The most obvious example is with game play itself. As developers and data analysts examine how players interact with the game, they're able to make subtle tweaks to improve the experience, something that can increase the time players spend in the game and, in some instances, help boost the number of in-game purchases they make, which gives publishers a revenue bump.
That's particularly true in mobile and free-to-play games. Because players of those titles are used to regular, small patches and updates, it's easier for developers to make changes—and make them quickly.
(Read more: When the boss is big data)
"It's critical to adapt a game very quickly after it launches based on how people are interacting with the game, versus how we imagined it initially," said Keith Kawahata, vice president of Kabam Studios. "It's very easy to make aggregate, broad assessments, but the real trick is slicing that data by a number of different vectors. We look at players socio-economically. We look at players geographically. We look at players by carrier. We look at players by device. And we will modify the game based on that information."
A real world example of how Kabam uses that data is in pricing. Players in one area of the world—or using a particular operating system—might be willing to spend more for in-game purchases, while others are reticent to do so. By adjusting the pricing in areas that spend less, the company can convince them to loosen their purse strings.
Similarly, players of different genres have different spending habits. Players of racing games have shown a broader willingness to pay for items than those who opt for strategy games, Kawahata said, but they do so only at a lesser amount.
The overall amounts of data being generated by gamers is staggering.
Riot Games' "League of Legends," for example, generates over 500 gigaytes of structured data and over four TB of operational logs per day—information the company has a team of roughly one dozen analysts pouring over regularly to improve the game. (It's a strategy that has worked: "League of Legends" has over 32 million active players each month— who play for over 1 billion hours each month.)
(Read more: Big data's intelligence on Snowden)
With last year's launch of "Halo 4," Microsoft incorporated tools to collect large amounts of data from the game, letting the company analyze how gamers were playing the title—and allowing the data and design teams to react.
Speaking at this year's Strata conference, Dave Campbell, CTO of the server & tools division at Microsoft, said, "Once the data team figured out what they could do, they took that to the designers... and it took some time for designers to realize that they could get answers to questions that they never really figured they could get an answer to that quickly. So they could turn things around in a couple of hours and actually do something about it versus the context where it didn't matter, you were going to ship a game and the only chance you have to update it is a year or two later when you changed the title."
Beyond game play improvements, data analysis is also proving useful in customer service.
"From the moment data is generated to when it is actioned, we want it to be as close to real time as possible," Taneja said. "We want to gain insight and prediction on game play—as opposed to looking at a description of what's happened in the past. So when [players] call us for support, we can have a full 360 [degree] view of their game play."
Additionally, by examining the data generated by users, publishers are better able to recommend other games they might enjoy.
While EA has collected data sporadically in the past, it was never a streamlined process. Now, with big data as a core component of the company's strategy, all future titles will filter information to the big data team, where it will be analyzed, then shared with developers responsible for maintaining (and updating) current games as well as those working on future versions of those franchises.
"It's game play understanding at a macro level," Taneja said. "And by understanding how gamers interact when playing by themselves or with friends—there is a tremendous amount of insight we can gather and share."
_ By Chris Morris, Special to CNBC.com