Video games have grown from singular personal diversions into a market that spans not just the globe but minds. Many games, such as Battlefield and Call of Duty, serve millions of players a day and numbers close to that concurrently. As they play, they leave behind traces of themselves, behaviors, activity, and even their own personalization–all of these digitally recordable as data that tells video game companies how consumers play their games.
In a speech at Stata conference 2013 yesterday EA CTO Rajat Taneja spoke to audiences about how Big Data has inspired EA to upgrade their game.
The video game industry has grown from 200M active users to 1.5B players worldwide–each one producing a vast virtual footprint every time they go onto one of the games that they play. If that isn’t the definitive “big data” then nothing in this industry is. All of these players are using code written to give them a particular experience, one that burgeons on some of the most complex scripting ever seen on this Earth; their experience is exactly what EA is selling and they want it to not just be perfect: they want it to be something that players will return to time and again.
Every time a player logs into Battlefield 3 and loads into a game, they can be tracked from start to finish. All of their activity regarded by the system as a digital footprint that can be analyzed after the fact or while they’re playing. Add together the potential 1.5B players (or the potential million plus concurrently on the system) and that’s a lot of data. So much data that it’s difficult to sift through it without a plan–or an analytics suite.
During his presentation, Taneja mentioned that EA has turned to the ever-popular Hadoop for collating and processing all that information, which is then distilled down using mapreduce in order to allow distinct elements of the gameplay to be visualized. A process that used to take EA days to boil down and recondense now can occur in a matter of minutes; allowing the game company to mix historical data and current models of player behavior to see what’s going on.
Player experience is part of product and data is part of customer service
In 2012, after Amr Adwallah, CTO of Cloudera, spoke about the mixture of Hadoop and gaming, I interviewed him about how Big Data could be used. At the top of the list is the fact that every movement, every bullet, every death is recorded and that could be used to understand why people play games and how they play them.
In his presentation, Taneja showed a “heat map” of a level from Battlefield 3. Heat maps are used to show a statistical visualization of quantity over a space–this one appeared to be about how often players moved through particular regions of the map. Many online shooter games (like BF3) allow players to have their maps chosen for them “randomly” and get paired up with other players. This sort of information could be used to better fit players and maps that they tend to use more–or match up players who have similar or complementary play styles.
From console to mobile to global all under the gaze of Big Data
He also mentioned EA’s upcoming single-sign-on service that will allow players to have a single username/password to authenticate with any EA game irrespective of device: PC, console, and mobile will all use the same source.
When I saw this announcement, it seemed obvious that EA was going with a big data solution to help collapse everything that happened through each of these venues into one place. No doubt, the maps in BF3 won’t mean much to what people do in Mass Effect 3 or Rad Racing or Trivial Pursuit.
Keeping in mind that players who play on PC and consoles also own mobile devices–and as a result also play games on mobile devices.
Pooling resources from PC and console and mobile, EA can and will be able to watch as players move from one device to another–still playing their games–and look at how they approach them. It will mean the company can suggest games to players of BF3 on the mobile device they own that other BF3 players also like. It also means that EA will be able to tell how that audience differs when they’re at home versus when they’re on the go.
Doing so would certainly boost sales for EA in their mobile spectrum by leveraging the already massively profitable console/PC market.
Big Data, Hadoop, and the ability to compare and contrast vast amounts of play (by distilling it down into the activities that correlate to customer interest) will most likely become EA’s next big win.