How Big Data Improved QA for 'Halo 5'
When it comes to high-profile software releases, a major bug can quickly turn a product into a punch line, which is why pretty much every developer on the planet forces their work through endless rounds of QA. Aside from each new edition of Windows and Office, there’s probably no higher-profile product launch for Microsoft than games in the Halo series. In order to minimize the chances of snafus and issues with Halo 5, due for release in October, the team has relied on two in-house Big Data platforms—Trill and Tempe—to quickly process and analyze massive amounts of game telemetry data. For example, analytics allowed the Halo team to quickly determine that people playing the beta were winning a rare medal far too frequently, and adjust the game’s parameters accordingly. The above video offers some interesting insight into how a team can use Big Data to make big software projects—not just games—run far more smoothly.