IBM and PlayFab will use machine learning to understand player behavior

Computing giant IBM and backend service provider PlayFab have formed a joint initiative to give game developers deeper insights into player behavior. 

The program will let developers feed data into IBM's Watson artificial intelligence platform for analysis, with the system capable of building models from structured and unstructured data, as well as open machine learning libraries. 

Watson will use that information to deliver lessons and feedback directly to devs through PlayFab, allowing them to make meaningful changes to their games in a bid to boost engagement by better understanding players. 

"Modern connected games offer immense artistic and commercial potential, but to realize that potential, developers need tools to analyze and act in real-time on the massive amounts of data these games-as-services continuously generate," said PlayFab CEO, James Gwertzman.

Atari is one of the first game companies to make use of the fledgling initiative, and has been talking up the practicalities of Watson's data-crunching capabilities in the fast-moving, fluctuating world of mobile games. 

"Understanding what players experience, and being able to react quickly, has become crucial to success in operating a modern game," explained Sylvain Constantin, executive producer for Roller Coaster Tycoon Touch.

"This is why we are so excited to now be collaborating with IBM, through their relationship with PlayFab. The Watson Data Platform takes us to a whole new level."