In the lead up to the recent successful launch of the EPSRC Centre for Doctoral Training in Intelligent Games and Games Intelligence, we have been formalising the concept of Game Intelligence. Some of our ideas on this concept were presented at the IEEE 2014 Conference on Computational Intelligence and Games and are available in our vision paper published in the conference proceedings. In this blog post we offer a quick summary of the work that involved a large survey of public game data sources and existing games for education and scientific discovery.
We define Game Intelligence as useful knowledge gained through playing games. This can arise directly for the player through in-game learning, or indirectly through analysis of the gameplay data of other players. Given the digital nature of many modern games and the growing ubiquitous presence of an internet connection, the analysis of vast quantities of gameplay data is now possible. Typically, existing applications of game data mining have focussed on better monetization or game design. However, a few proof of concept examples show that the analysis of in-game data could have far larger impact and would be an interesting direction for future research.
The conclusions of this study note several areas we are keen to promote improvement in as part of the NEMOG project. Specifically, we want to see more:
Educational games teaching advanced concepts to accommodate the increased average age of gamers;
Scientific discovery games for topics outside of Biology;
Game data mining studies extracting knowledge useful outside of the context of the game itself;
Professional game developers making games with social and scientific benefit.
The last of these goals, may require new business models for serious games / games with a purpose. A topic we will be discussing in depth at our upcoming event at Cass Business School in December.
A significantly extended and updated paper covering our vision for Game Intelligence is currently in preparation and will be available soon.