|Session Name||Evaluating Game Experiments: A More Robust Approach for Freemium Games|
|Track / Format||Business, Marketing & Management|
Data analytics is a core part of running a game's business today. As free-to-play games are living products that span multiple years, a large part of running them involves experimenting with different aspects of the game like introducing new features, improving user flows and changing variables to improve the tuning of the game.To conduct effective experiments, it is essential to understand the distributions of key metrics for players. This presentation will highlights weaknesses in the conventional methodology to evaluate experiments and propose a more reliable methodology that takes into account the actions of the most engaged players in a game. Utilizing the new methodology can significantly improve the accuracy of experiment results.