Valve to use machine learning for better game suggestions on Steam

Valve announced in a blog post that it will use machine learning on game platform Steam in 2019. With this, the company wants to improve the game suggestions that the platform presents to users.

Valve speaks of ‘Store Discoverability’ and describes this as a new engine for making recommendations. The goal is to better match Steam users to games based on their preferences.

Furthermore, Valve will adjust the game library in the Steam client, connecting in terms of interface and functionalities with the Steam chat released in July last year. This chat feature is getting a new app and the event system is being improved.

The company behind Steam has also announced that Counter-Strike: Global Offensive’s Trusted Matchmaking system will be upgraded and become available for all games on Steam. This means, according to Valve, that players will have more information at their disposal to determine whether a player is a cheater or not.

Valve also released some statistics. For example, based on data from 2018, the platform had 47 million daily users and an average of 90 million active users per month. In 2019, those numbers will probably increase further, partly because Valve is working with Perfect World to release Steam in China. Valve will announce more about this in the coming months.