Valve has announced a new featured section of its service called Steam Labs, which will give users a chance to test out different works-in-progress projects from the company and give feedback. To start with, the company is launching with three experiments, but more will be added later.
The three launch experiments are Micro Trailers, the Interactive Recommender, and the Automated Show. Micro Trailers are (as you might expect) ultra-short six-second game trailers. The Interactive Recommender uses machine learning to recommend new games based on your top-played games. The Automated Show is a half-hour showcase of the latest Steam games, generated through Steam’s algorithm.
While the trailers and automated show are focused on showing video clips in one manner or another, the Interactive Recommender is personalized to each user. Valve boasts that the model is based on millions of Steam users, and you can filter results with a series of sliders. But in a bit of behind-the-scenes information, Valve also explained that it doesn’t feed any data about the game into the system except for the release date. Instead, the machine learning will gather data on its own and pick out similarities, along with trends among buying habits with players who have similar gaming tastes. The new recommender won’t replace current recommendations on Steam, as right now it’s just in the Steam Labs phase.
Valve says it welcomes feedback from users, including which Labs experiments are worth pursuing further and how to further develop them. It has also opened a dedicated Steam Labs page where you can see all of the current active experiments.