Steam will now use machine learning to suggest games from your backlog. The same algorithm that is used by Interactive Recommender will look at your playtime across various titles and based on that will point you towards similar games.
Introducing Steam Labs Experiment 008: Play Next! 🔬
Got a bunch of unplayed games in your Steam library, unsure what to play next? 🤔
Let the Interactive Recommender system help you choose by predicting which of your games you'll enjoy the most!https://t.co/VrD1ZJgrYd
— Steam (@Steam) February 13, 2020
The system is called Play Next and it will present you with a list of three games from your backlog at a time. You can reroll the list for new variants.
To help you see at a glance what kind of game each of our suggestions is, we include a microtrailer (on hover), and the top tags for the game. If you’ve already played some comparable games, we show some of those too.
The algorithm we use for picking comparable games is also a new system we’re experimenting with in this lab, and it’s a work-in-progress. So don’t be surprised if it sometimes shows you some unexpected comparison points.
Valve via Steam community forum
Just like with other Steam Labs innovations, Valve will have to see how users respond to the Play Next feature before deciding its ultimate fate.
Steam Labs rolled out in 2019. It allows users to check out experimental features designed to increase games discoverability. Steam hopes to use customers’ feedback to tweak the features on the fly before officially implementing them. Steam Labs, in itself, is one such experimental feature.