Set a useful starting point
During onboarding, choose interests that describe the mechanics or moods you actually want. You can adjust what you look for as your taste changes; an old preference does not have to define every future session. Start with a few meaningful choices instead of selecting everything just to get through setup.
The recommendation system learns from interests, likes, saves, foreground play, replays, shares and skips. These are different signs: saving a title can mean you want to return, while a skip can indicate that a particular suggestion is not for you. Do not assume one action immediately dictates the next card or produces a fixed sequence.
Use feedback controls with intention
When a game suits you, use the available like or save action if that reflects what you want. When a recommendation is clearly unwanted, use Not interested; Undo is available if you change your mind. Those explicit actions are more useful than opening and leaving a title by accident, because they communicate a deliberate preference.
- Choose onboarding interests that match mechanics you genuinely enjoy.
- Like or save games you want the feed to learn from or help you revisit.
- Use Not interested for a poor fit, and Undo if that feedback was accidental.
Review the pattern, not one recommendation
Give the feed several clear signals before deciding whether it reflects your taste. If recommendations seem broad, interact with a few examples from distinct styles and use explicit feedback on the ones that miss. The service supports varied discovery, so seeing an unfamiliar game does not necessarily mean your stated preferences were ignored.
You can still browse the games catalogue or open a direct game page when you know what you want. Those routes complement the feed: recommendations offer another choice, while search and direct links put you in control of a particular title. Use whichever path fits the moment instead of trying to force every discovery through one surface.
Questions about shape game recommendations
Will liking one game make similar games appear next?
No exact next-game result is guaranteed. Likes are among the signals used to shape recommendations over time, alongside interests and other play activity.
Can I correct an accidental Not interested action?
Yes. The product provides Undo for that feedback. Use it when the unwanted signal was accidental, then continue giving clear preference signals.