Nine to Five

It reads how long you tend to stay: across every game you've played whose typical hours it knows (free games left out), how often you went past the typical player and how often you left before half their hours. It glows when you often stay past them, dims when you often leave early, and needs five played games to read at all. Then it shows the games you own most like this one, by the same measure as the shelf of games like it, each with your hours against the typical player's.

Why not a forecast for this game: on 36,766 Steam libraries, how a player did on games like one told little beyond how long they stay in games at all, so those games are shown as evidence and never scored. Nothing of the library is kept.