A look under the hood at the match score — what it weighs, what it ignores, and why your data never leaves your diary.
It learns from what you log, not from what you can’t control
Every recommendation you see in loggit starts with one number: the match score. It is the percentage on every card in the swipe deck — and the question we get asked most is, simply, where does it come from?
The score weighs the things you have actually told us about: the titles you rated highly, the genres you return to, the lengths and formats you finish. It deliberately ignores the stuff that makes other apps feel invasive — your contacts, your location, your browsing outside loggit.
A friend who knows your taste, not a billboard
A good recommendation should feel like a friend who knows your taste — not a billboard that has been following you around the internet.
That principle decides what we are willing to use. If a signal would make you lower your voice before mentioning it, it does not belong in a recommendation engine. Taste is enough.
What actually goes into a score
Roughly four things: the titles you rated highly and how strongly, the genres and themes you keep coming back to, the formats and lengths you tend to finish, and how a new title relates to things you already loved.
It is a blend, not a single rule. A great match usually lines up on several of those at once, which is why the number feels different from a generic “popular near you” list.
A number you can tune
The result is a score you can actually trust — and tune. Open your Insights, hit Tune, and nudge the weights toward what matters to you this season. Want shorter films on weeknights? Done. More books, fewer games? Easy.
Your diary stays yours. The score is just a lens on it.