A transparent explanation of the viewing signals, show attributes, related-show evidence and safeguards behind TV tiempo recommendations.
Recurring genres, networks, keywords and people across genuinely tracked shows become weighted evidence.
A candidate gains evidence when your shows recommend it and when it recommends shows already in your profile.
Favourite shows strengthen both their attributes and their related-show connections without overwhelming every other signal.
Dislikes, drops and blocks reduce similar candidates, while ordinary skips remain neutral so browsing does not damage the profile.
Ratings and popularity can order close results, but they are kept separate from the viewer-specific Taste Match percentage.
Recommendation versions and deterministic pagination keep long sessions consistent while short evidence summaries explain useful matches.
TV tiempo combines your viewing progress, watchlist, favourites and explicit reactions with show genres, networks, keywords, people and recommendation relationships. The result becomes more useful as your television history grows.
Actively watched and completed shows provide the main passive evidence. Watchlist-only and simply skipped shows do not train taste.
Likes and favourites strengthen positive evidence. Dislikes, dropped shows, blocked shows and disliked people provide progressively stronger negative evidence.
The model compares genres, networks, keywords, main credited people, original language, origin country and relationships between shows.
No. Vote quality and popularity are gentle tie-breakers between similarly relevant candidates; they do not create the personal match itself.
TV tiempo omits the percentage until a profile has enough relevant evidence. It does not display a fabricated cold-start score.