Your Personal Taste Match Score

A movie can be critically acclaimed and still bore you. It can have a 4.5 on IMDb and still be exactly your kind of guilty pleasure. Aggregate ratings tell you what most people think. They do not tell you what you will think. That gap between crowd consensus and personal enjoyment is exactly what the taste match score fills.

CineMan analyzes your viewing history, extracts patterns across 10 distinct tag categories, and produces a single percentage for every movie in the catalog. A film with a 91% taste match means its DNA closely aligns with movies you have already watched and enjoyed. A 34% means it sits far outside your usual preferences. Neither number says anything about objective quality. They are about fit.

How Taste Scoring Works

When you first install CineMan and let it process your Netflix history, the extension scrapes your viewing activity directly from Netflix's interface. It then enriches each title with detailed metadata from TMDB, including genre tags, style descriptors, plot themes, cast and crew information, keywords, and more. From this enriched data, CineMan builds a weighted taste profile that represents your viewing tendencies.

When you encounter a new movie on any streaming platform, CineMan compares that movie's tag profile against yours. The comparison happens across 10 separate categories, each contributing a different weight to the final score. The result is a single percentage that reflects alignment between your historical preferences and the movie's attributes.

The 10 Tag Categories

Each category captures a different dimension of what makes you like or dislike a movie. Here is what CineMan tracks and how much each category contributes to the final score:

• Style (20%) — The filmmaking approach and tone. Are you drawn to visually stylized films, gritty realism, documentary-style narratives, or polished studio productions? Style captures how a movie feels aesthetically, separate from what it is about.
• Plot Themes (18%) — The narrative structures and thematic elements. Revenge stories, coming-of-age arcs, heist narratives, survival tales, political intrigue. This category identifies the story patterns you gravitate toward.
• Genre (16%) — The broadest classification. Action, comedy, drama, horror, sci-fi, romance, thriller, and their many sub-genres. Genre matters, but less than you might think, because many people enjoy films across multiple genres.
• Keywords (12%) — Specific content markers that do not fit neatly into other categories. Time travel, artificial intelligence, small-town settings, ensemble casts, unreliable narrators. Keywords capture the granular details that often drive personal preference.
• People (12%) — Directors, actors, writers, and composers you consistently watch. If you have seen five Denis Villeneuve films and rated them all highly, his next project gets a boost. This applies to frequently watched actors, composers known for distinctive scores, and screenwriters with recognizable voices.
• Mood (8%) — The emotional register. Tense, lighthearted, melancholic, uplifting, unsettling, whimsical. Mood is distinct from genre: a comedy can be melancholic, a thriller can be whimsical. This category captures your emotional comfort zone.
• Era (5%) — When the film was made. Some viewers strongly prefer modern films. Others love 1970s cinema. Many do not care at all. CineMan detects whether era is a factor in your patterns and weights it accordingly.
• Pacing (4%) — Slow-burn or fast-paced. This is derived from runtime, scene density data, and genre conventions. If you consistently abandon slow three-hour dramas but finish every tight 90-minute thriller, pacing becomes a meaningful signal.
• Origin (3%) — Country of production and language. Some viewers actively seek out Korean cinema, Scandinavian noir, or Bollywood productions. Others stick to English-language films. CineMan detects these patterns without imposing them.
• Production (2%) — Studio, budget level, and production context. Independent films versus major studio releases, A24 versus Marvel, low-budget horror versus blockbuster spectacle. A small but sometimes meaningful signal.

Why the Same Movie Gets Different Scores

Consider a film like Blade Runner 2049. It is a visually stunning, slow-paced science fiction film directed by Denis Villeneuve with a moody atmosphere and philosophical themes. Here is how two different viewers might see it scored:

Viewer A has a history heavy in sci-fi, loves slow-burn films, has watched three other Villeneuve movies, and gravitates toward atmospheric, visually driven storytelling. Taste match: 94%.

Viewer B mostly watches romantic comedies and fast-paced thrillers, has never seen a Villeneuve film, and tends to abandon anything over two hours. Taste match: 31%.

Both scores are correct. They are not measuring quality. They are measuring fit. Viewer B might watch Blade Runner 2049 and discover a new favorite. But statistically, based on everything CineMan knows about their patterns, the probability is low.

Profile Maturity Tiers

The accuracy of your taste match depends on how much data CineMan has about your preferences. The system tracks this through maturity tiers:

• Starter (10-29 titles) — CineMan has enough data to generate basic scores, but they may shift as more history is processed. Genre and broad style preferences are detected, but nuanced patterns are not yet reliable.
• Developing (30-69 titles) — Core preferences are established. The score starts to reflect genuine patterns rather than small-sample noise. Most tag categories have enough data to contribute meaningfully.
• Confident (70-149 titles) — The profile is robust. Scores are stable and rarely change dramatically when new titles are added. Niche preferences (specific directors, mood patterns, era tendencies) are well-calibrated.
• Expert (150+ titles) — Full resolution. All 10 categories are contributing at their intended weights. The profile captures subtle distinctions like your preference for a specific subgenre within a genre, or your tendency to enjoy a particular director's early work more than their recent output.

You do not need to reach Expert tier for the scores to be useful. Even at the Developing stage, the taste match outperforms generic recommendation algorithms because it is built from your actual behavior rather than collaborative filtering.

Privacy: All Local, Always

Your taste profile is stored exclusively in chrome.storage.local on your device. CineMan does not have a server that stores user profiles. There is no database of taste data, no collaborative filtering across users, and no way for anyone (including CineMan's developers) to see your profile. The entire computation happens in your browser.

This is a deliberate architectural choice, not a cost-saving measure. Taste profiles contain deeply personal information about what you watch and enjoy. That data should not exist on someone else's server. For more detail, see the Privacy page.