Critic-mediated collaborative filtering

Stop arguing about what to watch.
Find the film you will both love.

Critic-Match builds your taste profile by measuring how closely you agree with 130 real film critics — then cross-references it with a friend's to surface only the films where both of your top critics agree.

600 films130 critics15,087 real reviews20 films to build your profile
Your profile, step 120 films · about 2 minutes
Joker2019
ActionAdventure68%
Toy Story 42019
AnimationComedy97%
Once Upon a Time In Hollywood2019
ComedyDrama85%

Like, shrug or dislike each one. Skip anything you have not seen — we only use films you actually have an opinion about.

What you get back
sim(u,c) = Σ ru,m·rc,m / ( √Σ ru,m² · √Σ rc,m² )

Your top 10 critics by cosine similarity, each with a real match percentage — then a shared deck scored across both profiles.

Three steps, and the arguing stops

The whole point is that neither of you has to compromise. A film only reaches the top of the shared deck when the critics you trust and the critics they trust both liked it.

Step 1

Rate 20 films

A cross-genre starter set, deliberately weighted toward films with dense review coverage so your profile has real evidence behind it from the first screen.

Step 2

Meet your critics

We compare your ratings against every critic's published verdicts over the films you have both covered, and keep the ten you agree with most.

Step 3

Share a deck

Pick a friend and get films neither of you has rated, scored by both sets of critics. Mark anything you have seen and the queue re-ranks instantly.

Try it immediately

Three demo accounts are already set up

Seeded with genuinely different tastes, matched against real critics, and already friends with each other — so you can open a shared deck in two clicks.

alice · horror & thriller
demo1234
bob · prestige drama
demo1234
carol · crowd-pleasers
demo1234
Provenance, plainly

Where the data comes from

Film metadata, critic names, publications, review text and Fresh/Rotten verdicts come from a real Rotten Tomatoes export (15,087 reviews across 130 critics). Each critic's era bias, mainstream bias, contrarianism and severity are computed from that corpus, not asserted.

Audience scores are supplemented from a Metacritic export and shown for context only — they are never fed into the matching or scoring maths.

Read the full method