Methodology
How studio scores, format hubs, and "vs" verdicts are computed. No paid placement, no opaque weights — the math is here, the data is the public catalog.
The composite score
Every studio with five or more scenes in the catalog gets a 1.0-5.0 composite score. Scores come from a weighted blend of five dimensions, computed nightly from the public catalog data:
Dimensions
Content (30%)
Catalog depth — total scenes from the studio. Smoothed via smooth(video_count, 100) so the curve saturates: 50 scenes ≈ 4.0, 300 ≈ 4.7. A studio with 1,000 scenes doesn't dominate one with 500.
Visual quality (30%)
Share of the catalog tagged 4K, 5K, 6K, 7K, 8K, or 12K. 70%+ hi-res gets a perfect 5.0; below 30% is hard-capped near 3.0. This single dimension separates studios still shipping 8K from those whose catalogs are stuck in 1080p HD.
Variety (15%)
Distinct category tags across the studio's scenes. Smoothed via smooth(distinct_cats, 30). A studio that shoots one genre exclusively scores lower than one that crosses formats and themes.
Freshness (15%)
Days since the studio's most recent release. <60 days → 5.0, <180 → 4.5, <365 → 4.0, <730 → 3.5, older → 3.0. Studios that stopped producing two years ago can never score above 3.0 here regardless of how good the back catalog is.
Format mix / extras (10%)
How many of POV, 180°, 360°, passthrough, and passthrough-AI the studio has at least one scene in. Five formats → 5.0, three formats → 4.25, one → 3.5. Rewards studios building across the format spectrum.
Verdict labels
- Excellent — composite ≥ 4.7
- Great — composite ≥ 4.3
- Good — composite ≥ 3.8
- Solid — composite ≥ 3.3
- Mixed — anything below
Head-to-head verdicts (/vs pages)
On a "studio A vs studio B" page, the page declares a winner if the composite gap is > 0.15. Below that threshold, the page calls it a "tie" and points the visitor at the category mix and per-dimension breakdown to decide.
Format hubs (/studios/by-format/[fmt])
Each format hub (8K, passthrough, POV, 180°, 360°, 4K) lists studios with at least N scenes in that format. The gate (10 for 8K, 3 for passthrough, 20 for POV, etc) is tuned per format so the list is informative — passthrough is rare so the bar is low, POV is common so the bar is high.
What is not computed
We deliberately do not score performer attractiveness, scene "hotness", or any subjective dimension. Those are personal. The score answers "how strong is this studio's catalog as a body of work?" — not "is this scene exciting to you?".
Affiliate payouts are not an input. The math runs identically whether or not a studio has a partner program.
Auditing the math
The implementation is at src/lib/scoring.ts in the public catalog repo. Every studio review page exposes its per-dimension breakdown so you can sanity-check the composite against the components. If you think a studio is mis-ranked, the most likely cause is missing format tags in the source catalog — open an issue and we'll fix the tagging.