A statistical portrait of the roster, on every site page
Site pages were good at the granular question (who is on this studio, how many scenes each) and silent on the shape of the whole roster. So we added a Roster analytics panel: median age on set, scenes per performer, share of the roster still shooting, share that appears nowhere else in our catalog, plus hair, ethnicity and nationality splits. Everything is computed at build time from the same enrichment that fills the performer pages, nothing is typed by hand. We deliberately cut the stats that read identical on every studio (median height was 160-165 cm everywhere, most common cup a B almost always) and kept the ones that actually swing site to site: seen-only-here ranges from 2 to 96 percent across our catalog, scenes per performer from 1.0 to 6.8, nationality from all-Czech to mostly-American. Every metric carries a fill gate, so a thin studio drops the rows it cannot stand behind instead of guessing.
Every site page on FapStarsDB now opens its lower half with a panel we are calling Roster analytics: a statistical portrait of the performers we have catalogued on that studio. Not a list of names, the page already had that. An aggregate. Median age on set, how deep the catalog runs per model, what share of the roster still shoots, the hair and ethnicity split, the nationalities that actually show up. It reads off the same profile data we collect for the performer pages, rolled up to the studio.
Why a site page needed this
A site page on FSDB was good at the granular question (who is on this studio, how many scenes each) and silent on the shape of the whole. If you landed on a studio cold, you could scroll a grid of faces and learn nothing about what kind of roster it is. Young or mixed-age? A handful of regulars with deep filmographies, or a long tail of one-shoot drop-ins? Mostly local talent or syndicated across the industry? That last one matters a lot to anyone deciding whether a site is worth their time, and nobody publishes it.
We had every number sitting in the database already. It just had nowhere to land on the page.
What is in it, and where the numbers come from
The top row is five tiles. Share of the roster seen only here (nowhere else in our catalog), scenes per performer, share filmed in the last year, the age span on set with a median, and the share of D-cup or larger. Under that, two bar charts for hair colour and ethnicity, then a row of nationality chips. Every figure is computed at build time from the performer profiles, the same enrichment that fills hair, age, measurements and nationality on the individual pages. Nothing is typed in by hand, and nothing is cached. Rebuild the site and the panel recomputes against whatever the data says that day.
We threw out the numbers that look the same everywhere
The first draft had more tiles, and most of them were useless. We pulled the stats for all 166 studios that qualify and looked at the spread. Median height came back at 160 to 165 cm on basically every site. Most common cup size was a B almost everywhere. A stat that reads identical on every page is not information, it is filler, and it makes a page look auto-generated. So those got cut.
What stayed are the dimensions that actually swing studio to studio. Share seen only here runs from 2 percent to 96 percent across our catalog. Scenes per performer goes from 1.0 on a casting site that books each girl once to 6.8 on a deep archive like Watch4Beauty. Share filmed in the last year spans 3 percent to 61 percent. Nationality is the loudest signal of all: one site is a wall of Czech names, the next is mostly American, another leans Colombian and Venezuelan. Those are facts that tell you what the studio actually is, and they are the ones we lead with.
When the data is thin, the panel shrinks itself
Enrichment is not uniform. Some studios we have indexed deeply, others are barely filled in. So every metric carries a fill gate: a stat shows only when we hold that attribute for at least half the roster, and the whole panel stays hidden under eight performers. The point is to never print a median age computed from three models and pass it off as the studio average.
You can see it work above. On a 37-performer studio where birthdates and measurements are mostly missing, the age, cup and ethnicity rows simply do not render. Three tiles, a hair chart and a clean nationality row remain. We would rather show less than guess.
A couple of honest caveats
The active-roster number drifts on its own. It measures who filmed in the last twelve months relative to today, so a studio that stops releasing will watch that figure slide and eventually the panel calls it dormant. That is by design. Ethnicity is bucketed coarsely, five groups, not the full granularity we keep per performer. And the whole thing only reflects what we have enriched, which is why the copy under the heading says every share is measured against the performers we hold that attribute for, not the whole roster.
This is the next layer on the site pages, after the performer-block changes we tested a few weeks back. The featured strip and the catalog context answered who works the most here. The roster panel answers what the roster is. The numbers it leans on are also what feeds the FapStars score, so if you have read how we rank, none of the inputs will be a surprise.