Method
The sources, the formula, the comparison groups, the date policy, and what does not get published.
Where the numbers come from
Two sources, both open. Wikidata (CC0) supplies the roster and a notability signal. English Wikipedia (CC BY-SA 4.0) supplies every measurement, record and fight, parsed from the article's own infobox and results table.
Wikidata is not used for the measurements, because it cannot be: it has no arm-span property at all, and competition class is filled in for a few dozen fighters out of twenty-odd thousand.
Deliberately not used: BoxRec, UFCStats, Sherdog and Tapology — proprietary databases whose terms do not permit republication, which routing through a third-party API would not change.
One consequence worth stating plainly, because it is the site's most-quoted figure: reach comes from the English Wikipedia infobox and nowhere else. There is no arm-span property on Wikidata (P2050 is wingspan, for aircraft and animals) and no schema.org property for it either, which is why a reach figure appears on the page under a dated source line and never inside this site's structured data. Reach is also self-reported by promotions and measured to no standard. Treat a difference under an inch as noise.
Dates, and what happens when there is not one
7,034 of the 168,293 bouts published here carry no date this site can establish. They are printed with no year, sorted last in every table, and drawn past a double rule in the record line rather than placed in the chronology. No year is ever inferred for them.
Two classes of impossible date used to reach these pages and no longer do. A year in the future — a scheduled bout, or an editor's typo for a year twenty years earlier — cannot be told apart from the article, so the date is dropped and the bout kept. A year before 1850 is always a stray figure from a neighbouring cell. Both now leave the bout undated rather than dated wrongly.
A pre-1993 MMA date, by contrast, is not rejected. Vale tudo and shoot-style bouts genuinely predate UFC 1, and the records that carry them are correct.
The decade shown on a fighter's page is the decade holding more of their dated bouts than any other — a count, not a projection. It used to be estimated as the decade a fighter would turn 27, which produced eras that cannot exist (MMA fighters filed under the 1950s, boxers under the 2030s) and inherited every wrong birth date wholesale. 4,497 of 5,754 fighters have three or more dated bouts and therefore have a decade; the rest have none, and their pages say so.
24 fighters carry a birth date on Wikidata that contradicts their own bout table — a 1925 birth against a career that ran 2012 to 2017. This site cannot correct that upstream, but it does not republish those dates as structured data and it does not substitute a guess.
No contests, and when sources disagree
A no contest is not a win, a loss or a draw, and it is never folded into one. It is counted and printed separately, so a record of 20–0 with two no contests is shown as exactly that.
This site publishes one source per fighter and does not silently reconcile it against anyone else's database. Where an article's own infobox total disagrees with its own results table, both figures are shown and each is labelled with where it came from. Nothing is averaged. No field is filled in from a second source to make a total balance. If a figure here is wrong it is almost always wrong on Wikipedia too, and fixing it there fixes it here at the next build — which is a slower correction path than an editorial one, and it is the honest one.
Reach advantage
reach advantage = (reach − height) ÷ height × 100
A fighter whose reach equals their height scores 0%. Because it is a ratio it is comparable between a strawweight and a heavyweight, which a raw reach in inches is not.
What each fighter is compared against
A percentile is meaningless without a stated reference group. Every fighter is placed in the most specific group of at least 30 that they belong to:
| Compared by | Fighters |
|---|---|
| division | 5,449 |
| era | 205 |
| country | 86 |
| discipline | 14 |
Where no weight class is on record, the comparison falls back to era, then country, then the discipline as a whole — and the page says so, in those words, rather than quietly comparing a heavyweight against flyweights.
92 fighters had no weight class stated but did list a fighting weight; their division was recovered from the published limits, and only where the weight sits on a limit. Catchweights are left unresolved rather than rounded into a division nobody competed in.
What gets excluded, and why
Of 8,656 candidates, 5,754 were published and 2,902 were not:
| Reason | Fighters |
|---|---|
| no computable finding (no measurements, no record, no dated bouts) | 2,902 |
A fighter is published only when at least one genuine finding is computable — a reach advantage, a finish rate, or a career shape. An empty template with ads on it is not a page.
Known limitations
- Reach is not measured to any standard. Promotions self-report it. Treat differences under an inch as noise.
- Fight lists are parsed from article tables and inherit their gaps; a table missing a bout means this site misses it too.
- The dataset skews toward fighters notable enough for an English Wikipedia article.
- Percentiles describe this dataset, not the sport.
- No page predicts the outcome of a fight.
SOURCE·English Wikipedia infoboxes + Wikidata roster·REVIEWED AUG 2026
