Polymarket MMA data: fight-night pricing at 250ms

MMA trades on Polymarket as winner and method-of-victory questions on fight nights. Live books are free to read from the official API; historical books are not available from it at all.

The /book endpoint returns only the current ladder, so a post-event study of how a title fight was priced round by round requires depth recorded during the bout. PolyOrderbooks captures full ladders four times per second for the MMA series it lists.

Fight-night contracts trade as binaries like any other market — winner YES/NO questions and method markets — and resolve to the shared archive record: full bid/ask ladders with sizes, a captured_at timestamp, and the winning outcome.

No MMA-specific count is published on this page because the family is measured as long tail; the honest split lives on the sports data coverage page.

• Live ladder — official /book, free and keyless, current moment only (what the API serves).
• Historical depth — full L2 books at 250ms over REST, free Starter with 3 days, paid windows to 120 days.
• Discovery — /v1/series for fight cards; /v1/markets for bout questions and slugs.
• Coverage — measured counts per sport on the sports data page.

Fight night: fighters open the headline bout at a stable line, the pre-fight weigh-in barely moves it, and round one either holds the favorite or ends the book in a single strike. The official API serves that ladder live; the archive keeps the round-one knockout's second-by-second book for replay after the arena empties.

MMA questions span winner and method-of-victory across the fight-night slate, inside the 700k+ long tail with the other counted-together families. A single-night card closes several independent books, which makes the fixture clock — not the calendar — the unit of study.

The behaviors worth measuring against the record: how second-round odds reprice a first-round drubbing, whether underdog method-of-victory lines hold their premium until the judges' cards, and how instantly an early finish settles every co-main book on the same night.

The pull is mechanical — /v1/series for the fight card, /v1/markets for the night's questions, archive window around the main event — and the return is the full ladder with sizes at every recorded second.

The finish clock is the signal: a round-one ending reprises every other book on the card before the next bout even walks out.

As a long-tail family, one card is a case study; a full season in one window is the only scale that supports general claims.

3 days of free 250ms covers a fight weekend; Pro extends to 120 days and resolved cards stay queryable.

Study one: measure how a round-one stoppage reprises every other book on the card, counting frames from the stoppage to each fixture's new executable side.

Study two: compare method-of-victory premiums before and after the weigh-in — the record holds the pre-event book intact for the whole fight week.

Study three: test whether underdog finishes produce a full-band jump within a single recorded second, matching the one-second reprises measured in the crypto studies.

One fight card on Starter starts the analysis; the Pro window to 120 days holds a full schedule of cards for the study that needs more than a single night.

The entry exercise is one fight night: pull the main-event window from the archive and count the executable-price moves across the rounds. A round-one stoppage settles the book mid-stream, and that single event teaches the fixture-clock model of this family.

Then compare the winner book to the method-of-victory book at the same second during a late-round exchange; the premium divergence under pressure is the bout-read in its purest form.

Resolve every bout and keep the card queryable; one card is a case study, and an evening slate of such cards in the Pro window is the smallest legitimate dataset for fight-night conclusions.

• Sports data
• Market data
• Order book data