Polymarket esports data: 31k markets on CS2, LoL, and Dota 2

Polymarket prices the big three esports titles as match winners, map winners, and map handicap markets. Live books are free to read from the official API with no key.

History is where the official API stops — the /book endpoint is a live object with no past. Reconstructing a Counter-Strike map's price ladder after the fact requires capture while the map traded, which is what the PolyOrderbooks 250ms archive provides for the esports series it lists.

The map-and-set handicap shape is the esports tell — 'Map 2 Rounds Handicap: Spirit Academy (-6.5) vs MORROW (+6.5)' is a real Polymarket contract — and it trades as its own book alongside match and round markets.

Every contract stores identically: full bid/ask ladders with sizes, a captured_at timestamp, and the winning outcome on resolution. That uniformity is what makes a multi-map series replayable at 250ms from opening price to final horn.

• 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 CS2, LoL, and Dota 2; /v1/markets for questions and slugs.
• Coverage — measured counts per sport on the sports data page.

A Counter-Strike 2 Bo3 at a weekend major: the first-map moneyline tightens through the pistol round, the second map swings the series line, and the decider reprices the winner inside one force-buy round. Live frames are free from the official API; past frames live only in the archive.

The catalog covers the three big titles — CS2, League of Legends, and Dota 2 — as match winners, map winners, and map handicaps; each map settles independently, so a Bo3 produces a sequence of smaller books before the series market closes.

Three repeatable reads appear in the record: map-two winners correcting the series line before map three, early-game gold leads repricing LoL mats within seconds, and CS2 pistol-round upsets showing as instantaneous single-band jumps on the executable side.

The pull is mechanical: /v1/series for the league and stage, /v1/markets for the match questions, and the archive query on any map window to replay the full ladder second by second.

Map boundaries are the volatility seam: the series line reprices at map transitions, and the first engagements of each map set the new executable side.

At 31k-plus markets esports is a measured family, not long tail, so sample sizes for map-level studies are genuinely usable.

Free Starter holds days of map windows; Pro stretches to 120 and keeps resolved series queryable.

Study one: test whether the series line reprices on a map-two winner within seconds of the final round, using map-ending frames and the venue's one-sided share as the control.

Study two: replay a pistol-round upset in CS2 and count the frames to the first executable-price jump; sub-second reprises are the established pattern in the crypto corpus.

Study three: compare map-handicap books against match-winner books across one tournament weekend to see which market holds the tighter grid at the same moment.

One major weekend at 250ms is the honest first sample — the tournament week fits the Pro window and the deciders stay queryable after the trophy is lifted.

Start with one Bo3: pull the map-one and series windows from the archive and locate the exact frame where the map winner flipped the series line. One pistol round, one recorded second, one executable-price jump — that is the esports unit of study.

Then pull a map-handicap book at an identical second and compare its width to the match-winner book; the handicap grid discipline is where esports diverges from the other quarter-style families.

Make it a weekend habit: one major, one window per decisive map, and every pulled market resolved and queryable, so the weekend becomes a dataset rather than a memory.

• Sports data
• Market data
• Order book data