Bitcoin (BTC)
Polymarket BTC Data
How Bitcoin prices, trades, and order books moved on Polymarket, captured every 250 milliseconds from 2024 onward. BTC markets make up a large share of the archive's crypto contracts, and they are the most-measured of any asset in our published research.
Figures measured as of 2026-08-26 on the published PolyOrderbooks archive.
What's in the archive
The BTC market types the archive records
BTC appears across several instrument types, and the archive records each at the same 250ms cadence with the full order book attached. Up/Down contracts on 5-minute, 15-minute, and 4-hour windows are the most common, at roughly 62% of the crypto portion of the archive. Price thresholds ("BTC above $120,000 by Friday") make up about 17%, and event-linked crypto markets tied to macros or headlines about 14%.
The remaining contracts sprawl across longer-window variants and correlation or ratio markets. Every row carries the same columns: bid, ask, and mid prices, top-of-book depth on both sides, and cumulative traded volume, stamped with a wall clock and a 1-based sequence number.
Measured
What BTC five-minute books look like at settlement
The most detailed published measurement of the BTC archive covers settlement behavior. On 8,349 BTC five-minute contracts that settled between 2026-08-26 and 2026-09-23, a 298-contract resolved sample with 17,880 one-second snapshots showed books ending one-sided at least 90% of the time. The winning side rested at 0.99 or better in 98.6% of observations, with a median top-of-book size of 42,254 shares and a 90th percentile of 65,081.
The two sides were never simultaneously empty across 17,880 snapshots — the book does not vanish, it tilts. A strategy that trades 0.99 to a 42,000-share bid in the final minute is not fading weakness; it is joining the side that already knows the answer.
Measured
Strike ladders can reprice in a single second
BTC strike ladders move together. In a documented instance, every rung from 0.044 to 0.996 repriced on the same one-second tick, and a 99.6-cent chasm in the middle of the ladder produced the appearance of a 0.50 repricing that was not executable — the mid sat over an empty book.
The lesson for backtests is that price alone is a lossy representation for BTC on Polymarket. You need the book that sits behind the mid to tell executable from notional.
Schema
Fields on every BTC row
- timestamp (RFC3339, UTC) and sequence number
- asset: BTC, and the contract ID
- bid price, ask price, and mid
- top-of-book bid and ask depth in contracts
- cumulative traded volume for the contract
- market type flag: Up/Down, threshold, or event-linked
Access
How to get the BTC data
The BTC series is included in the full archive, the order book history download, and the API. For live BTC prices, Polymarket's own free API is usually the right answer; the archive exists for what the free API cannot give you: depth and history. A free sample covering any single market or a slice of the strike ladder is available without an account.
Practices
The BTC workout
- Pull one day of BTC Up/Down contracts at 250ms and separate the books into the five-minute contract frames; one day holds 288 contracts of active book and the final-second behavior repeats 288 times.
- Compute the per-contract one-sided share over its final minute; the measured corpus (8,349 contracts between 2026-08-26 and 2026-09-23, 298 resolved in the sample) shows the high-90s one-sidedness is normal, not anomalous.
- Benchmark the executed size distribution against the median of 42,254 shares and the p90 of 65,081; trades in that band are the ones that dominate the tape.
- Verify the round-grid share (98.6% of frames at 0.99/0.995 is the measured baseline) for your day; a day that departs strongly is a data-recheck day, not an anomaly story.
Deeper
The measured reality
The BTC five-minute market is the venue's most-studied series because it is the largest and most sampled: thousands of contracts, 250ms books, and a strike ladder that compresses to a single level near resolution — the perfect anatomy of a prediction market product.
The measured contrast with the other coins is instructive: on the August 23 event, BTC's lower bound took 37 minutes to fully reprice while ETH, BNB, DOGE, and SOL collapsed in 26 seconds — BTC liquidity is deeper, more patient, and slower to commit, and both qualities read directly from the archive.
Conclusion
The honest takeaway
BTC data is the reference series for every other coin page in the catalog: the 8,349-contract corpus, the 250ms books, and the measured one-sidedness give every claim about Up/Down markets a benchmark to run against.
The patient-repricing contrast (37 minutes versus 26 seconds) is a reminder that liquidity depth and reprice speed are different dimensions; BTC's depth buys it time that the smaller coins do not have.
Start any BTC study from the final-minute flags and the executable floor; the measured patterns — one-sidedness, round-grid pricing, median size — will reproduce against your own day, which is how you know the pipeline is right.
FAQ
Does the archive include BTC 5-minute Up/Down markets?
Yes. 5-minute, 15-minute, and 4-hour Up/Down contracts are the core of the crypto portion, roughly 62% of it, with full order book snapshots every 250ms. The published nine-week settlement study alone drew on 8,349 settled BTC five-minute contracts.
What fields come with BTC order book history?
Timestamp, sequence number, contract ID, bid, ask, mid, top-of-book bid depth, top-of-book ask depth, and cumulative traded volume. Depth is the part Polymarket's free API does not serve.
Is the free Polymarket API better for BTC prices?
For live prices, usually yes. The archive exists for depth, history, and research-grade reconstruction — the sample is free and needs no account.