Polymarket Makers and Takers, Explained

• Maker: an order that rests on the book — a limit order at or behind the current price, adding depth.
• Taker: an order that crosses the spread and immediately fills against resting size, removing depth.
• Polymarket documents maker rebate mechanics on the official CLOB, so providing liquidity can pay instead of cost.
• In a recorded row you see the result: maker-resting depth before a fill, and the same level reduced after.

The maker side is the book you can see. When markets invert — the measured ETH case dropping bid depth 96% in a minute — it is makers leaving before takers ever arrive. Liquidity disappears in the maker book and the price follows.

For backtesting, taker vs maker changes fill realism: a taker fill at the ask is immediate; a maker entry at the bid waits on the queue. Replaying against recorded depth models both; replaying against mids models neither.

The cross-venue style of analysis extends here: cross-marker studies (Polymarket vs Binance, e.g. the ETH same-second study) rest on the same maker/taker logic on both venues.

Recording at 250ms keeps maker depth over time — the resting size every taker paid into. That is the dataset story behind the alignment of prices, metrics, and books in one archive.

To study the rebate/reward trade in practice, pair the recorded book with the on-chain fills: makers reward resting size, takers pay for immediacy, and the tape shows who did which.

• From the tape, classify fills by direction against the resting book direction: a fill that removes resting liquidity is a taker, and a fill that adds a level is a maker; the archive tape plus books disambiguates both.
• Count maker versus taker share in a reprice window; the measured pattern (median trade 42,254 against thin tops) implies most volume is taker against small resting size.
• Compute realized spread to a taker: fill price minus the executable floor at receipt, per fill; the page's own fill study found slippage at reprice spikes.
• Record the resting side before and after your fills to confirm whether liquidity replenished or drained; the drain pattern (final-minute 3-of-18 ask presence) is the extreme.

Maker and taker are roles relative to the book: makers rest liquidity, takers consume it, and on Polymarket the balance is lopsided because resting size is usually thin and one-sided.

The measured venue texture — high-90s one-sided frames, reaffirmed round-grid pricing, and fills that slip at reprice spikes — is the description of a venue where takers dominate and makers earn the (tick-thin) spread for taking the one-sided risk.

The realistic simulation for a maker candidate on this venue is: quote one tick inside today's two-sided gaps, replay your quotes against the 250ms frames, and count how many fills you would have earned on the reprice side versus how many you would have eaten on the one-sided one.

Maker and taker are defined by the book, and the archive's paired books-and-tape make the classification auditable.

Themaker's edge on this venue is the reprice flow and the tick-thin two-sided gaps; the taker's cost is the one-sided floor-to-bound gap.

Simulate on the archive before quoting: replay your maker logic against the frames and measure realized spread per fill.

The maker-taker split is the role lens of the tape; execution and market-making pages build on its classification.

Pair it with the slippage and mid-price pages for the trading mechanics, and with market-making for the strategy side.

The measured fill behavior (clean fills, reprice-spike slippage) is documented through this lens.

Maker and taker are roles defined by the book, auditable from the paired tape and books in the archive.

Simulate before quoting: replay your maker logic, measure realized spread, and let the frames decide.

• Order book data
• Price history
• Crypto data
• Provider matrix