Polymarket Bid-Ask Spread, Explained

• The best bid and best ask are the top two rows of the book; the spread is ask minus bid.
• Prices step in 0.01 ticks, so the tightest common spread is one tick.
• Depth at the inside determines how much size the tight spread actually serves.
• Crossed states — best bid above best ask — occur in fast repricing and are left flagged in recordings rather than filtered.

The empty-book spread: a measured BTC contract printed a 99.6-cent-wide chasm in the middle of its strike ladder, making a nominal 0.4995 mid with a 0.998 ask — tradeable-looking, nothing there.

The inversion spread: an ETH book lost 96% of bid depth and multiplied ask depth almost sixfold in a minute, so the spread exploded before the price followed. Neither move exists in a mid-only series.

Both are why measured spread studies on this site use the recorded ladder, where spread width, depth between levels, and crossed flags are rows, not anecdotes.

For a live check use the official GET /book: read best bid and best ask directly. For research, spread history belongs with the book: query 250ms rows and compute width over time, because a minute-average spread hides the gap that existed for nine seconds.

The spread is the price of immediacy. When the spread is wide, you are paying for speed or standing aside — and the recorded book is the only way to know which you would have done historically.

• Pull the books for an Up/Down and a sports market at 250ms and compute the per-frame top-of-book spread (ask minus bid); print the median spread per series.
• Mark the frames where the spread is effectively unquotable because one side is empty; the measured one-sided share (90%+) makes these frames the majority, and your count is the analysis.
• For a two-sided frame, record whether the executable gap is one tick (0.5 or 1 cent) or wider; the measured round-grid tightness (98.6% at 0.99/0.995) predicts most two-sided frames are one-tick.
• Add a reprice window to the same spread series and watch the spread blow out in the seconds of the ladder pass; the 99.6-cent move is the extreme form of spread widening.

The spread on Polymarket is mostly a full-frame minority event because one side is usually missing: the "spread" that matters is the distance between the executable floor and the far bound, which can be enormous exactly when the other side exists.

The round-grid tightness is the positive side: when both sides rest, the gap is one tick and the venue behaves like a tight, honest market — the measured 98.6% at 0.99/0.995 is that behavior in a single number.

Spread on Polymarket means two different things — the two-sided tick gap (tight) and the one-sided floor-to-bound gap (huge) — and any analysis should state which.

Use the one-sided share and the two-sided tick-gap median as the pair of numbers describing a market's spread regime.

A reprice compresses the whole story into a second; the ladder frame at that second is the spread's true extreme.

Spread is the price of immediacy and this page is the bridge between the book explainer and the execution pages.

Read it with the maker-taker and slippage pages for the trading loop, and with bid-ask data for the raw series.

The two-spread distinction (two-sided tick gap vs one-sided floor-to-bound gap) is used by every execution claim later.

Spread has two meanings on this venue — the tight two-sided tick gap and the one-sided floor-to-bound distance — and analysis must state which.

Use the one-sided share and the two-sided tick median as the pair that describes the spread regime.

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