Polymarket Price Level Data

• Tick size is 0.5 cents on the tighter grids and 1 cent deeper in the ladder — the smallest executable step on Polymarket.
• Up/Down ladders span the strike range (0.044 to 0.996 in the measured extreme), with mid-adjacent levels 0.50/0.60/0.70/0.80/0.90/0.95/0.99/1.00 as the standing shape.
• Sports ladders carry 35–36 levels per side from 0.01 to 0.99 on the connected grids.
• The price-bound structure is in the archive: high/low bounds per contract, with the ladder between them.

The strike ladder compressed from 0.044 to 0.996 in a single second of a documented reprice — a 99.6-cent width traversed in one frame.

The 2026-09-09 ETH repricing moved the executable floor to the new futures level at 14:42:50, and the mid to 0.065, within that same second.

Level-adjacent trading is what a one-sided book leaves you: with no resting ask, the buyable level is the far bound, and its price is the level data, not the mid.

• Executable-boundary math: fills reference the level list as-of the frame, not the mid — the only honest price for an impact study.
• Bound tracking: mark each frame's high/low bound against the 250ms level rows and compute where the ladder actually traded within it.
• Grid health: the 98.6%-at-0.99/0.995 metric is computed from the level data and is the cheapest book-quality measure per market.

• Pull one Up/Down contract's ladder frames and list the distinct price levels present over its life; the grid plus its empty states is the market's anatomy.
• Measure level-to-level gaps: between mid-adjacent levels the gap should be 0.5 or 1 cent on the tight grids; record the extremes.
• Reproduce a documented one-second ladder pass (0.044 to 0.996) from raw frames: find the two consecutive frames whose levels span that width.
• Compute the round-grid share (frames whose executable level is 0.99 or 0.995) — the book-quality metric that validates a market's tightness.

Price levels are what actually executes; the mid is a summary that often does not exist as an executable price, which is why level data, not mids, is the right substrate for impact work.

The grid is a design AND a risk: half-cent ticks make tight quotes cheap, and one-sided books make those same quotes takeable only by walking a ladder that may jump the width in a second.

The 98.6%-at-round metric is real and measured: a market that prices 0.99/0.995 almost always is a market whose behavior you can describe from level rows alone.

Level data feeds executable-boundary math, bound tracking, and grid-health metrics; store it as the ladder frames and derive the top-of-book view on read.

Every fill study should pin its prices to levels as-of frame, not to a dashboard mid; that single habit kills a whole class of slippage measurement error.

Price levels are the executable truth: fills hit levels, and the mid is a summary that often is not an order you can take. Level data is the substrate for slippage, impact, and grid-health work precisely because it happens to match the venue.

The grid is both the offer and the warning of the venue: half-cent ticks make tight quotes cheap, and ladders that jump the width in one second (0.044 to 0.996 recorded) are the risk inside that efficiency.

The measured round-grid share — 98.6% of frames resting at 0.99/0.995 — is the single most useful market-quality metric on the site, and it computes in one pass from the level rows.

Pin every fill to its level as-of frame, not to a dashboard mid, and the worst class of slippage measurement error disappears; the level record is the one column your models should trust most.

• price bounds explained
• tick size explained
• order book explained