Polymarket Slippage Data

The archive computes fill-cost inputs per row — quoted spread, depth at the executable price, cumulative size before the bound — so slippage for any order size is reconstructable second by second, not estimated from a market average.

On a venue where executable prices sit at the 0.99 bound 98.6% of the time, slippage concentrates in one place: consuming the resting supply at the bound or crossing to the other side.

The measured bound structure tells the slippage story: a median 42,254 shares rest at the winning bound while 90% of final-second books have one side only — so a size above the rest crosses the venue at the loser side's price, and the series captures exactly where that threshold sat each second.

One-sided books (16.9%) bound slippage at a single price, and the honest performance number for a strategy is its slippage at the sizes it actually trades.

A slippage study runs the order's size against each retained 250ms frame: consume the executable side, price the remainder against the next step or the other side, and record the total cost for that instant.

The archive makes each frame's full ladder available (depth history), so slippage is a computation, not a convention.

Slippage is size-dependent and the honest number belongs to the order: the data products and the slippage explainer agree that one-size estimates mislead, and the archive exists so the estimate becomes a per-frame measurement.

• Slippage explained
• Fill simulation
• Market making