The price on the screen is not the price you pay

The price a prediction market shows you is the midpoint of its order book — not a number anyone can actually trade at. What you pay is whatever price the book fills your order at, plus whatever the venue's own fee schedule charges on that fill, and the difference between those two numbers is a cost that is entirely computable and that almost nobody computes before they click buy.

This isn't a rounding error. On a market with real depth, the gap is a percentage point or two. On a thin market — a long-shot outcome, an obscure sports line, a geopolitical question few people are trading — the gap can be the difference between a position with real edge and one that was underwater the moment it filled.

What “the price” actually means

A Polymarket contract has four separate numbers, not one: the best bid, the best ask, the midpoint, and the last traded price. The screen shows you one of these — and which one depends on how wide the market is.

Executable price is the volume-weighted average price your actual order fills at, walking through the order book one level at a time until your requested size is filled. It is never the ask and never the midpoint; on anything but a trivially small order in a deep market, it is worse than both. A 500-share order doesn't fill at one price — it eats through the cheapest shares first, then the next-cheapest, and the average of all of that is what you actually pay.

Break-even probability is the executable price plus the fee, expressed as the probability your position needs to be right just to come out flat. It is not the midpoint's implied probability — it is a higher bar, and the gap between the two is exactly the cost this article is about.

Polymarket's own published rule for which price it displays makes the problem sharper: the displayed number is the bid-ask midpoint, unless the spread between best bid and best ask exceeds ten cents — in which case the site shows the last traded price instead. On a wide market, the number you're looking at may not be a live quote at all. It can be a trade from an hour ago, on a book that has since moved.

The fee, made concrete

Polymarket publishes a taker-fee schedule: a rate, multiplied by the size of your trade, multiplied by how close the price is to 50 cents — the fee is highest at a coin-flip price and shrinks toward the extremes. The rate itself varies by category, and it is not a secret:

That last line matters beyond the arithmetic: some markets — a meaningful share of the geopolitical and conflict questions among them — carry no fee at all. Treating every market as if it charges the same rate, or assuming a fee where none exists, produces a wrong answer in either direction. The honest position, when a market's own fee fields don't resolve cleanly, is to say the fee is unknown — not to guess.

A worked example

The screen said 62 cents. The position actually cost 65.6 cents to break even on — a gap of more than three and a half points that existed before any judgment about whether the trade itself was any good. Every article explaining spread and slippage in the abstract stops here. None of them run this arithmetic on your actual position, for your actual size, on the actual market you're looking at.

Check it on a real position

What you actually pay on Polymarket runs this exact arithmetic — displayed price and whether it's a midpoint or a stale trade, your volume-weighted fill, the category fee, and the break-even probability — on any live market, for any size, free and without an account.

The calculator answers what a position costs. It does not tell you whether the position is worth taking — that judgment is yours, and it depends on far more than arithmetic: your own thesis, what the resolution contract actually says, and what would have to be true for you to be wrong. That is what the full position audit is built to examine.