Press note
We Tested Polymarket Prices Against 6,776 Resolved Markets. Here's Where They Bend.
6,776 marketsValidation sample
Resolved Yes/No markets with known outcomes.
Original Vultax research: in an experimental test, a calibrated fair-value model beat raw Polymarket prices by ~9.6% on held-out Brier score. Where the favorite-longshot effect lives, what the model can honestly flag, and why it still isn't a trading edge. Prediction-market prices are good probability estimates on average — that is the standard result, and it held in our data. But 'good on average' leaves room for systematic, repeatable distortion, and the best-documented one is the favorite-longshot effect: low-probability outcomes trade above their true frequency and near-certain outcomes trade slightly below it. A fair-value model, honestly defined, answers one question: given where this market is priced, where did markets priced like it historically resolve? That is a calibration question, and calibration questions have a standard scoring rule.
The observation window and population are stated in the study's method section.
Supporting figures
- ≈ +9.6% — Brier improvementExperimental, held-out markets: 0.0352 → 0.0319 versus raw market prices.
- Near 0 and 1 — Where prices bendThe calibrated mapping's improvement concentrated at the longshot and near-certain ends of the price range.
Images

Vultax Research
The study’s illustration, free to run with that credit and a link. It is artwork, not a figure.
Share card, 1200×630 — the headline figure and the title, rendered. Credit Vultax Research.
Reproducing one of the study’s figures instead? Caption it “Chart: Vultax Research”. Higher resolutions and other crops on request.
Links
- The study
- Full text, sources and method
- Markdown edition
- Plain text for quoting and for assistants
- Browse markets in Vultax
- Every open Polymarket market with its price history, liquidity and Vultax analytics.
Press contact
Vultax Research — [email protected]
Questions on method, a figure checked before you print it, a chart at another size, or a cut of the data this study does not publish. Say what you are writing and when it runs.
How to cite
Vultax Research, We Tested Polymarket Prices Against 6,776 Resolved Markets. Here's Where They Bend., 2026-09-08, https://vultax.com/research/prediction-market-fair-value-calibration
Published 2026-09-08. Figures are the published study; live screens move. Plain-text version · Use our data