Beatpoly is built for professional use. That means our scores, statistics, and market notes must be explainable, testable, and limited by clear assumptions. This page documents how we evaluate prediction markets, how we classify risk, and what evidence is required before a statistic appears in Beatpoly research.
Review Scoring Framework →Every published metric should include enough information for a reader to understand where it came from, what it includes, what it excludes, and how much uncertainty remains.
For a statistic to be used in enterprise product decisions or cited as research, the following fields must be documented:
Beatpoly's reliability rating is not a prediction of whether an event will happen. It is an assessment of whether the market price is clean, usable, and defensible for professional use.
Measures: How likely the market is to settle cleanly and in line with the plain-English interpretation of the question.
A low resolution risk score means the market has clear, unambiguous rules, a reliable data source, and a well-defined settlement process. A high score means the market has risk of contested settlement, source failure, or wording ambiguity that could affect how the contract resolves.
Measures: Whether the displayed price is usable for trading, monitoring, or public citation without significant distortion.
A thin market may show a price that cannot be executed without moving it significantly. Beatpoly models what a professional-sized trade would actually cost after spread, fees, and slippage.
Measures: Whether two similar-looking markets across venues are actually equivalent — asking the same question, resolved under the same rules, using the same data source.
Most cross-venue "arbitrage" opportunities fail this test. A price gap between Kalshi and Polymarket is compensation for basis risk, not free money. Beatpoly quantifies how much the markets actually differ.
Measures: Whether market movement looks unusual, concentrated, or potentially disconnected from public information flow.
Abnormal flow detection does not prove wrongdoing. It identifies patterns that warrant additional scrutiny — particularly useful for compliance teams and media researchers who need to understand what is driving a price before relying on it.
The following statistics appear in Beatpoly's research library. Enterprise product decisions rely only on statistics whose methodology has been fully documented. Research library content may cite statistics before full audit publication, but those statistics require the documentation fields described in the Research Standard above.
| Statistic | Use | Required documentation | Status |
|---|---|---|---|
| 72% of contracts resolve NO | Market-structure research / educational baseline | Venue list, date range, market count, binary vs. multi-outcome treatment, cancelled market treatment, category breakdown | Provisional — methodology card in progress |
| 87.3% of traders lose money | Behavioral research / educational baseline | Wallet sample, PnL definition, realized vs. unrealized treatment, fees, wash-trading filters, minimum activity threshold | Provisional — not used in enterprise scoring |
| 22-point maker/taker execution gap | Execution-quality research | Source paper, exact sample, venue applicability, fee period, definition of maker and taker in context | External claim — not used in scoring |
| 161,000 contracts analyzed | Dataset credibility | Market IDs, source, date range, venue split, category split, exclusion criteria | Dataset card required before homepage use |
| $21.5B Polymarket 2025 volume | Market context | Source, platform-reported vs third-party measured, time period, methodology | Source required — remove until cited |
Market data, liquidity metrics, and abnormal-flow alerts are designed for frequent updates as market conditions change. Methodology pages and research statistics are updated when datasets, scoring rules, or venue mechanics change materially. Last methodology review: April 2026.
Prediction markets can be powerful information tools, but they can also be distorted by poor liquidity, ambiguous rules, insider information, and market manipulation. Beatpoly is designed to help professional users identify these risks before relying on market prices.
Using Beatpoly data responsibly means understanding what each score does and does not measure, acknowledging the limitations documented on this page, and applying appropriate judgment before making decisions based on market prices.