Historical metrics
Compute descriptive statistics from public on-chain activity emitted by prediction-market smart contracts.
Read-only research dashboard
Historical on-chain analytics for prediction markets.
Polymarket Backtest computes reproducible statistics from public smart-contract activity, then compares the same historical window across transparent baseline methods.
Educational and research use only. Not financial advice. Not a prediction service. Not affiliated with Polymarket.
Methodology preview
90 day window
Historical metrics
Matched trades18.4k
Markets observed712
Benchmark paths5
Benchmark distribution
Disclosure checks
Lookback90d
Universefixed
Bait filterflag
Future claimnone
Reproducible pipeline
1. IngestPublic on-chain events
2. NormalizeShared schema and prices
3. CompareSame window, multiple baselines
4. DiscloseLimits, assumptions, caveats
What it is
A research layer for public prediction-market activity.
The product is built for people who want to inspect historical behavior, compare baseline assumptions, and understand methodology before making claims from the data.
Comparable baselines
Run the same historical window through multiple benchmark methods, so comparisons are explicit instead of implied.
Disclosure-first analysis
Surface future-performance caveats, limited wallet universes, bait-wallet risk, and lookback constraints as part of the product.
Read-only workflow
No execution, custody, brokerage, market making, trading recommendations, wagering, or prediction issuance.
Methodology
Make the assumptions inspectable.
The beta is designed around a simple idea: historical analytics should show the data path, the comparison window, and the caveats before it shows any headline number.
analysis.yml
read-only
window: 90_days
source: public_onchain_events
mode: descriptive_statistics
benchmarks:
- baseline_hold
- random_reference
- filtered_activity
disclosures:
future_performance: required
wallet_universe: limited
bait_wallet_risk: flagged
execution: none
Same window, same schema
Benchmark comparisons are only meaningful when the input window and normalization rules are explicit.
Caveats shown with the result
Historical performance is not presented as a signal, recommendation, or forecast.
Independent third-party analytics
The service is not affiliated with Polymarket or any prediction-market venue.
Who it is for
Analysts who need evidence, not a pitch.
The dashboard is intentionally narrow: historical, methodological, and inspectable.
Researchers
Explore public prediction-market behavior with reproducible windows and transparent assumptions.
Journalists
Use descriptive evidence and caveats when covering wallet behavior and market activity.
Market analysts
Compare historical baseline methods without receiving trading advice, predictions, or execution tools.
What this is not
Clear boundaries are part of the product.
Polymarket Backtest is a software analytics service for educational and research use.
Not a gambling, gaming, or wagering operator.
We do not operate markets, accept wagers, or settle prediction-market outcomes.
Not a prediction service.
We do not forecast outcomes or tell users what will happen in any market.
Not investment advice.
Historical analytics are not recommendations to buy, sell, copy, trade, or participate.
Not a how-to-profit advisory.
The product is not designed to teach users how to profit from prediction markets.
Not a trading or wagering venue.
We do not route orders, provide execution, hold funds, or custody assets.
Not affiliated with Polymarket.
We are an independent third-party analytics provider, not endorsed by or partnered with Polymarket.
Beta access
Join the research beta before the public launch.
We are inviting a small group of researchers, analysts, and prediction-market observers to review the methodology, stress-test the disclosures, and shape the dashboard.
Expected beta: August 2026
Share your use case and we will follow up when the research beta opens.
Request beta access