Price-relative scoring
Every fill is scored against the price actually paid. Calling a near-certainty right earns almost nothing, and being right against a contested price earns the full distance, so easy calls cannot pad a record.

OVERROUND ranks public forecasters by skill-adjusted accuracy, so a hot streak can’t pass as a track record. You get an alert when a consistently accurate forecaster changes position. Use it as one research input on the venues you independently choose. OVERROUND does not recommend trades, execute them, or decide what is suitable or lawful for you.
| # | Forecaster | Skill | Resolved |
|---|---|---|---|
| 1 | 54 0x54a2…1d43 | +21.14 | 1813 |
| 2 | ma | 1895 | |
| 3 | ma | 230 | |
| 4 | ma | 3517 | |
| 5 | ma | 690 |
Live figures from the public leaderboard API.
Anyone can look sharp for a week. Consistent, well-calibrated accuracy across many independent, resolved events is the signal, and it’s buried in millions of public on-chain records. We do the separating.
Every fill is scored against the price actually paid. Calling a near-certainty right earns almost nothing, and being right against a contested price earns the full distance, so easy calls cannot pad a record.
We track predicted vs. observed frequency across an entire history, so a single lucky window can't inflate a ranking.
Built entirely on public on-chain data. Every number traces back to a resolved event you can independently check.
From the wallets with the strongest verified historical records, straight to your inbox or Telegram.
We continuously watch the forecasters that rank highest on skill-adjusted accuracy in each category, the ones with a long, verified record of being right.
When a tracked forecaster opens or adds to a position, our on-chain reader picks it up within minutes of the move landing on-chain.
You get a clean message: the market, the side, the entry price, and the size, with context on the forecaster's track record in that category.
Rank every tracked forecaster by skill-adjusted accuracy, filter by category, and open any wallet’s per-category record and current positions. Free to read; paid subscribers get every score, full calibration detail, and graded event history for the top of every category’s board.
| # | Forecaster | Skill | Resolved | Cal. |
|---|---|---|---|---|
| 1 | 9c 0x9c4f..2a1b Politics | 0.94 | 412 | |
| 2 | 7a 0x7a3e..e920 Crypto | 0.91 | 287 | |
| 3 | 1d 0x1d88..04cf Sports | 0.90 | 356 | |
| 4 | b2 0xb2f0..7733 Iran | 0.89 | 198 | |
| 5 | 6e 0x6e51..aa10 Finance | 0.88 | 301 | |
| 6 | 3f 0x3fa9..51d2 Politics | 0.86 | 244 | |
| 7 | c0 0xc08d..9e4b Esports | 0.85 | 177 |
For any forecaster, see how their stated confidence lines up against what actually happened. Points on the diagonal mean a forecaster who says “70%” is right about 70% of the time, the gold standard of a reliable predictor.
It’s the most honest one-glance answer to “does this wallet’s record show consistent calibration?”
When a tracked forecaster makes a move that clears the conviction gate, you hear about it on email and Telegram, typically within minutes of the move landing on-chain, so you can review it alongside your own research. No setup: alerts are on from the moment you subscribe.
Representative examples · informational only · not a recommendation
Every number on OVERROUND is derived from public on-chain data with a transparent, versioned method. There is no editorial judgement: the method is documented and versioned, with a small set of proprietary anti-gaming screens. Calibration is tracked separately from rank, so you can always see the working.
Read the full methodology →Out-of-time validation results are published on the research page, committed before the results were known.
Positions and resolutions are read directly from on-chain Polymarket activity, with nothing self-reported.
Each resolved call is weighted by how contested the market was, so easy calls count for less than hard ones.
Calibration uses a Brier-style proper scoring rule that rewards confidence which proves correct and penalizes confident misses, and it is tracked separately from the rank.
Rankings discount thin track records, so a short, lucky run can't reach the top of the board.
A high-scoring forecaster moves. Subscribers know within minutes. One subscription covers all of it, and the public board stays free, so you can check the record before you pay a dollar.
Billed monthly at the founding rate. The price stays $19 for as long as your subscription stays active.
By subscribing you agree to the Terms and authorize recurring billing until you cancel.
Not ready to pay? Browse the live board free. You see each category’s #1 forecaster in full, with the other identities, skill scores, full ranking, and alerts saved for subscribers.
The Weekly Skill Report. Leaderboard changes, notable wallet moves, and every newly graded alert from the featured feed, the alerts subscribers receive. Wins and losses both, one email each Monday.
Free. Unsubscribe anytime. Privacy.
No. OVERROUND is an information and analytics subscription. We surface historical accuracy and notify you when tracked wallets change position. We never tell you what to do, accept stakes, or hold funds. Past accuracy does not predict future outcomes.
OVERROUND is an information and analytics subscription built on publicly available prediction-market data. We track and rank the historical accuracy of forecasters, publish a verifiable calibration ledger, and send alerts when tracked forecasters change position. We do not accept stakes, hold customer funds, custody cryptocurrency, or pay money to users. Nothing here is financial or investment advice, and past accuracy does not predict future outcomes. OVERROUND is independent and not affiliated with Polymarket, Kalshi, or any trading venue.
Join at the founding rate and lock in $19/month for as long as you stay subscribed.