OVERROUND vs Arkham Prediction Market Elo
Arkham launched Elo-style accuracy ratings for Polymarket traders in June 2026, and the framing is one this site shares: profit leaderboards mostly reward bankroll and luck, so rank accuracy instead. The two products then diverge on four checkable points. Claims about Arkham below come from their own announcement, as verified 2026-07-22.
Where the two agree
Arkham’s announcement says correct predictions at lower odds raise a trader’s rating more than correct predictions at higher odds. That difficulty weighting is a sensible idea, and it is a cousin of what OVERROUND does: every graded position here is scored against the price actually paid, so calling a near-certainty right earns almost nothing. Both products reject raw P&L as the ranking metric. Arkham also states its own limitations plainly in print, which deserves credit.
Difference one: bet size
Arkham’s announcement, verbatim: “Ratings do not take bet sizing into account, so smaller but more accurate traders will rate higher than larger bettors with a lower win-rate.” OVERROUND’s scoring works on the actual fills: position sizes flow through the per-event grading, the conviction gate that triggers alerts is sized against each wallet’s own history, and correlated markets in one event family count once, so a hundred legs of a single question read as one piece of evidence.
Difference two: what the top of the board rewards
Arkham’s launch announcement named as its #1 a trader of crypto up/down 5-minute markets, at 64.3% across 2,644 bets. OVERROUND excludes that market class by policy: an entire up/down series collapses to a single event under the independence rules, and its outcomes track short-horizon price noise rather than judgment. The exclusion is documented on the methodology page, coverage boundaries included.
Difference three: the graded record
A rating system asks you to trust its math. OVERROUND publishes the output: every alert ever sent is recorded at detection and graded on resolution, losses included, on the receipts ledger, and the scoring method carries a pre-registered out-of-time validation whose results were committed to publication before they existed. As of 2026-07-22, Arkham has published no graded record or backtest for its Elo ratings.
Difference four: depth per category
OVERROUND ranks within categories, because a politics specialist and a tennis specialist are different claims: 18 category boards, each with calibration records on top-ranked wallets. Arkham’s Elo is one global rating.
Who should use which
Arkham fits a reader who wants a free accuracy-ordered list from a large analytics brand and already lives in its ecosystem. OVERROUND fits a reader who wants the ranking claims auditable: size-aware grading, category depth, and a public record of what the alerts actually returned. Run any wallet through the free checker to see the grading logic firsthand, or read the full review and alternatives.
Competitor statements verified 2026-07-22against Arkham’s published announcement. Corrections welcome: hello@overround.pro. Nothing here is financial advice. Past accuracy does not predict future outcomes.
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.