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 announcements, as verified 2026-09-22. The live leaderboard stopped at a Cloudflare verification page that day, so nothing here describes its current standings.
At a glance
| Arkham Elo, per its June 29 announcement | OVERROUND | |
|---|---|---|
| Ranks by | An Elo rating, with more credit for correct calls at lower odds | Accuracy scored against the price actually paid |
| Bet size | “Ratings do not take bet sizing into account” | Fills weighted by size within each event |
| Crypto up/down 5-minute markets | Its launch #1’s strategy centered on them, at a 64.3% win rate over 2,644 bets | Excluded by policy |
| Per-category ratings | None mentioned in the announcement | A board for each Polymarket category |
| Published graded record or backtest | None in the announcement, and no later post in its announcements index as of 2026-09-22 | Every alerted position graded in public, and a pre-registered out-of-time validation |
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 once it has five or more recent positions, and each set of mutually exclusive outcomes counts once, so a hundred candidate legs of a single race 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
OVERROUND publishes the output. Every alerted position 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. Arkham’s June announcement publishes no graded record or backtest for its Elo ratings, and its announcements page lists no later post on them as of 2026-09-22.
Difference four: depth per category
OVERROUND ranks within categories, because a politics specialist and a tennis specialist are different claims. It runs a board for each Polymarket category, each with calibration records on top-ranked wallets. Arkham’s announcement mentions no per-category ratings.
Who should use which
Arkham fits a reader who wants an accuracy-ordered list from a large analytics brand and already uses its platform. 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.
Get the next conviction alert, typically within minutes
Pro sends conviction alerts on Telegram and email when a tracked forecaster makes a move that clears the conviction gate, typically within minutes. Each wallet and each category has a daily limit, and alerts held back by these limits go into the daily digest. Pro also opens every skill score on the full leaderboard. The receipts ledger stays free.
Sources
Competitor statements verified 2026-09-22 against Arkham’s published announcements. Corrections welcome: hello@overround.pro. Nothing here is financial advice. Past accuracy does not predict future outcomes.