CrowdIntelDefinition
What is Scope edge?
The gap between a wallet's win rate inside a narrow scope (one topic, one category, one market) and its global win rate — a signal that the wallet has information advantage in that specific scope.
In detail
A wallet with a 55% global win rate is mediocre. But if that same wallet hits 85% inside Middle East conflict markets over 40 resolved bets, the scope-specific edge is a strong signal of domain knowledge or inside information. Scope edge formalizes this: it's the difference (or ratio) between scope-narrow win rate and global win rate, weighted by sample size. Finding scope-edge wallets is the most reliable way to surface domain-specific insiders that generic leaderboards miss.
How CrowdIntel measures it
CrowdIntel implements scope-narrow investigations (type=wallet_scope, scope_narrow=true) that freeze the wallet's stats inside a specific topic. The investigation page shows scope-specific win rate, PnL and top markets, with the global overlay disabled — so a wallet's iran-war edge is not diluted by their losing bitcoin bets.
Frequently asked
Why does scope matter for insider detection?
Insiders usually have information in one domain (their industry, their country, their network). A person with edge on pharma FDA decisions shouldn't have edge on World Cup matches. Scope-narrowing separates real insiders from lucky generalists.
What's the minimum sample for scope edge to be credible?
CrowdIntel requires ≥10 resolved bets in the scope AND p-value ≤ 0.1 before opening a scope-narrow investigation. For thick clusters (3+ wallets), the threshold rises to ≥20 combined bets.
Can a wallet have high scope edge but negative PnL?
Yes — a wallet betting heavily on heavy favorites can hit 85% win rate but lose money (winning cheap bets, losing expensive ones). CrowdIntel's orchestrator rejects scope-edge subjects with negative PnL as 'heavy-favorite bettors, not insiders.'
Related terms
- Polymarket insider
A Polymarket trader who bets with non-public information or against the consensus and wins at rates statistical models cannot explain by skill alone.
- Bayesian win rate
A shrinkage estimate of a wallet's true win rate that combines raw wins-over-bets with a prior — preventing a wallet with 9 wins out of 10 bets from being treated as a 90% sharpshooter.
- P-value (trading)
The probability that an observed win rate could occur by chance alone if the trader had no edge — used to separate genuine skill from lucky streaks.
- Polymarket whale
A Polymarket wallet that trades significantly larger positions than the median user — typically $5,000+ per bet or $100,000+ lifetime volume.