Polymarket wallets ranked by how far their settled baseball buys beat the price they paid, on a conservative lower bound, beside the recent large trades on MLB games, World Series futures, and player props. Sabermetrics, pitch tracking, and advanced batting metrics give a quantitative trader plenty to model.
Ranked by the 95% lower bound of each wallet's edge on its first large buys, observed as they filled and equal-weighted by independent event. Coverage is Polymarket only and partial: large taker buys since model launch, not reconstructed history, and not raw win rate.
A live rank needs 30 resolved independent events. Each row uses model category-skill-v2.1.0, and observations begin , from partial Goldsky primary taker fills. Edge metrics refreshed 57m ago.
Large trades in baseball, ranked by how unusual they are rather than by time, each with the wallet behind it. Without Pro a trade reaches this flow once it is 24 hours old, and the suspicion score behind the ranking stays with Pro.
Wallets first seen in the last 30 days, placing large directional trades. They have little settled history yet, so this is flow, not edge.
The wallets ranked at the top of this page. The order is the 95% lower bound of each wallet's edge against the price paid, measured on its first large buy in every settled baseball market as it filled. A short lucky run has a wide interval and a low bound, so it rarely reaches the top.
Sample size. MLB's 162-game regular season is the longest of the major US leagues, so a statistical model converges on accurate probabilities sooner, and the starting pitcher decides enough of a game that a good pitcher-matchup read makes single games more predictable.
Starting pitcher quality (FIP, xFIP, SIERA), bullpen availability, park factors, weather, and lineup construction. A late lineup change or a bullpen worn down by the previous series is where the edge shows up.
Every other category we track.
Moderate in the regular season, much deeper in the playoffs and the World Series. Marquee matchups and rivalry games draw more trading than a mid-week series between small-market teams.
They can. Models built on Stuff+ for pitchers, Barrel% for hitters, and sprint speed look for prices that lag the numbers, most often early in the season, when samples are small.
