Where unusual Polymarket wins concentrate

Thirteen published Polymarket wallet addresses were matched to a 714-million-fill archive and scored against the prices they paid. Pooled once per company reporting cycle, their initial positions won 32 of 37 cycles on companies audited by KPMG, against 23.3 wins implied by entry prices, and 28 of 55 cycles on other known auditors' clients, against 30.7 expected. A separate subject-concentration screen over 22,322 eligible wallets selected 321, against a simulated mean of 51.3. The paper keeps the three results apart, states the units, reports the screens that failed and the corrected strict-screen figures (1,316 bets, a matched gap of 5.40 cents a share against established wallets), and lists the checks still needed before any stronger reading. It does not establish insider trading, common ownership or a way to identify informed traders in advance. Version 1.1 names the author and adds a sector-and-size check: auditor labels are shuffled only among companies of the same SIC sector group and public-float tercile, and none of 20,000 shuffles reaches the observed excess. It also adds acknowledgements, declarations, a citation block and four references. No figure or table of version 1.0 changed. Vultax Research published the study at https://vultax.com/research/polymarket-excess-wins-domain-concentration on 8 October 2026. This record holds the paper (version 1.1, Vultax Research Note 2026-025) and its BibTeX entry, the version 1.0 paper and companion data package, the two aggregate tables and the script of the new check, a static HTML snapshot and a PDF print of the page, its proof scope record and a CITATION.cff.