maz26

$3.2M · 30-day displayed profit

Reconstruction error for all 25 wallets this cycle is published in full in the reconstruction workpaper.

Our original fetch was incomplete and our own marketCapped flag had already caught it; the reconstruction is unchanged and the gap stays unresolved.

We are not refusing this wallet because its book is unreadable. We are refusing it because our original fetch of it was incomplete (68.8% of the fills we have since measured), and we published figures computed from it before noticing. The fill and market counts on this page were rebuilt 2026-07-18 from the most complete re-fetch obtained (provenance: u2-wallet-completeness-provenance.json); the reconciliation figures were not rebuilt and are not evidence about the trader. This is a data-completeness gap in our own fetch, not a finding about the trader — logged as U2 in the standing unreconciled-items table, alongside the full 24-wallet completeness ledger (U2).

Version history →

Corrected 2026-07-25. Re-fetching this wallet with the corrected pipeline measured 128.68% of displayed — an over-reconstruction — and a blind second implementation, run without access to ours, measured 128.68%. The -220.4% above was our data gap, not the wallet’s: it came from the fetch that silently omitted maker-side fills, the same failure we retracted in the endpoint case. The corrected figure still lands outside the ±15% band, so the verdict does not move, and the July 2026 cycle stays frozen on the fetch that produced it — the correction is published beside the original, not adopted. Source: scorecard-refetch-2026-07-25.json; the dated correction is on the Scorecard.

55756 fills · 185 markets · avg buy $0.656 · top market: Will Canada win on 2026-07-04?