Wallet Audit · free sample of preflight_copy
Pick any wallet on Polymarket. We simulate copying its last two months from our tape — at the trader's entry prices, with 2026 taker fees and a conservative slippage floor on top — validated on a second month the wallet never saw coming. Below is a complete real audit, free to read. The live product is the same verdict as a tool call: preflight_copy at $0.10, paid by your agent in USDC via x402. The API returns structured JSON — aggregate fields, not trade rows; this page is a frozen human-readable sample of the same audit, plus an excerpt of the evidence run behind it. Shareable per-call receipt pages are on the roadmap.
COPY. Positive net copy-return in two independent months, adequate sample, moderate concentration risk. Not investment advice — see caveats below.
This is the human-readable form of preflight_copy — the same verdict your agent gets today as a live tool call: $0.10 via x402, sub-second for covered wallets, computed from our own identity-tagged tape running since December 2025. Profiles are recomputed nightly with an hourly delta; every response carries next_refresh_at — re-buying before that timestamp returns the same data. Wallets below the coverage floor return QUEUED free of charge — typically ready in minutes, worst case about an hour. A COPY verdict is issued only with out-of-sample validation; in-sample evidence caps it at WATCH.
The grade weighs all five — and weak sub-scores drag it down harder than strong ones lift it. A wallet cannot buy its way past a C in concentration.
| Window | Copied | W / L | Win rate | Net ROI |
|---|---|---|---|---|
| Jun 12 – Jul 11 | 132 | 83 / 49 | 63% | +11.3% † |
| May 12 – Jun 11 (out-of-sample) | 107 | 69 / 38 | 64% | +2.5% |
† Two bases, both shown: +11.3% after the 2c slippage floor · +9.2% after explicit taker fees ($14.02 on 132 trades). Across all 24 screened wallets, fee-inclusive: 6 clear positive, not 7.
Category mix of the 132 copied trades: esports 42% · sports 35% · other 22% · crypto 1%. The P&L concentrates in esports — +$96 of the +$75 total (sports +$6, other −$23). Top-5 markets are only 14% of total absolute P&L: healthy diversification across 130 markets, inside a category-concentrated book.
In 57% of his entries, a 30–60 s copy delay moves the entry price by at most one cent (median drift: 1c). At a 5-minute delay, 44% stay within 1c and the median drift is 2c. He does not enter early: his median first entry lands 88% of the way from market creation to resolution, and the account is clearly a bot — active 24/7, bursting up to 17 prints in a single second. This is drift evidence — how far the market moved after his entry — not a fill simulation at your size. Wallets that make their money in fast markets fail this test — and the audit says so.
Correction, on the record: an earlier draft of this page said 84% — our own recompute corrected it to 57%. That correction is the product working.
Six real rows from the 132-trade evidence run behind this sample:
| Timestamp (UTC) | Market | Outcome | His entry | Sim entry | Fee | Result | P&L (net) |
|---|---|---|---|---|---|---|---|
| 2026-06-12 13:04 | cs2-fut-g2-2026-06-12 | FUT Esports | 0.31 | 0.33 | $0.17 | loss | −$5.17 |
| 2026-06-12 20:14 | fifwc-can-bih-2026-06-12-can | No | 0.77 | 0.79 | $0.05 | win | +$1.28 |
| 2026-06-13 20:09 | mlb-ari-cin-2026-06-13-total-9pt5 | Over | 0.49 | 0.51 | $0.12 | loss | −$5.12 |
| 2026-06-13 22:19 | wnba-ind-conn-2026-06-13 | Indiana Fever | 0.76 | 0.78 | $0.07 | win | +$1.34 |
| 2026-06-20 18:25 | cs2-ts7-fal2-2026-06-20 | Team Falcons | 0.60 | 0.62 | $0.10 | win | +$2.96 |
| 2026-07-08 05:53 | lol-g2-t1-2026-07-08 | G2 Esports | 0.13 | 0.15 | $0.21 | win | +$28.12 |
These six rows are an excerpt of the 132-trade evidence run this sample was computed from — each row checkable against the public record. The API returns aggregate fields (verdict, fee-inclusive ROI, win rate, entry-price buckets, category mix, machine-readable reasons); a trade-level appendix is not part of the API response today.
Past copyability does not guarantee future results; wallets change strategies, seasons end. Simulation fills at wallet-price + $0.02; live fills can be worse in thin books. One wallet is concentration risk — that is exactly why the Leaderboard tracks all survivors forward, and preflight_copy runs the same checks programmatically. Context: of 24 wallets screened identically this month, 17 were net-negative to copy after the 2c slippage floor — 18 of 24 after explicit 2026 taker fees, including the two biggest whales by volume.
No farming patterns found. Our tape-level screen ran four independent tests on this wallet; all four came back clean.
Honest caveats: he is a bot (active 24/7, up to 17 prints in one second). He builds positions in ~24 clips on average — iceberg-style execution, which is exactly why we mirror the first entry only. An existing copy bot already mirrors him in 102 of his 319 markets, so copy-crowding is a real risk. And our tape is taker-side: a wallet funding-graph check on Polygonscan is the one test this screen cannot do.
Another service publicly flags this wallet as high farming risk. We publish our tests and the evidence above so you can judge both for yourself.
Method
Big wallets split one position into dozens of clips. We copy the decision — the first entry per market — the way you actually would.
Your simulated fill = their entry + a flat $0.02 slippage floor, plus 2026 taker fees (rate × p × (1−p)). Crypto-second markets excluded as uncopyable.
Anyone looks good on the month you picked them. We re-run the prior month blind — most "winners" die here.
Every simulated trade in the evidence run carries a timestamp, market, his entry, your simulated entry, fee and result — six rows are excerpted above, each checkable against the public record. The API returns the aggregates computed from them.
What the simulation is — and is not. Simulation: your copy fills at the trader's entry price plus a conservative 2-cent slippage floor, $5 per trade, 2026 taker fees. We do not yet replay your exact size and delay against the historical order book — that replay is on the roadmap (we record full-depth books for it). Until then, treat the number as a conservative floor, and the latency panel as the drift evidence. Every verdict today uses the +2c floor; exact book replay at your size and delay is roadmap.
What every call returns
Your agent gets this as a structured tool call — preflight_copy, $0.10 via x402 — returning structured JSON. This page is a frozen human-readable sample of the same audit; shareable per-call receipt pages are on the roadmap. See For agents.
This page is the free sample. The live product is the same verdict as a tool call: sub-second for covered wallets, computed from TickTape's own identity-tagged tape running since December 2025 — not reproducible from Polymarket's live API. Paid by your agent in USDC via x402; no card, no account, no subscription. Wallets below the coverage floor return QUEUED free of charge — typically ready in minutes, worst case about an hour.
Point your agent at the APIHumans read for free — paying requires an agent (any x402 client works). Live calls are paid from the first request; test the exact response shape free with ?sandbox=1 — unlimited, never charged. Support: hello@ticktape.cc