Research › Traders
What if I had copied this trader?
Two questions a copy trader asks, answered from the fills the site has: what $1 on every entry of a given trader would have returned, held to resolution, after a 1¢ half-spread and Polymarket's taker fee — copied at their price and, more honestly, an hour later — and whether a trader betting far above their own normal size tells you anything. Every figure comes with its n and a 95 % interval. Past results, not advice.
Copy a trader
Pick a dataset and a trader. The site export holds each tracked wallet's fills from the day it was tracked; the one-year set, when present, is today's leaderboard replayed over twelve months — those wallets were picked with hindsight, so their figures are the figures of known winners.
Traders with the most resolved entries in this dataset:
Dataset-wide figures
Before any one trader: what copying every entry of every wallet returned, how the per-wallet means spread out, and how much the hour between a trader's fill and yours costs. A wallet whose interval lies above zero is not proof: with hundreds of wallets tested, a few pass by chance.
Site export (tracked wallets, from the day each was tracked)
The wallets are the traders the site tracked, each from the day it was added (curated wallets since March 2026, the leaderboard set since October 2 2026). Short histories: most figures have few resolved entries.
Fills 2022-12-12 to 2026-10-22 · 2,427 wallets (2,293 with entries) · 2,132,120 entries: 1,986,292 resolved, 145,828 open · 444,606 hedged markets and 160,639 exited markets left out · computed 2026-10-25 04:13 UTC
Copied at entry, after costs
−6.6 %
95 % CI −6.7 % to −6.4 % · $1 on each of 1,986,292 resolved entries
Copied 1 h later
−6.0 %
95 % CI −6.3 % to −5.7 % · 938,987 entries had a price 1–24 h later; 1,047,305 did not (most short markets had ended)
What waiting an hour costs
+0.2 %
95 % CI +0.0 % to +0.4 % · later minus entry on the same 938,987 entries — the honest copy-trading number
Per-wallet means (1,580 wallets with 30+ resolved entries)
| p10 | p25 | median | p75 | p90 | 95 % CI above zero | below zero | |
|---|---|---|---|---|---|---|---|
| Copied at entry | −28.3 % | −13.8 % | −4.9 % | +1.0 % | +9.4 % | 95 of 1580 | 480 of 1580 |
| Copied 1 h later | −36.4 % | −19.9 % | −7.0 % | +1.9 % | +13.0 % | 75 of 1355 | 461 of 1355 |
"Above zero" counts the wallets whose whole 95 % interval of the mean return lies above zero: with 1,580 wallets tested, about 40 would pass by chance alone.
By entry price
| Resolved | Hit rate | Copied at entry | Copied 1 h later | |
|---|---|---|---|---|
| under 30¢ | 383,183+32,061 open | 10.0 %9.9 % to 10.1 % | −40.7 %−41.5 % to −39.9 % | −47.3 %−48.3 % to −46.2 %176,571 priced · −13.3 % vs entry on these |
| 30–50¢ | 362,464+13,867 open | 40.9 %40.8 % to 41.1 % | −7.5 %−7.9 % to −7.2 % | −9.3 %−10.0 % to −8.7 %190,725 priced · −3.5 % vs entry on these |
| 50–70¢ | 392,135+17,141 open | 64.5 %64.3 % to 64.7 % | +5.0 %+4.8 % to +5.3 % | +2.3 %+1.8 % to +2.8 %201,830 priced · +1.1 % vs entry on these |
| 70¢ and up | 848,510+82,759 open | 94.2 %94.1 % to 94.2 % | +3.9 %+3.8 % to +4.0 % | +11.0 %+10.7 % to +11.2 %369,861 priced · +8.2 % vs entry on these |
By topic (the 12 with most resolved entries)
| Resolved | Hit rate | Copied at entry | Copied 1 h later | |
|---|---|---|---|---|
| Crypto | 579,141+8,026 open | 67.1 %66.9 % to 67.2 % | −3.2 %−3.5 % to −2.9 % | +11.2 %+10.1 % to +12.3 %56,126 priced · +16.1 % vs entry on these |
| Soccer | 462,703+21,940 open | 59.8 %59.7 % to 60.0 % | −6.6 %−6.9 % to −6.2 % | −4.7 %−5.3 % to −4.2 %263,834 priced · −0.5 % vs entry on these |
| Tennis | 123,206+718 open | 58.1 %57.8 % to 58.4 % | −5.3 %−6.0 % to −4.7 % | −4.5 %−5.4 % to −3.6 %66,924 priced · −4.3 % vs entry on these |
| Esports | 120,691+1,770 open | 55.7 %55.4 % to 56.0 % | −6.2 %−6.8 % to −5.5 % | −16.0 %−16.9 % to −15.2 %61,287 priced · −10.5 % vs entry on these |
| Weather | 82,828+10,894 open | 72.5 %72.2 % to 72.8 % | −8.7 %−9.5 % to −7.8 % | −8.0 %−9.0 % to −7.0 %62,587 priced · −0.2 % vs entry on these |
| NBA | 70,693+2,517 open | 55.0 %54.7 % to 55.4 % | −3.7 %−4.5 % to −2.8 % | −7.5 %−8.6 % to −6.4 %50,848 priced · −3.2 % vs entry on these |
| Geopolitics | 61,042+8,824 open | 65.4 %65.1 % to 65.8 % | −9.6 %−10.7 % to −8.5 % | −4.4 %−5.7 % to −3.1 %55,560 priced · +4.3 % vs entry on these |
| Mentions | 55,420+1,707 open | 63.1 %62.7 % to 63.5 % | −13.8 %−15.0 % to −12.6 % | −6.9 %−9.0 % to −4.9 %41,427 priced · +4.4 % vs entry on these |
| Elections | 43,925+40,292 open | 64.6 %64.1 % to 65.0 % | −15.2 %−16.4 % to −14.0 % | −8.2 %−9.9 % to −6.5 %30,537 priced · +3.4 % vs entry on these |
| Finance | 41,831+7,013 open | 63.7 %63.2 % to 64.2 % | −14.9 %−16.2 % to −13.5 % | −14.1 %−15.9 % to −12.2 %22,806 priced · −0.6 % vs entry on these |
| MLB | 38,051+822 open | 53.3 %52.8 % to 53.8 % | −5.9 %−7.0 % to −4.7 % | −3.2 %−4.9 % to −1.5 %23,106 priced · −1.1 % vs entry on these |
| College Basketball | 37,899+64 open | 59.6 %59.1 % to 60.1 % | −2.4 %−3.7 % to −1.2 % | −3.6 %−5.4 % to −1.8 %22,042 priced · −0.9 % vs entry on these |
Markets the traders left before resolution (160,639): holding them would have returned +10.8 % a trade (+9.6 % to +12.0 %) on the 143,904 that resolved; held side won 61.9 %.
Does conviction predict outcomes?
A conviction bet is an entry of at least 3× the trader's own 75th-percentile position cost over the previous 90 days (at least 10 earlier positions; at least $1,000). The baseline is point in time: only earlier entries feed it. The question is whether those bets won more often, and returned more after costs, than the same traders' normal-size entries at the same prices.
Site export (tracked wallets, from the day each was tracked)
2,108,734 resolved entries had a baseline (10+ earlier positions in the previous 90 days); 23,386 did not.
Conviction bets: held side won 79.0 % of 119,805 and returned +2.0 % a trade after costs (+1.7 % to +2.4 %); normal-size entries won 61.4 % of 1,844,689 and returned −7.2 % (−7.3 % to −7.0 %). Conviction bets sit at higher prices on average (75¢ vs 59¢), so compare within a band before reading anything into the hit rates.
| Resolved | Hit rate | Copied at entry | Copied 1 h later | |
|---|---|---|---|---|
| Conviction (≥ 3× own p75, ≥ $1,000) | 119,805+5,731 open | 79.0 %78.7 % to 79.2 % | +2.0 %+1.7 % to +2.4 % | +12.2 %+11.3 % to +13.0 %90,066 priced · +9.9 % vs entry on these |
| Normal size | 1,844,689+138,509 open | 61.4 %61.3 % to 61.4 % | −7.2 %−7.3 % to −7.0 % | −7.9 %−8.2 % to −7.6 %834,052 priced · −0.8 % vs entry on these |
| No baseline (under 10 earlier positions in 90 d) | 21,798+1,588 open | 57.6 %56.9 % to 58.2 % | −4.8 %−6.8 % to −2.7 % | −6.8 %−9.1 % to −4.4 %14,869 priced · −1.5 % vs entry on these |
By entry price
| Resolved | Hit rate | Copied at entry | Copied 1 h later | |
|---|---|---|---|---|
| Conviction · under 30¢ | 3,543+294 open | 16.8 %15.6 % to 18.1 % | −27.5 %−33.9 % to −21.2 % | −19.2 %−30.2 % to −8.2 %3,112 priced · +6.1 % vs entry on these |
| Conviction · 30–50¢ | 17,571+488 open | 43.5 %42.8 % to 44.2 % | −4.4 %−6.1 % to −2.8 % | −0.6 %−3.1 % to +1.9 %15,114 priced · +3.3 % vs entry on these |
| Conviction · 50–70¢ | 25,964+958 open | 64.7 %64.1 % to 65.3 % | +4.7 %+3.8 % to +5.7 % | +11.2 %+9.5 % to +13.0 %21,712 priced · +6.7 % vs entry on these |
| Conviction · 70¢ and up | 72,727+3,991 open | 95.6 %95.4 % to 95.7 % | +4.1 %+3.9 % to +4.3 % | +18.4 %+17.4 % to +19.3 %50,128 priced · +13.6 % vs entry on these |
| Normal · under 30¢ | 375,254+31,329 open | 9.9 %9.8 % to 10.0 % | −41.1 %−41.8 % to −40.3 % | −48.0 %−49.1 % to −46.9 %170,429 priced · −13.7 % vs entry on these |
| Normal · 30–50¢ | 340,147+13,191 open | 40.8 %40.6 % to 41.0 % | −7.7 %−8.1 % to −7.3 % | −10.1 %−10.7 % to −9.4 %172,420 priced · −4.1 % vs entry on these |
| Normal · 50–70¢ | 360,932+15,947 open | 64.5 %64.4 % to 64.7 % | +5.1 %+4.9 % to +5.4 % | +1.3 %+0.8 % to +1.8 %176,511 priced · +0.4 % vs entry on these |
| Normal · 70¢ and up | 768,356+78,042 open | 94.0 %94.0 % to 94.1 % | +3.9 %+3.8 % to +4.0 % | +9.8 %+9.5 % to +10.0 %314,692 priced · +7.3 % vs entry on these |
Conviction bets right now
The same rule applied live: tracked traders who opened a new position in the last 7 days at 3× their own usual size or more. A market card shows a chip when such bets land on it. Whether that means anything is the table above, not the size.
60 bets in the last 7 days by 27 traders · 1,663 tracked wallets have a baseline (10+ positions in 90 d), 462 do not · updated 2026-10-25 07:18 UTC
- rabbitfoot1 bought Yes in Will Estoril Praia win on 2026-10-25?at 51¢ · 2026-10-25 07:16 UTC · 6 fills · their usual position: median $286, p75 $1.6K$5.5K3.5× their p75
- UpTheBlues bought Under in Urawa Red Diamonds vs. V-Varen Nagasaki: O/U 3.5at 96¢ · 2026-10-25 07:16 UTC · their usual position: median $59, p75 $303$1.8K6.1× their p75
- UpTheBlues bought Under in FC Tōkyō vs. FC Mito Holly Hock: O/U 3.5at 71¢ · 2026-10-25 07:09 UTC · 3 fills · their usual position: median $59, p75 $303$1.1K3.7× their p75
- THG. bought Up in Bitcoin Up or Down - October 25, 3:05AM-3:10AM ETat 99¢ · 2026-10-25 07:08 UTC · 15 fills · their usual position: median $117, p75 $552$2.5K4.6× their p75
- YYYYY999 bought FUT Esports in Counter-Strike: FUT Esports vs Liquid (BO3) - PGL Masters Bucharest Group Stageat 73¢ · 2026-10-25 07:07 UTC · 11 fills · their usual position: median $2.0K, p75 $13K$62K4.8× their p75
- outloudm bought Yulia Putintseva in Pan Pacific Open, Qualification: Sayaka Ishii vs Yulia Putintsevaat 67¢ · 2026-10-25 07:07 UTC · their usual position: median $838, p75 $1.1K$5.0K4.7× their p75
- RN1 bought Over in Kawasaki Frontale vs. Fagiano Okayama: O/U 2.5at 54¢ · 2026-10-25 07:07 UTC · 8 fills · their usual position: median $237, p75 $1.4K$4.2K3.1× their p75
- UpTheBlues bought No in Urawa Red Diamonds vs. V-Varen Nagasaki: Both Teams to Scoreat 73¢ · 2026-10-25 07:06 UTC · 2 fills · their usual position: median $59, p75 $303$2.3K7.6× their p75
- THG. bought Up in Bitcoin Up or Down - October 25, 3:00AM-3:05AM ETat 99¢ · 2026-10-25 07:03 UTC · 6 fills · their usual position: median $117, p75 $552$2.5K4.6× their p75
- UpTheBlues bought Harriet Dart in Guangzhou Open, Qualification: Harriet Dart vs Mei Yamaguchiat 94¢ · 2026-10-25 07:01 UTC · 2 fills · their usual position: median $59, p75 $303$8.5K28.2× their p75
- UpTheBlues bought Shenzhen Leopards in Shenzhen Leopards vs. Beijing Royal Fightersat 52¢ · 2026-10-25 07:01 UTC · 18 fills · their usual position: median $59, p75 $303$2.5K8.1× their p75
- ghostorder bought Tatiana Prozorova in Guangzhou Open, Qualification: Tatiana Prozorova vs Joanna Garlandat 89¢ · 2026-10-25 07:00 UTC · 14 fills · their usual position: median $183, p75 $716$7.7K10.7× their p75
- SnakeBall bought Under in Urawa Red Diamonds vs. V-Varen Nagasaki: V-Varen Nagasaki O/U 1.5at 92¢ · 2026-10-25 06:59 UTC · 2 fills · their usual position: median $44, p75 $210$1.1K5.5× their p75
- hoodquant bought Up in Bitcoin Up or Down - October 25, 2:55AM-3:00AM ETat 97¢ · 2026-10-25 06:59 UTC · 6 fills · their usual position: median $70, p75 $291$1.1K3.9× their p75
- omnibus-076daa bought TSG 1899 Hoffenheim in Spread: Bayer 04 Leverkusen (-1.5)at 73¢ · 2026-10-25 06:56 UTC · their usual position: median $84, p75 $331$1.4K4.3× their p75
- 0x50f7 bought Down in Bitcoin Up or Down - October 25, 2:50AM-2:55AM ETat 100¢ · 2026-10-25 06:53 UTC · 9 fills · their usual position: median $20, p75 $235$6.0K25.6× their p75
- BoneOhio bought Down in Bitcoin Up or Down - October 25, 2:50AM-2:55AM ETat 99¢ · 2026-10-25 06:52 UTC · 9 fills · their usual position: median $73, p75 $260$1.6K6.3× their p75
- 0x50f7 bought Up in Bitcoin Up or Down - October 25, 2:45AM-2:50AM ETat 100¢ · 2026-10-25 06:49 UTC · 6 fills · their usual position: median $20, p75 $235$3.9K16.4× their p75
- BoneOhio bought Up in Bitcoin Up or Down - October 25, 2:45AM-2:50AM ETat 99¢ · 2026-10-25 06:49 UTC · their usual position: median $73, p75 $260$2.5K9.8× their p75
- LuckyLoo bought No in Will FC Mito Holly Hock win on 2026-10-25?at 91¢ · 2026-10-25 06:49 UTC · their usual position: median $21, p75 $85$6.7K79.0× their p75
- THG. bought Up in Bitcoin Up or Down - October 25, 2:45AM-2:50AM ETat 99¢ · 2026-10-25 06:49 UTC · 3 fills · their usual position: median $117, p75 $552$2.5K4.6× their p75
- Cannae bought No in Will Getafe CF win on 2026-10-25?at 87¢ · 2026-10-25 06:41 UTC · their usual position: median $44, p75 $182$3.0K16.4× their p75
- CosmicCedar bought No in Will Vanderbilt have more than 7.5 wins during the 2026 CFB season?at 100¢ · 2026-10-25 06:39 UTC · their usual position: median $71, p75 $242$1.7K7.1× their p75
- 0x50f7 bought Down in Bitcoin Up or Down - October 25, 2:35AM-2:40AM ETat 100¢ · 2026-10-25 06:38 UTC · 9 fills · their usual position: median $20, p75 $235$3.0K12.8× their p75
- SnakeBall bought Under in Cerezo Ōsaka vs. JEF United Ichihara Chiba: 1st Half O/U 2.5at 97¢ · 2026-10-25 06:36 UTC · 20 fills · their usual position: median $44, p75 $210$2.8K13.4× their p75
- looppau bought Dayana Yastremska in Guangzhou Open, Qualification: Dayana Yastremska vs Emerson Jonesat 100¢ · 2026-10-25 06:32 UTC · 13 fills · their usual position: median $69, p75 $300$2.5K8.3× their p75
- GoalLineGhost bought Over in FC Internazionale Milano vs. ACF Fiorentina: O/U 0.5at 97¢ · 2026-10-25 06:31 UTC · 4 fills · their usual position: median $143, p75 $631$2.1K3.3× their p75
- BoneOhio bought Up in Bitcoin Up or Down - October 25, 2:15AM-2:30AM ETat 99¢ · 2026-10-25 06:29 UTC · their usual position: median $73, p75 $260$3.0K11.4× their p75
- hoodquant bought Up in Bitcoin Up or Down - October 25, 2:15AM-2:30AM ETat 99¢ · 2026-10-25 06:29 UTC · 13 fills · their usual position: median $70, p75 $291$2.1K7.1× their p75
- SnakeBall bought Under in Kataller Toyama vs. Hokkaidō Consadole Sapporo: Kataller Toyama O/U 2.5at 98¢ · 2026-10-25 06:29 UTC · 3 fills · their usual position: median $44, p75 $210$1.8K8.4× their p75
- The-Joker bought No in Will Student List – Students Win win the most seats in the 2026 Serbian parliamentary election?at 72¢ · 2026-10-25 06:27 UTC · 21 fills · their usual position: median $940, p75 $2.6K$14K5.5× their p75
- hoodquant bought Down in Bitcoin Up or Down - October 25, 2:20AM-2:25AM ETat 97¢ · 2026-10-25 06:24 UTC · 13 fills · their usual position: median $70, p75 $291$1.1K3.8× their p75
- UpTheBlues bought Under in Nagoya Grampus vs. Yokohama F·Marinos: O/U 4.5at 95¢ · 2026-10-25 06:19 UTC · their usual position: median $59, p75 $303$1.5K4.8× their p75
- SnakeBall bought Under in Cerezo Ōsaka vs. JEF United Ichihara Chiba: O/U 4.5at 90¢ · 2026-10-25 06:19 UTC · their usual position: median $44, p75 $210$1.3K6.2× their p75
- GoalLineGhost bought Angers SCO in Spread: FC Lorient (-1.5)at 82¢ · 2026-10-25 06:17 UTC · 3 fills · their usual position: median $143, p75 $631$2.1K3.2× their p75
- UpTheBlues bought Under in Sagan Tosu vs. Vanraure Hachinohe FC: O/U 5.5at 98¢ · 2026-10-25 06:17 UTC · their usual position: median $59, p75 $303$2.2K7.4× their p75
- Cannae bought No in Will Sunderland AFC win on 2026-10-25?at 67¢ · 2026-10-25 06:17 UTC · 3 fills · their usual position: median $44, p75 $182$2.0K10.8× their p75
- 0x5b624e3E0426F7AA2Ca57F7451fbcEa4b125D87f-1769796221989 bought No in Will Sunderland AFC win on 2026-10-25?at 67¢ · 2026-10-25 06:17 UTC · their usual position: median $390, p75 $700$2.7K3.8× their p75
- UpTheBlues bought No in Will Vegalta Sendai win on 2026-10-25?at 87¢ · 2026-10-25 06:15 UTC · 3 fills · their usual position: median $59, p75 $303$1.6K5.3× their p75
- UpTheBlues bought Yusuke Takahashi in Seoul: Christian Langmo vs Yusuke Takahashiat 53¢ · 2026-10-25 06:14 UTC · 23 fills · their usual position: median $59, p75 $303$2.0K6.7× their p75
A conviction bet is a NEW position of at least 3× the trader's own 75th-percentile position cost over the previous 90 days, and at least $1,000; adding to an old position does not count. Size is not skill: see the research table above for what such bets returned. Past results, not advice.
How it is computed
- Entries. A market the trader bought a side of and ended the dataset a net buyer of (kept more than 1 % of the shares bought). Markets only sold are not entries. Both sides kept with the smaller side at 25 % or more of the larger side's money is a hedge, left out and counted. Markets bought and sold out of entirely are "exited": counted apart, valued as if held to resolution. Buys after the market settled are dropped.
- Entry price and time. The volume-weighted average of the trader's buys of that side over the whole dataset, and the first buy. A copier at the first buy pays the first price, so this flatters traders who average down well.
- Return. $1 buys 1 / (price + 1¢ + fee) shares; the fee is Polymarket's taker fee for the market's category at that price (schedule). A share pays $1 if the side won, $0 if it lost, $0.50 when Polymarket resolved the market 50-50 (neither a win nor a loss). Resolution comes from the market's stored winner, else settlement prices, else the site's resolved list.
- Copied 1 h later. The same, at the dataset's first fill on the market by anyone between 1 h and 24 h after the entry (fills under 5 shares or $1 set no price). Without one, the entry has no later figure — most 15-minute and in-play markets had ended. A later price is only as good as the fills' timestamps: the site estimates them from block numbers, and rows read before the October 22 2026 repair can be hours off for old fills, so for 15-minute crypto markets treat the later-price column with suspicion until the dataset has been re-exported.
- Intervals. 95 %: Wilson for hit rates, a normal approximation for means; none under 10 resolved entries. Per-wallet distributions use wallets with 30+ resolved entries. Many entries of one trader fall in one event (sets of a match, legs of a parlay), so the real uncertainty is larger than the intervals.
- Hold times are the time from the first buy to the settlement, read off a histogram with edges at 1, 3, 6, 12, 24 … hours; the medians are approximate.
- Live conviction. The worker computes each tracked wallet's median and p75 position cost from its positions touched in the last 90 days (every 6 h) and checks the last 24 h of buys every 2 minutes; a position counts as new when at most 5 % of its dollars were bought before the window.
- What this is not. Not a ranking of traders to copy. Selecting the best past copy return out of thousands of wallets and expecting it to continue is the mistake every table here is built to show.
Data: /api/research/traders/copy · /api/research/traders/conviction · All research