“I analyzed 100 Chinese traders”: the two it names won about as often as their prices said
An article on X says a group of Chinese-speaking Polymarket traders keeps landing on the right side through discipline, steady sizing and better information
The claim“A statistically noticeable group with a similar style (medium size, multi-market, focus on sports and asymmetric information, low emotionality + technical automation) is currently outperforming.” Its example: “36 of its 50 biggest wins are ‘No’ on a big favorite… The account keeps betting that it’s overpriced, and over thousands of matches that adds up.” — X article (@BimbaCrypto)
The short version
The article was posted on X on October 26, 2026 and had about 5,900 views when we read it. Its author says they went through about 100 Polymarket accounts that “appear tied to the Chinese-speaking segment” and found a common style: systematic $2–15k positions, consistent sizing, no bigger bets after a winning streak, and an edge from Asian football, Challenger tennis and Chinese-language information. It names two accounts as examples: jtwyslljy, up $1.95M, which mostly bets “No” on famous clubs winning, and Woaifacai, up $909.2K.
The 100 accounts are not listed, and the article does not say how it decided they were Chinese, so that part cannot be checked. We downloaded the two named wallets’ full histories from Polymarket’s data API: 109,620 records for jtwyslljy from May 2025 and 10,761 for Woaifacai from June 2026, plus the result of every market they traded. The article’s figures for both are right: jtwyslljy was at $1.95M before fees on the day, with $104M of volume and 105,543 trades, and Woaifacai’s $909.2K, $30.2M volume and $354.7K biggest win all match. After fees, jtwyslljy was at $1,710,332 and Woaifacai is now at $792,254.
The pattern the article picks out for jtwyslljy is real but proves less than it says. We count 37 of its 50 biggest wins as “No” on a team winning, bought at 33.8¢ on average; the article says 36 at 34¢. But a list of biggest wins only shows the bets that won. Across all 546 of its “No” bets, it won 185 times where their prices implied 189, so the market priced those favorites about right. The bets made $1.46M overall because more money happened to sit on the winners: the 36 best made $5.02M and the other 510 lost $3.56M. These are 546 bets, not “thousands of matches”.
The records also contradict the article’s picture of steady sizing and “no blow-ups”. After its best month, January, jtwyslljy’s larger bets roughly tripled; it then lost $852,478, about half its profit, between March 7 and April 26, and $675,660 more between August 10 and October 19. Woaifacai’s stakes ran from about $260 to $419,498. It came back in October with bigger stakes than in the World Cup, lost $300,674 on one tennis match and $245,238 on one NBA game on October 24, and gained $432,088 the next day. Neither wallet trades much of what the article says these traders specialise in: jtwyslljy put 0.3% of its money into Asian football and almost none into Challenger tennis, and its largest profit came from England’s Premier League.
Making a lot of money is not the same as having an edge. If every price jtwyslljy paid had been fair, the usual swing either way from luck alone, at the size it bets, would be about $1.4M, so a $1.7M profit would happen by chance roughly once in eight tries. Its share of wins, 43.1% against 41.5% implied, is close to the prices it paid. Woaifacai made most of its money in five weeks of the World Cup, and one bet, $419,498 on Switzerland, made $354,708 of it. Verdict: the two wallets are real and profitable, but their records do not show that they beat the market beyond luck. They do not show the discipline the article describes, and the group of 100 cannot be checked.
Technical appendix
Exhibit A jtwyslljy's “No” bets on teams to win
Wallet 0x9cb990f1862568a63d8601efeebe0304225c32f2 (jtwyslljy), linked from the article. These are all its bets on "No" in markets of the form "Will [team] win on [date]?", the pattern the article describes. Price is the average paid per share before the fee; "implied wins" is the number expected if each bet won exactly as often as its price. P/L is cash: sales, payouts and merges minus what was spent.
| Price paid for "No" | Bets | Spent | P/L | Wins | Implied wins |
|---|---|---|---|---|---|
| under 20¢ | 106 | $883,692 | −$199,144 | 11 | 14.0 |
| 20–30¢ | 124 | $2,824,232 | +$289,907 | 29 | 31.4 |
| 30–40¢ | 132 | $3,868,601 | +$1,101,927 | 46 | 45.2 |
| 40–50¢ | 88 | $2,640,066 | +$358,165 | 44 | 39.1 |
| 50¢ or more | 96 | $1,746,506 | −$92,348 | 55 | 59.2 |
| All "No" bets | 546 | $11,963,096 | +$1,458,508 | 185 | 188.9 |
The 36 most profitable of these bets made +$5,019,140; the other 510 lost $3,560,632. Its largest losses on the same pattern include $140,400 on "No" to Barcelona winning on March 18 and $134,899 on "No" to Atlético Madrid on March 10. Everything else the wallet bought, $20,521,340 across 3,130 positions, made +$211,407.
Exhibit B jtwyslljy over time
jtwyslljy's profit after fees, from Polymarket's daily profit history (data-api.polymarket.com/v2/user-pnl, taken at 00:00 UTC).
| Date | Profit after fees | Change |
|---|---|---|
| Oct 1, 2025 | $55,728 | |
| Jan 1, 2026 | $160,117 | +$104,389 |
| Mar 7, 2026 | $1,673,378 | +$1,513,261 |
| Apr 26, 2026 | $820,900 | −$852,478 |
| Jun 1, 2026 | $1,419,995 | +$599,095 |
| Aug 10, 2026 (highest) | $2,277,732 | +$857,737 |
| Oct 19, 2026 | $1,602,072 | −$675,660 |
| Oct 26, 2026 (day of the article) | $1,710,332 | +$108,260 |
| Oct 28, 2026 | $1,687,254 | −$23,078 |
Its worst single day was April 25, 2026: −$271,001. Stake size by the month positions were opened (the 90th percentile is the size that 9 in 10 positions were below):
| Opened | Positions | Median stake | 90th percentile | P/L |
|---|---|---|---|---|
| Dec 2025 | 246 | $1,267 | $26,992 | −$63,183 |
| Jan 2026 | 273 | $1,486 | $22,000 | +$631,871 |
| Feb 2026 | 96 | $4,838 | $67,812 | +$356,793 |
| Mar 2026 | 254 | $2,100 | $59,310 | −$32,188 |
| Apr 2026 | 250 | $3,753 | $42,720 | −$36,797 |
| Sep 2026 | 286 | $4,352 | $22,010 | −$273,986 |
| Oct 2026 | 431 | $3,038 | $11,048 | −$249,914 |
After its best month, January, its larger bets roughly tripled in size in February and March, and March and April were its worst stretch.
Exhibit C Woaifacai
Wallet 0x9b89ef1f344d3e17b37ccdff3bccf62d04f04244 (Woaifacai), linked from the article. 10,761 records from June 17 to October 28, 2026.
| Positions opened | Positions | Spent | Median stake | Largest stake | P/L |
|---|---|---|---|---|---|
| Jun 17 – Jul 15 (World Cup) | 199 | $5,121,189 | $8,868 | $419,498 | +$670,667 |
| Oct 13 – the article | 179 | $5,823,368 | $13,866 | $300,674 | +$135,383 |
| After the article | 5 | $37,757 | $9,486 | $11,777 | −$13,966 |
It opened no positions between July 15 and October 13. Its stakes ran from about $260 (1 in 10 were smaller) to $419,498; the median was $11,142, and 1 in 10 were above $79,228. Its largest results:
| Market | Pick | Price | Stake | P/L |
|---|---|---|---|---|
| Will Switzerland win on 2026-07-02? | Yes | 48.9¢ | $419,498 | +$354,708 |
| Notre Dame vs. BYU (Oct 18) | BYU | 50.9¢ | $187,460 | +$175,395 |
| Japan Women's Open: Ann Li vs Ashlyn Krueger (Oct 24) | Ann Li | 69.5¢ | $300,674 | −$300,674 |
| Grizzlies vs. Warriors (Oct 24) | Warriors | 61.0¢ | $245,238 | −$245,238 |
| Fulham vs. Hull City: O/U 2.5 (Oct 17) | Over | 51.8¢ | $133,734 | −$133,734 |
By Polymarket's daily figures (UTC days) it lost $269,626 on October 17, lost $289,529 over October 23 and 24, and gained $432,088 on October 25, the day before the article. Its tennis bets, including the Asian swing the article mentions, lost $231,718 in total.
Exhibit D Where jtwyslljy's money went
The article says these traders' biggest results come from Asian football, lower-tier European cups, Challenger and ITF tennis, and political markets. Where jtwyslljy's $33.0M actually went, by the event each market belongs to:
| Markets | Positions | Spent | Share | P/L |
|---|---|---|---|---|
| Tennis, main tour (Wimbledon, the Slams, Masters, Asian tour events) | 1,075 | $5,774,745 | 17.5% | +$299,508 |
| England's Premier League | 246 | $4,453,758 | 13.5% | +$1,008,903 |
| Esports | 764 | $4,249,938 | 12.9% | −$427,220 |
| World Cup | 231 | $3,654,575 | 11.1% | +$384,500 |
| Spain's La Liga | 106 | $2,188,722 | 6.6% | −$476,122 |
| Champions League | 94 | $1,743,469 | 5.3% | −$283,223 |
| Bundesliga, Serie A, Ligue 1, Europa League | 156 | $2,251,786 | 6.8% | +$1,017,623 |
| US sports (NFL, NBA, MLB, NHL, UFC, college) | 907 | $7,468,837 | 22.6% | +$381,861 |
| Asian football | 38 | $114,188 | 0.3% | +$11,716 |
| Challenger and ITF tennis | 5 | $12,205 | 0.04% | −$3,251 |
| Everything else | 226 | $1,119,384 | 3.4% | −$209,533 |
We found no political markets in its history.
Exhibit E The article's claims, one by one
| The article says | The record shows |
|---|---|
| jtwyslljy: +$1.95M all-time, "built steadily", "drawdowns but no blow-ups" | +$1.95M is right before fees on the day; after $236,197 of fees it was $1,710,332. It lost $852,478, half its profit at the time, between March 7 and April 26, 2026, and $675,660 between August 10 and October 19. |
| jtwyslljy: $104M volume, 105,000+ trades, 3,483 markets | Matches: 105,543 trades; 3,498 markets by October 28. |
| 36 of its 50 biggest wins are "No" on a big favorite, bought at 34¢ on average | We count 37, at 33.8¢. True, but a list of biggest wins only shows bets that won. Across all 546 such bets it won 185 times; their prices implied 189. |
| "Over thousands of matches that adds up" | 546 bets, not thousands. They made $1.46M, but the 510 outside the top 36 lost $3.56M. |
| Woaifacai: joined June 2026, +$909.2K on $30.2M volume, 345 markets, biggest win $354.7K | Matches. $909,209 is Polymarket's figure before fees at the midnight after the article; after $103,406 of fees the wallet is at $792,254. |
| Woaifacai made most of its money in the World Cup, then sat out the summer | True: +$670,667 from positions opened during the World Cup, and nothing opened from July 15 to October 13. |
| "Size is consistent"; systematic $2–15k positions | Woaifacai's stakes ran from about $260 to $419,498; 1 in 4 was over $37,000. |
| They "do not increase size after a winning streak" | jtwyslljy's larger bets roughly tripled after its best month. Woaifacai came back in October with bigger stakes than it used in the World Cup. |
| All-time leaderboard: rank 1 $22.1M, rank 50 $3.1M, rank 200 $0.9M | Matches the leaderboard on October 28: $22.05M, $3.10M, $0.94M. |
| About 100 accounts that "appear tied to the Chinese-speaking segment" | Not checkable: the article does not list them or say how they were identified. |
Exhibit F Links and sources
Both profile links in the article carry a via=search-gem tag. Polymarket says any profile page can be shared as a referral link, and that referrers earn 10% of the trading fees their sign-ups pay for 30 days (Polymarket's referral program); its documentation does not say how those links are formed, so we cannot confirm whether this tag is one. The article's other figures (market volumes, the share of traders in the red, bot shares of crypto volume) are attributed to Pew, The Block, The Defiant, Bloomberg, CNBC, Dune and Rest of World; we did not check them.