World Cup prediction market winner profiles: Behind a $470 million bet, someone made $8.25 million in two games

Author: Frank, PANews

On July 20, the USA-Canada-Mexico World Cup finally came to an end. This global spectacle that drew in hundreds of millions of football fans not only sparked massive attention on the field, but also tugged at the wallets of countless speculators off the field. Previously, PA Beacon had tallied the performance of “smart money” forecasts during the group-stage period. At the very start of the World Cup, large amounts in the prediction market showed no clear edge. When PA Beacon took a snapshot of the first 20 matches on June 17, 2026, the total pre-match BUY amount was $89.5457 million, with an amount hit rate of only 48.5%. Estimated by holding through to settlement, the overall loss was $1.7594 million, with an ROI of -2.0%.

A month later, the results changed. After all 104 matches of the World Cup were completed, PA Beacon recorded $474.04 million in total pre-match BUYs. Of this, $298.49 million was placed on the correct direction, and the amount hit rate rose to 62.97%. If all these positions were estimated to be held through to settlement after entry, total payout was about $502.93 million. The estimated net profit was $28.8858 million, with an ROI of 6.09%.

From losses to profits, it looks like large capital gradually regained judgment as the schedule progressed. But once broken down by account level, the answer isn’t that simple: a small number of wallets took most of the profits by relying on one or two heavily weighted matches; other accounts effectively bet across the entire World Cup, with higher hit rates, yet their final returns were relatively limited. The profit leaderboard and the ability leaderboard are not the same list.

This article’s data scope: covers 104 matches and 312 win/lose/draw markets. Trades came from the Polymarket Data API. Pre-screening condition: each trade amount must be no less than $5,000; only BUY transactions made before kickoff are counted. Multiple buys from the same wallet in the same match, same market, and same direction are combined. Profit and loss are estimated based on holding through to settlement. Midway sells, in-match trades, hedging across other markets, and fees are not included, so this is not the same as a trader’s real final P&L.

Figure 1: Overview of PA Beacon’s World Cup full-tournament capital observations. Data as of July 20, 2026.

Large capital executed a turnaround, but profits were concentrated in a few accounts

Among the 1,856 wallets that participated in pre-match BUYs, 1,003 were estimated to be profitable and 853 losing under the above methodology; profitable wallets accounted for 54.04%. Those profitable accounts collectively earned $99.6338 million, while the losing accounts collectively lost $70.7480 million. After offsetting, these “smart money” participants delivered a net profit of $28.8858 million in total.

More noteworthy is the concentration of profits. The top 5 profitable accounts together earned $37.7064 million, accounting for 37.85% of the gross profit of all profitable wallets; the top 10 together earned $55.1869 million, reaching 55.39%. In other words, the market’s 6.09% profitable expectation does not mean that most large capital consistently gained a stable advantage. Instead, a small number of heavily weighted winners clearly lifted the average performance.

In addition, the differences between matches were also huge. Belgium vs Egypt attracted $12.3905 million in pre-match BUYs, but only $0.6670 million was placed correctly, for an amount hit rate of just 5.4%. Belgium vs Senegal saw $12.0438 million in pre-match BUYs, of which $10.5326 million was placed correctly, for a hit rate of 87.5%. Similar capital size, entirely different outcomes. Large amounts do not equal correctness; what determines returns is still the entry direction, the odds, and the position sizing.

Figure 2: Accuracy of pre-match large capital by match.

One or two heavy bets created the most eye-catching winners

The account “mintblade” is the most typical case. This account participated in only 2 matches, with 3 aggregated positions. It投入ed $7.2889 million and generated an estimated profit of $8.2535 million, for an ROI of 113.23%, and an amount hit rate of 100%. One match—Iran vs New Zealand—accounted for a $6.7738 million profit, representing 82.1% of its estimated World Cup profit.

The logic behind his trades isn’t complicated. “mintblade” bought “Iran to not win” with $6.4705 million at an average price of about $0.49. The match ended 2-2, so this position was settled at $1, yielding an estimated payout of $13.2443 million. Afterwards, he bought “Uruguay to not win” and a small share of draw exposure in Saudi Arabia vs Uruguay; the two together earned another estimated $1.4797 million. With both matches correctly bet, he landed among the most profitable accounts in this World Cup. But why could he be so bold to heavily re-stake on just these two matches is intriguing.

By contrast, the address named “GRIMDRIP” was even more extreme. It participated in only one match, Czech Republic vs South Africa, and bought two related directions: “Czech Republic to not win” and “match draw.” The final score was 1-1, and both positions were correct. It invested $5.8490 million and produced an estimated profit of $7.4497 million, for an ROI of 127.37%.

Behind these big bets lies conviction about the match outcomes, leading to speculation that there may be unknown secrets behind the seemingly mysterious profitability of these addresses.

Another account, “DEEDDIT,” represents a different heavy-betting approach. It covered 8 matches and 9 aggregated positions, investing $20.9493 million and estimating profit of $8.0636 million. In the round of 32, Belgium beat Senegal 3-2. In the Belgium win market, it invested $7.1610 million, earning an estimated $7.7495 million from a single bet. In the semifinals, France lost 0-2 to Spain; it bought “France to not win” and earned another $3.4259 million.

However, “DEEDDIT” wasn’t correct throughout. It bet the wrong draw direction in matches like Mexico vs Ecuador and Switzerland vs Colombia, and several major losses exceeded $4.3000 million. Its single largest winning trade was equivalent to 96.1% of its World Cup net profit. Ultimately it ranked high, but its high performance depended heavily on just the Belgium vs Senegal match.

The accounts “sparklingwater123” and “endlessFate” show similar features. The former covered only 2 matches, investing $8.5005 million and estimating profit of $7.7469 million; the latter covered 5 matches, investing $11.4454 million and estimating profit of $6.1929 million. They repeatedly bought “not win” or draw when the prices were too high for the popular teams to take the win, but they were not mechanically opposing favorites—they expressed their judgment concentratedly in just a handful of matches. Overall, operations by these “whales” may involve multiple-platform hedging or price arbitrage.

Figure 3: Aggregated positions in the largest profit/loss rankings.

The “gambler” splashed $15.99 million to participate in 97 predictions, but its returns were less than others’ by just two games’ worth

If we look only at the profit leaderboard, the account “swisstony” isn’t especially prominent. But in terms of how frequently it participated, it is absolutely king. Its predictions covered 97 matches with 267 aggregated positions. It bought $15.9912 million pre-match, reaching an amount hit rate of 79.29%, estimated profit of $1.2482 million, and ROI of 7.81%. However, even though the number of correctly predicted matches was far greater than “mintblade” and “GRIMDRIP,” its profitability was only about 15% of the former’s.

The difference mainly comes from position structure. “swisstony’s” largest estimated profit per single bet was only $0.2227 million, while its largest loss was $0.3058 million. Its largest profitable position accounted for just 17.84% of its estimated net World Cup profit. It didn’t rely on a single match to stage a turnaround; instead, it accumulated gains across many matches with smaller position sizes. The outcome wasn’t as dramatic, but it more clearly demonstrates whether an account has a sustained advantage.

The accounts “AV23IUa” and “Latina” also belong to profitable accounts with relatively broader coverage. “AV23IUa” participated in 46 matches with 46 positions, investing $2.1245 million and estimating profit of $0.9060 million, for an ROI of 42.65%; “Latina” covered 11 matches, investing $3.4293 million and estimating profit of $1.2920 million, with an amount hit rate of 88.64%. In both cases, their profits weren’t concentrated in a single match.

Another noteworthy account is “zhqzhq.” Using the criterion of whether the funds on the correct direction exceed half per match, it covered 14 matches—all judged correctly. But with $1.0182 million invested, it generated only $0.0557 million in estimated profit, for an ROI of 5.47%. The hit rate is very high, but the returns are not. The reason is that when it bought, matches were often already in a highly certain phase.

Figure 4: Ranking of pre-match BUY amounts by account.

Outside the winners, losing accounts provide a more direct contrast. “LEEEROYJENKINS” had an estimated profit of $4.7976 million in Australia vs Turkey, but then it heavily bet Belgium to win vs Egypt; the match ultimately ended 1-1, resulting in an estimated loss of $8.3943 million in that single match. The prior big profit was offset by a single wrong bet, and its World Cup journey ultimately ended in an estimated loss of $3.2464 million.

More losing players mainly suffered from consecutively wrong predictions. The account “coldsway” invested $13.7255 million across 9 matches with an amount hit rate of only 26.22%, resulting in an estimated loss of $7.3437 million—the largest account-level loser. The account “FlickRaw” participated in only 2 matches, and in both it was wrong; it invested $4.7983 million, which went to zero.

These cases show that beneath the surface, these whales are still gamblers who bet big. One overly heavy wrong position is enough to wipe out multiple prior correct calls.

The championship favorites in the eyes of whales: France was seriously overestimated, Spain is the steadiest

As the excitement of the USA-Canada-Mexico World Cup faded, this 40-day “prediction frenzy” also came to an end. Revisiting these on-chain data, we see an extremely real yet brutal speculator world: here, capital size doesn’t equal foresight; “smart money” can also be punished by collective bias. Here, some people get rich overnight from one or two upsets, while others who seemingly thought everything through still lose to one heavy-position mistake.

Figure 5: Support capital for direct wins of France, Spain, Argentina, and England across five rounds of knockout matches. Unit: millions of USDC.

Spain: Among the four teams, Spain is the most stable sample in terms of capital judgment. In five knockout matches, the side with the larger amount was always the one that won. But the support strength wasn’t consistently rising: from the round of 32 to the quarterfinals, capital supporting Spain to win rose from $1.1430 million to $2.3900 million. In the semifinal against France, it fell to $0.8790 million, and the median buy cost was only $0.297. Ultimately, Spain won 2-0, and this actually became the match with the largest odds space. However, this $0.8790 million was not a broad consensus: the largest account contributed 46.7%, and the top five accounts accounted for 73.5%. Going 5-for-5 suggests that a few heavily positioned funds kept getting the direction right in a row, while the overall market still judged incorrectly.

Argentina: Argentina’s curve is the opposite. Although the team fought back and reached the final, the direct support funds gradually slid downward. In the round of 32, support amounted to $2.8040 million with a median cost of $0.86; in the semifinals it dropped to $0.3930 million and $0.315, respectively. Before the final, the funds supporting Argentina to win were only $0.4340 million, while the opposing funds reached $1.5180 million, making support just 22.2%. This time, capital shifted early to “not win” and was correct. However, the top five opposing accounts held 92.5% of that side’s funds, and the largest account alone accounted for 44.4%—still indicating that the conclusion was dominated by a minority of large holders.

Figure 6: Median buy costs of capital supporting the direct win markets of the four teams, weighted by trade amount.

France: France shows the most obvious pattern of “a minority correct, the crowd wrong.” In the semifinal vs Spain, capital supporting France to win was $1.6260 million, while opposing funds were $7.1940 million. The big money won, but the capital composition mainly came from a single wallet contributing about $5.7600 million, which was 80.1% of the opposing funds. In the third/fourth place match, capital shifted sharply back to France: $2.9997 million supported France to win, $0.1793 million opposed, and support share rose to 94.4%. The result: France lost 4-6 to England.

England: England is the most easily underestimated team. In the round of 16 vs Mexico, the capital supporting England to win was $1.1440 million, while opposing funds were $1.1870 million. The amount consensus leaned toward “not win,” yet England advanced 3-2. In the third/fourth place match vs France, capital supporting England was only $0.1590 million, while opposing funds were $1.0320 million; support share was just 13.4%, and the median buy cost was only $0.210. Finally, England won 6-4. But neither of these was a case of “smart money collectively getting it wrong”: in the third/fourth place round, the top five opposing accounts made up 98.3% of the funds on that side, and the largest account accounted for 60.4%—a minority of heavily positioned accounts pulled the amount consensus toward the wrong direction.

Figure 7: In the same team’s direct-win market, whether the larger amount side corresponds to the final match result.

As the noise of the USA-Canada-Mexico World Cup died down, this 40-day “prediction carnival” also came to a close. In replaying these on-chain data, we see an extremely real yet cruel speculation world: here, the size of capital doesn’t equal prophecy, and “smart money” can be反噬 by group biases; here, some people get rich by one or two upsets overnight, while others who scheme carefully still can’t beat one heavy-position mistake.

Once you cut through the fog of “survivorship bias,” those most eye-catching profit myths still have the same blood-curdling core: reckless big-bets. In football, there is no eternal champion, and the same is true for prediction markets.