Share
The 2026 FIFA World Cup was supposed to be the prediction markets’ coming-out party. In many ways, it still is. Trading volume has surpassed $5 billion across Polymarket and Kalshi in the first two weeks alone. Kalshi recorded three consecutive days with over $1 billion in daily volume during the group stage.
The World Cup champion market on both platforms combined has crossed $2.34 billion in cumulative trading. Macquarie forecasts total worldwide wagering on the tournament at more than $50 billion, the largest betting event on record.
But underneath the headline numbers, a more complicated story is emerging. Prediction markets offer an unusual level of transparency. Every position is visible onchain and every trade can be traced through blockchain analytics. That same transparency is also documenting a wave of multi-million-dollar losses that traditional sportsbooks rarely expose publicly.
The World Cup is not just growing prediction markets. It is exposing in real time exactly how dangerous they can be for undisciplined or unlucky participants.
The case that crystallized the risk problem involves a pseudonymous Polymarket account called FlickRaw. The account joined Polymarket this month and had made just three predictions as of Monday evening.
🚨BREAKING: Someone named “flickraw” just put $2.7M on the Netherlands to win their match vs Japan
This pays out is $5,827,897.87 on Polymarket pic.twitter.com/fHxKNM6fem
— Polymarket Sports (@PolymarketSport) June 14, 2026
Less than 24 hours after losing $2.7 million on a Netherlands position, FlickRaw risked another $1.5 million on Belgium to beat Egypt. Egypt’s 1-1 draw wiped out that position too. The back-to-back defeats cost FlickRaw roughly $4.2 million based on the stakes Polymarket publicized before each match. Polymarket promoted both trades publicly before kickoffs.
That last detail matters. Polymarket actively promoted the positions as pre-match content, turning a participant’s outsized wager into marketing material. The practice generated attention before the matches and documented the losses after. The promotion highlighted how large trades can become marketing content while also leaving losses permanently visible onchain.
FlickRaw is not an isolated case. It sits at the top of a loss ledger that has grown substantially during the tournament.
Trader “Supersob” lost roughly $5.8 million in a 24-hour period, wagering more than $2.6 million against Norway winning a match. Norway won 3-2, erasing approximately $2.7 million from his position. A second large bet against Iran compounded the damage.
Arkham Intelligence ranked Supersob as the third-biggest all-time loser on Polymarket with a 25% win rate across 5,230 trades and $16.7 million in total volume since June 2026 alone.
He lost $5.8 Million in a day after betting against Norway, then against Iran.
‘Supersob’ had both of his largest betting losses in the last 24 hours.
He’s the third biggest loser on Polymarket ALL-TIME. pic.twitter.com/p2HaKzVgFJ
— Arkham (@arkham) June 23, 2026
Trader “weatherman12” lost approximately $1.8 million when England drew 0-0 with Ghana in Group L, a scoreless result that wiped out his entire position on an England win.
An anonymous Polymarket user lost $13 million after betting Belgium would beat Egypt. The teams drew 1-1 instead. That is the same match that caught FlickRaw’s $1.5 million. Two separate traders, apparently with no connection to each other, placed material positions on Belgium and both lost.
The wins are equally dramatic but considerably rarer. One trader made $9.24 million in a single day after winning four consecutive bets, including a $7.03 million bet that Iran would lose against New Zealand.
Trader “fishalive” turned approximately $400,000 betting against a Spanish victory into roughly $4 million when Spain drew 0-0 with Cabo Verde, a result prediction markets had priced at roughly a 6.6% probability. Total gains from that match reached approximately $4.7 million after he added a second position.
One trader just turned $400k into $4.7 million in 90 minutes during the World Cup.
On Polymarket, the wallet “fishalive” went all-in betting Spain would NOT beat Cabo Verde.
Spain was sitting at 90% implied probability.
The match ended in a 2-2 draw.
He cleaned up roughly $4… pic.twitter.com/bj6s8zu9gI
— Xander (@baltimore465) June 22, 2026
The asymmetry in these stories is not accidental. It reflects the mathematical structure of prediction markets: low-probability outcomes that pay at long odds generate outsized gains for those who bet against the crowd. High-probability outcomes that pay at short odds generate multi-million-dollar losses for those who bet on the favorite with maximum size and no hedge.
Prediction markets are not like traditional sports betting in one critical structural way. At a sportsbook, a bettor pays fixed odds, knows the maximum loss at the moment of placing the bet, and cannot add to a losing position mid-event in the same market. The loss is bounded at entry.
Prediction market binary contracts work differently. A trader buying “Yes” on Belgium at 90 cents is saying Belgium has greater than a 90% chance of winning. If they are right, they earn 10 cents per contract. If they are wrong, they lose 90 cents per contract.
That asymmetric payout structure means that betting on strong favorites with maximum size, exactly what multiple World Cup traders have done, creates a risk profile where the upside is small and the downside is total.
The psychology reinforces the danger. A trader who has bet heavily on a 90% probability outcome and seen it fail once faces a specific cognitive trap. The temptation to re-enter the same type of position, reasoning that the sequence of losses is unlikely to repeat, is precisely the behavior documented in compulsive gambling research. FlickRaw re-entered within 24 hours of a $2.7 million loss. The second position cost $1.5 million more.
A Bloomberg analysis published in April 2026 found that since January 2025, more than 100,000 Polymarket accounts had recorded losses of at least $1,000, nearly double the number of wallets posting comparable gains.
A University of Toronto-led academic study covering 2.4 million users found that 68.8% had lost money since 2022, and that users who lose money disproportionately trade at extreme prices, below 10 cents or above 90 cents. That statistical pattern describes exactly what the highest-profile World Cup losers were doing.
The blockchain infrastructure underlying Polymarket makes the losses public in a way no traditional betting platform ever has. Every position is visible. Every loss is traceable. Arkham Intelligence can rank the all-time biggest losers. Polymarket itself can surface large positions as pre-match content and drive engagement through the spectacle.
That transparency was initially framed as one of prediction markets’ defining advantages over opaque sportsbooks. It makes manipulation harder to hide and market pricing more reliable when large positions are public.
But it also creates a dynamic that regulators and harm-reduction advocates are beginning to scrutinize closely: the publicization of extreme losses as entertainment, which normalizes the scale of risk and may encourage risk-taking by normalizing unusually large wagers.
European regulators have responded with concern. Switzerland’s Gespa regulatory authority stated that platforms including Polymarket offer 24/7 access with no betting limits, no time restrictions, lax age and identity verification, and no safeguards against addiction or insider trading.
Any unlicensed foreign website offering bets in Switzerland will be blocked, according to Gespa. The statement asked clubs, leagues, and sports federations to verify the legitimacy of prediction market platforms before signing commercial partnerships.
The World Cup losses are not occurring in a vacuum. Kalshi, the CFTC-regulated platform, has implemented employment disclosure requirements, risk scoring on high-volatility markets, and a whistleblower channel following its own insider trading investigation, which blocked over 100 potential insider trades in Q1 2026 alone.
Polymarket announced a compliance audit after the WSJ revealed its fake-win influencer campaign. The House Committee on Oversight and Government Reform launched a probe into both platforms in May.
None of those measures cap individual position sizes, require income or net worth verification before a participant can place a seven-figure bet, or introduce mandatory cooling-off periods after significant losses. Jordan Bender, a gaming equity research analyst at Citizens, stated that individual investors in prediction markets tend to incur losses in the long term and have lower returns compared to those who use traditional sportsbook operators.
The data from the World Cup’s first two weeks supports that assessment at every level of position size from retail to eight figures.
The prediction market industry is entering its most commercially significant phase just as evidence of widespread retail losses becomes harder to ignore. Trading volume has surged from about $50 billion in 2025 to more than $130 billion in 2026.
Meta is reportedly developing a competing product, Cboe has launched a regulated offering, and new entrants such as TruthPredict.ai continue to emerge. The sector now has the kind of growth that attracts capital, competition, and regulatory scrutiny in equal measure.
NEW: Meta is developing a standalone prediction-market style app, internally dubbed “Arena,” inspired by the rapid growth of Polymarket and Kalshi.
The app would initially use a points-based system rather than real-money wagering, though Meta has not ruled out betting features… pic.twitter.com/fpuHr4Czq4
— Frank Chaparro (@fintechfrank) June 23, 2026
The World Cup brings a broader question into focus: are prediction markets a valuable financial tool that improves price discovery and information aggregation, or are they, at scale, a system that transfers money from the many to the few under the language of market efficiency?
The University of Toronto’s finding that 68.8% of users lose money suggests both can be true at once. Prediction markets may offer genuine value for institutions and sophisticated traders while exposing retail participants, whose activity provides much of the liquidity, to a high probability of losses.
For FlickRaw, that debate is already settled. The account placed just three trades and reportedly lost $4.2 million. Whatever prediction markets ultimately become, they are clearly not a low-risk product.
Share
preferred on
