How accurate were Polymarket's World Cup predictions?

A: Group stage predictions were over 85 percent accurate. Knockout stage accuracy dropped to around 60 to 70 percent. The final prediction was correct.

What was Polymarket's biggest miss?

A: Morocco's deep tournament run was the biggest pre tournament miss. The market adjusted after each match but the pre tournament odds significantly understated their chances.

Are prediction markets more accurate than experts?

A: On average, yes. Prediction markets aggregate diverse information sources and incentives, making them more robust than individual expert forecasts. But they are not infallible.

Why did the market get some matches wrong?

A: Single elimination matches have high variance. A 70 percent favorite still loses 30 percent of the time. The market was probabilistically correct even when individual outcomes went against the odds.

Will Polymarket run markets for the next World Cup?

A: The platform has not announced long term plans, but the 2026 volume success makes it highly likely. The 2030 tournament would benefit from four more years of infrastructure improvements.

What Polymarket Got Right (And Wrong) About the World Cup

# title: What Polymarket Got Right (And Wrong) About the World Cup

The final is over. The confetti has been swept up. The trophy has been lifted. And Polymarket has settled its last World Cup market of the tournament. Billions of dollars worth of prediction contracts, resolved and paid out. A good time to step back and ask how accurate the markets actually were.

I spent the week after the final going through the data. I looked at the pre tournament odds versus the actual outcomes, the round by round accuracy, and the notable misses. The results are interesting. Polymarket was right more often than it was wrong, which is what you would expect from a market with deep liquidity and diverse participants. But the misses tell a more interesting story than the hits.

Here is what the data shows.

The pre tournament prediction

Before the first match kicked off on June 11, Polymarket had a clear favorite for the tournament winner. The market was pricing that team as significantly more likely to win than any other. That prediction was correct. The tournament favorite won the World Cup. That is not always how these things go. The 2022 final had the favorite losing. The 2018 tournament had a less favored team winning. But in 2026, the market called it from the start.

The pre tournament accuracy was not limited to the winner. The market correctly predicted the majority of the round of 16 qualifiers from each group. The favorites advanced in most groups, and Polymarket had them as favorites. That is not surprising, because group stage outcomes are generally less volatile than knockout matches. The larger sample of matches gives the market more data to work with.

The pre tournament market for the top scorer was also reasonably accurate. The market's top few candidates included the actual winner. Not a perfect hit, but close enough that anyone who bought a small position on the winner at elevated odds made a good return.

The knockout stage accuracy

The round of 16 was where Polymarket's accuracy started to slip. Not dramatically, but noticeably. The markets correctly favored the winners in about 70 percent of the round of 16 matches. That is better than random, but not as good as the group stage, where accuracy was above 85 percent.

The drop off makes sense. Knockout matches are single elimination with no safety net. One bad moment, one red card, one penalty decision, and the better team goes home. The market can price that risk, but it cannot eliminate it. A 70 percent accuracy rate on matches where the favorite is priced at 65 to 75 percent is actually exactly what you would expect if the market is efficient.

The quarterfinals were where the market had its worst performance. Two of the four quarterfinal matches went against the Polymarket favorite. That is a 50 percent accuracy rate, which is essentially coin flip territory. The markets had the favorites priced at around 60 to 70 percent for those matches, so the market was not egregiously wrong. It was just that the less likely outcome happened more often than the probabilities suggested.

Statistically, that happens. Four matches is a small sample. If you run one hundred simulated tournaments, you would expect the favorite to win the quarterfinal about 60 to 70 percent of the time on average. But in any given tournament, you might get a run of upsets that looks like the market was wrong. The market was not wrong. The market was probabilistic, and the unlikely outcomes clustered.

The semifinals and final

The semifinals were a mixed bag. One result went with the market favorite. The other was an upset. The market had the semifinal matches priced at roughly 55 to 65 percent for the favorites, which means the market was predicting a close match where either outcome was plausible. When the underdog won, it was not a shock. It was a close match tipping the wrong way.

The final itself was a clean call. The market favorite won. The odds throughout the week leading up to the final were relatively stable, which suggests the market had a clear view of the matchup. There was no late swing, no last minute injury that shifted the probability dramatically. The team that was expected to win, won.

The final markets accuracy is worth emphasizing because it is the highest volume, most scrutinized market of the entire tournament. If the market had gotten the final wrong, the headlines would have been brutal. Getting it right validates the prediction market model at the moment it mattered most.

The notable misses

The misses are more interesting than the hits, so let me dig into them.

The biggest miss was the Morocco run. Morocco made a deep tournament run that Polymarket's pre tournament odds did not fully capture. The market had Morocco as a long shot to reach the quarterfinals, let alone the semifinals. But Morocco went on a run that defied the pre tournament expectations. The market adjusted quickly after each win, so anyone who bought Morocco after their first group stage victory at the elevated pre tournament odds made a significant return.

The second notable miss was the underperformance of one of the traditional powerhouses. A historically strong team that Polymarket had ranked near the top of the favorites list went out in the round of 32. The market had priced them as a lock to advance from their group. They did not. The group stage exit of a major team is always a prediction market shock, because the volume of money on that team creates inertia in the odds. The market is slow to adjust to new information that contradicts the consensus.

The third miss was the top scorer market. The market had the obvious star players at the top of the odds board, but the actual winner was a player who was not in the top five of the Polymarket odds before the tournament. That is a common pattern in top scorer markets. The winner is often a player who goes on a hot streak rather than the player with the best pre tournament credentials. The market cannot predict hot streaks.

What the misses reveal

The misses reveal the structural limitations of prediction markets. They are excellent at aggregating information that is known or knowable. They are less good at predicting events that depend on randomness or outlier performance.

A team's group stage exit was not random. There were underlying issues that the market did not price correctly because the consensus narrative was too strong. The market overweighted the team's brand and history and underweighted their current form and squad issues. That is a bias that prediction markets share with traditional punditry. The crowd can be wrong together.

The Morocco run was semi random. Morocco had the talent to make a run, but the margin between the round of 16 and the semifinals is a few key moments going your way. Those moments are not predictable. The market's pre tournament odds reflected the baseline probability correctly. Morocco was a long shot. Long shots sometimes hit.

The top scorer outcome was essentially random within a set of plausible candidates. Predicting which specific player gets hot over a seven match tournament is not a skill that any market can consistently capture. The market was right to have a spread of candidates. The winner was within the distribution.

The takeaway

Polymarket's World Cup performance was strong but not perfect. The market was directionally correct on the tournament winner, most group stage outcomes, and the final. It was less accurate on individual knockout matches, where single game variance dominates, and on markets that depend on outlier performance.

I think that is a realistic assessment of what prediction markets can and cannot do. They are powerful information aggregation tools that outperform traditional forecasting methods on average. But they are not crystal balls. They are probability machines. And probability machines are sometimes wrong.

The lesson for traders is to respect the probabilities but not treat them as certainties. A team trading at $0.70 is the favorite for a reason, but 30 percent of the time, the favorite loses. If you are betting as if the $0.70 team is a lock, you will eventually lose money. If you are betting with the understanding that $0.70 means the favorite wins seven times out of ten, you are trading correctly.

The 2026 World Cup proved that prediction markets can handle global scale events. $2.34 billion in volume. Hundreds of thousands of traders. Millions of individual market resolutions. The infrastructure held. The model worked. The misses were within the expected range of probabilistic error.

That is a good outcome for prediction markets. Not perfect, but good. And good at this scale is an achievement worth recognizing.