Original title: The Favorite-Longshot Bias in Prediction Markets: Evidence from Polymarket
Marcos Cardozo, José Ignacio Rivero-Wildemauwe · September 2026
People who bet often pay too much for risky, unlikely bets and avoid safer choices. This pattern is known as the favorite-longshot bias. Researchers wanted to see if this happens on Polymarket, a major prediction platform where people trade on future events. To find out, they analyzed millions of trades across the platform. They found that buying cheap, low-probability bets usually loses money, while buying expensive, highly likely favorites makes a small profit. However, this bias looks different depending on how you group related bets together. The bias shows up strongly in categories like politics and crypto, but it is surprisingly missing in sports bets. Even though some people regularly keep buying risky longshots, they perform about the same as everyone else.
Imagine you are betting on horse races. You notice that betting on the longshot horses (the ones with low odds) almost always loses you money, while betting on the favorites (the high odds horses) makes you a small profit. But then you realize that if you group all the horses from the same race together, the longshot bets actually become profitable. In this paper, the contracts are like the horses, and grouping by the race (the parent event) changes whether the longshot bets lose or gain money.
Prediction markets like Polymarket are used to forecast events, from elections to sports. If longshots consistently lose money, traders can adjust strategies. But the fact that grouping changes the result means the bias is not simple—it depends on how you look at the data, which matters for anyone using these markets to gauge probabilities.
The results are specific to Polymarket and may not generalize to other prediction markets. The bias is absent in Sports, and its magnitude depends on how contracts are grouped. The study does not explain why the bias exists or why it varies by category.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
The Favorite-Longshot Bias in Prediction Markets: Evidence from Polymarket — Marcos Cardozo, José Ignacio Rivero-Wildemauwe
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.