With Tadgh Hegarty, Journal of Prediction Markets, 2023.
Covid produce an interesting natural experiment for assessing the ability of betting markets to forecast. For a long stretch, soccer was played with no crowds. Everyone knows home advantage exists, and crowd support one of a few obvious candidate explanations. So, the question was simple: did betting markets adjust, and did they get it right?
We look at matches from the top four European leagues (England, Spain, Italy, Germany) from 2018/19 to 2020/21 and compare outcomes with what the Asian handicap market implied about expected goal differences.
The headline result is that markets adjusted quickly. They priced in a large drop in home advantage once games went behind closed doors—roughly halving the pre-Covid home goal advantage—and that adjustment turned out to be close to what the data eventually showed.
The subtle part is what happened early on. Some of the first “no crowd” matches, especially in Germany, looked like home advantage had collapsed far more dramatically. But markets didn’t wildly overreact to a handful of strange early games. As more matches were played, the data settled down, and it turned out the market’s initial judgement was basically right.
Key research finding: Betting markets processed the “no crowds” shock quickly and accurately, and they did not get misled by extreme early results.
Practical advice: Even when faced with an incredibly complex pricing question, with very little evidence to go on, bookmakers were able to figure out the right answer.