Sports betting and financial market data shows how people predictably misinterpret new information

Let’s say it’s a Golden State Warriors home game and Stephen Curry shows he’s still got what he’s got by sinking back-to-back 3-pointers in the first few minutes of the first quarter. Chase Center fans are taking note, and so are the betting markets, where the odds are moving in the Warriors’ favor.

It’s a long game though. The visitors fought back, and with 10 seconds left, the Warriors were up by two points and just missed one more shot. The odds of winning are low and the betting odds should change to reflect that near certainty. But that’s not the case.

“If you look at the history of NBA games, the team that has the ball with a two-point lead with 10 seconds left wins more than 90% of the time,” said Eben Lazarus, an assistant professor of finance at the Haas School at the University of California, Berkeley. . of business. “But what is emerging in the betting market is that people treat baskets as too similar during a game. They overreact to less important information, early baskets, and They underreact to strong signals.”

According to a paper published in the Quarterly Journal of Economics, interesting patterns in how people interpret new information apply to a variety of environments, from sports betting to financial markets.

Lazarus and co-authors Ned Augenblick of the University of California, Berkeley-Haas, and Michael Thaler of University College London conducted three experiments, analyzing the prices of millions of bet trades and option contracts. They found that people consistently overreact to weak information and underreact to strong information.

“There are all kinds of situations where you might know whether the news is good or bad, but have a hard time determining exactly how important it is,” Lazarus says. “We saw this pattern everywhere, but this was surprising to us given the stakes involved in gambling and financial markets.”

Built on decades of behavioral science and economics research

Lazarus and his co-authors wanted a way to unify different theories about how people act irrationally when processing new information.

The study builds on decades of behavioral psychology and economics research into how people update their beliefs given new information, and found that people are overly cautious about updating. It goes back to the classic paper in 1966 that argued that people tend to be, and the paper by Dale Griffin and Amos Tversky in 1992 that showed that people tend to They focus too much on information that seems dramatic, but not on how reliable it is.

Recent papers have shown that people make systematic errors as a result of miscalculating probabilities, and that when in doubt about which decision to make, they tend to choose an intermediate option. The paper also relates to research on how financial markets can sometimes overreact and sometimes underreact to news.

“We believe we have a simple framework for thinking systematically about financial markets and various real-world situations,” he says.

As humans, we are constantly taking in information, whether it’s a new poll supporting our preferred candidate or feedback from our boss. The researchers theorized that most of the time, we don’t have the information to judge exactly how important that information is, so we tend to default to the middle ground.

“When it’s easy to know in which direction to update your beliefs, but not quite sure how much to update, people tend to treat all ‘good’ information somewhat the same,” he says. “Given this difficulty, some people will overreact to fairly weak news and underreact to news that should approach certainty.”

experimental evidence

The research team first tested the theory with laboratory experiments. These included a classic experiment to determine which deck a particular card came from, and a novel sports-related experiment that gathered 500 NBA fans and showed them a sequence of events in a mock basketball game. Both are included. .

The simulation began with 2 minutes and 40 seconds left in each quarter, and participants then watched four consecutive possessions. After each possession, participants had to predict the probability that each team would win (winning a $50 bonus for accuracy).

The researchers used data from the website inpredictable.com to establish the “correct” probability of winning in each scenario. But even though people understand that late-game baskets are more important than baskets scored early in the game, they still overreact to first-quarter baskets, making them 60% more important than they should be. The researchers found that players underestimated fourth-quarter baskets by 33 points. %.

“This gives us a good idea in experiments that people overreact or underreact to information, but we need to come up with some ways to test this in more high-stakes situations in the real world. “I did,” Lazarus said.

Sports betting data from Betfair

To do so, the researchers focused on sports prediction market Betfair, analyzing more than 5 million betting transactions across 260,000 basketball, soccer, football, and ice hockey games. Researchers developed a new empirical method to measure whether prices are over- or under-reacting to information because there was no way to reliably determine the “correct” probability of winning. .

Again, we find that events such as scores early in a match cause more fluctuations in betting odds than they should, given the high uncertainty about the outcome. On the other hand, important events such as the Q4 targets caused smaller changes in the market than warranted.

Data from options market

Lazarus and his coauthors tested their theory in sophisticated financial markets using option price quotes for S&P index options traded on the Chicago Options Exchange from 1996 to 2018. After applying multiple filters, they obtained over 4 million option prices corresponding to: The expiration date is 955 days. To provide a clear time horizon, they focused on stocks that expire within 100 business days (approximately 4.5 months).

They found the same pattern they observed in the sports betting market.

“…there appears to be relatively little information in today’s news about the value of the S&P over several months, but the market is behaving as if it were,” the authors write. “However, within two weeks of contract resolution, the relationship reverses…As the signal becomes stronger, the market begins to underreact.”

Real world impact

The study describes some puzzling patterns in how people and markets react to news, but Lazarus says being aware of these patterns eliminates all risks. I warn you that this is not the case.

“Even if we know that the market is, on average, underreacting or overreacting at various points, it’s still not a free lunch,” he says. “If you bet against these movements, you can still lose a lot of money.”

Still, the findings suggest that it is wise to be careful about how much weight we give to different pieces of information, even in much more ambiguous situations.

“Let’s say you had a negative interaction with your boss and it’s been bothering you all week,” he says. “How important is it really to my future with this company? I think it’s better to take a step back and think about how responsive I should be.”