The midterm elections this month didn’t just deliver a decisive swing in political power in the U.S. House and Senate. They might also represent a fundamental shift in how elections are thought about and analyzed.
Prediction markets like Polymarket and Kalshi, where people bet on elections just as they might bet on a basketball game, use math to turn betting data into predictions of real-world outcomes.
For example, on Nov. 1, Polymarket used its data to predict that Democrats had a higher than 70 percent chance of taking control of the U.S. Senate. Now, according to the article “How Accurate Were Prediction Markets at Picking Midterm Winners?,” whether prediction markets can accurately anticipate election results has become a key question:
An analysis by The New York Times, using statistical methods, found that for most of the midterm race, prediction markets performed no better than election forecasts, which produce odds using polling and other indicators instead of betting activity. But as Election Day neared, prediction markets performed better as forecasts and polling became stale.
The nature of prediction and the role of prediction markets
Ask your students these questions:
First, take a look at the graphs at the top of the article about the predictive accuracy of Kalshi and Polymarket in the midterms, clicking through the four categories (House & Senate, Senate, House, Governor). Note: The green lines represent four different election forecasts, which produce odds using polling and other indicators instead of betting activity. What do the graphs show? What conclusions can you draw?
According to the article, Polymarket assigned Senator Darline Graham a 70 percent chance winning her election in South Carolina. What does it mean to say that someone has a 70 percent chance to win an election? How is this kind of statement different from a statement like “When a fair six-sided die is rolled, there is a roughly a 33 percent chance that a 5 or a 6 will appear”?
In fact, Ms. Graham lost her election to the Democratic challenger Annie Andrews by around 90,000 votes. Given the outcome, how do we think about the prediction that she had a 70 percent chance to win? Does it mean that the prediction was incorrect? What exactly constitutes evidence for, or against, a claim like this?
The prediction markets publicly celebrated their successful predictions. But Nate Silver, who himself found fame in 2008 for accurately projecting the election of Barack Obama, was less impressed:
Mr. Silver, the forecaster behind Silver Bulletin, said the accuracy of prediction markets was a direct result of work by pollsters and forecasters, who create the information that bettors have mostly relied on.
“In a world where these models didn’t exist, or if the models kept their forecast private, I don’t think the prediction markets would be as accurate,” he said in an emailed statement.
What does Mr. Silver mean by this?
If you’d like to go further:
Online prediction sites allow you to wager money on the existence of aliens, the return of Jesus and even the chances of nuclear Armageddon. We asked students this spring: Has betting gotten out of hand? To address some of the ethical questions around this topic, invite your students to read the piece and post their thoughts to our forum.
How do more traditional methods of election analysis compare to relying on betting markets? Students can see how pollsters and pundits did with their midterm predictions in July and September, and check them against the November results. Did they fare better or worse than the prediction markets? And were late polls suggesting big Democrat gains reliable, or were they overstating leads?
And when it comes to polling, students can learn more about margin-of-error calculations in this Upshot piece. They can also read about everything from data sources to how a winner is declared to concerns about how news reporting on polls might affect the polls themselves in this Times Insider response to reader questions about election coverage.
To learn even more about prediction markets, check out this article about a Polymarket “sharp” who bet over a million dollars on the midterms; this piece, “How Everything Became Gambling,” to delve into cultural issues; this one about the ethical and legal issues accompanying the expansion of sports gambling; and, finally, this piece about the possibility of corruption in prediction markets.
Our 15-minute lesson plans are new this year. Find more lessons here, or send in your idea.
