Did prediction markets call the U.S. midterm election?: State of Crypto

Prediction markets have garnered attention for at least the past six years. Are they accurate? Do they provide useful information on election results above and beyond what polls tell us? I'm not sure. If you have thoughts, I'd love to hear them — email me.

You’re reading State of Crypto, a CoinDesk newsletter looking at the intersection of cryptocurrency and government. Click here to sign up for future editions.

Confidence over predictions?

The narrative

The "blue wave" election was sweeping. In the wake of it, there's been a debate on X (formerly Twitter) whether prediction markets like Kalshi and Polymarket were better or more accurate than polling models. It's a question that's come up since at least 2020, when this newsletter's predecessor tracked what polls and early prediction market platforms were projecting for that year's presidential contest. Four election cycles later, it's fair to say prediction markets have, at the least, carved out a niche for themselves.

Whether that niche is tailored toward degenerate gamblers or providing legitimate market insight is maybe still a question.

Why it matters

There's a chicken-or-egg question in play here. Did prediction markets follow polling sentiment or did they — to use the term their name implies — predict where polling sentiment may lead?

I've been thinking about prediction markets less as an indicator of polling sentiment itself and more as an indicator of confidence in a given outcome. That doesn't really answer the above question, but I've found it a useful frame for my own reporting purposes. And to be clear, I've been following prediction markets for quite a while. I wouldn't say they've replaced polling data or other tools as part of my personal workflow.

So how'd they do this election?

Breaking it down

Prediction markets and polls generally matched across most of the key races that CoinDesk tracked and also generally matched the eventual outcome, with a few exceptions.

In Wisconsin's First House District, Rep. Bryan Steil lost to Democrat Mitchell Berman. Both polls and prediction markets had that race in a dead heat the Friday before.

In Alabama's Second House District, Rep. Shomari Figures defeated challenger Rhett Marques. Both polls and prediction markets showed Marques as leading against Figures in the redrawn district.

In the Iowa Senate race, State Rep. Josh Turek defeated Rep. Ashley Hinson, flipping the seat; prediction markets had given Hinson the edge, while polls, when averaged, had a 50/50 response.

Outside of these three, both prediction markets and polling projected the eventual winner in the main races CoinDesk was watching.

Other races didn't show up much in either polling or prediction markets.

The Polymarket contract for the Mississippi Senate race favored incumbent Republican Cindy Hyde-Smith until vote results started coming in on Tuesday night. Kalshi did not have a contract for that contest.

There was even less information available on South Carolina's Senate race. Neither Polymarket nor Kalshi had a contract on the outcome, though Polymarket did have a parlay for the winners of the Montana, South Carolina and Mississippi Senate races. Spokespeople for both companies did not respond to questions about these races.

Both those races became upsets as Democrats Scott Colom and Annie Andrews, respectively, won their seats.

It's not as though polling data was more helpful. According to The New York Times' archive of polls, the most recent South Carolina race had Senator Graham up by 1. There were no polls available on the Times page for Mississippi, though other sites cited polls showing Senator Hyde-Smith leading narrowly.

I'm leaving the remainder of this newsletter as an exercise to you, dear reader: How should I be thinking about prediction markets, especially with the context of seemingly limited intelligence (polling or otherwise) on these races that appear to have been unexpected flips? I want to hear from you!

This week

Thursday

• 15:00 UTC (10:00 a.m. ET) The Senate Banking Committee is holding a nomination hearing on a number of individuals, including President Donald Trump's nominees to the SEC.

If you’ve got thoughts or questions on what I should discuss next week or any other feedback you’d like to share, feel free to email me at nik@coindesk.com or find me on Bluesky @nikhileshde.bsky.social.

You can also join the group conversation on Telegram.

See ya’ll next week!