How prediction markets mirrored the chaos of the college football coaching carousel
SOURCE:The Athletic|BY:Dan Santaromita
Prediction markets reflected the drama, and rumor-mongering, of the college football coaching carousel.
In November, Missouri football coach Eli Drinkwitz’s name popped to the top of a list. The list? A ranking on the prediction market Kalshi for who would be Penn State’s next permanent head coach after James Franklin’s firing. On Nov. 19, Drinkwitz peaked with a 43 percent likelihood of getting the job on that site, making him more than twice as likely as any other coach in the market.
The catch? Drinkwitz, according to him, never even interviewed for the role.
“We got Twitter trending with bets on who’s gonna be leading or get this job,” Drinkwitz said after Mizzou beat Arkansas in its regular-season finale on Nov. 29. “I was the leading bet-getter for a job I never interviewed for. That’s annoying. That’s bullcrap. It’s just speculation.”
Drinkwitz had signed an extension with Mizzou a few days prior to his comments. On Dec. 5, Matt Campbell left Iowa State to become Penn State’s new head coach.
This example is one of many from this year’s very active college football coaching carousel. The fact that people can now legally put money on the candidates, thanks to prediction markets like Kalshi and Polymarket, has made the carousel even more carnivalesque.
In essence, you can now bet on whether or not you believe a message board rumor.
Prediction markets are exchanges, like stock markets, where people can risk money on the outcomes of future events. The fluctuations in price can be shown as percentages that demonstrate the prevalent belief of the market in the likelihood of a specific outcome, in this case, a coaching hire. The stronger the belief, the higher the price to purchase the stock. But as the Penn State example demonstrates, that belief can often be misplaced.
Traditional sportsbooks typically avoid taking on the risk of offering markets that can be easily manipulated by insider information, such as a coaching position, where hiring managers and school insiders could have information on leading candidates before the sportsbook.
The Lane Kiffin drama is behind us, Penn State ended its long hiring process and now Michigan has a full-time replacement for Sherrone Moore in Kyle Whittingham. With the biggest jobs settled, how right or wrong were the prediction markets, and what happened along the way?
While notable sports betting companies like DraftKings, FanDuel and Fanatics launched prediction markets in December, this article uses numbers from Kalshi because it had next coach markets live and was active in the U.S. for the entire college football season.
The Penn State market was all over the place
While Kiffin’s will-he-stay-or-go drama at Ole Miss garnered plenty of attention, the Penn State hiring saga had more plot twists and candidates connected to the job. The prediction market for the Nittany Lions’ position, which opened after the school fired Franklin on Oct. 12, was a reflection of what the rumor mill was producing.
Campbell ended up getting the job on Dec. 5, but he wasn’t the favorite until the very end. However, he was priced at a 29 percent likelihood as early as Oct. 21, reflecting the fact that Campbell was viewed as a logical candidate early on, even if other candidates were mentioned for the job more frequently.
Nebraska Coach Matt Rhule was the early favorite. He played at Penn State and has improved the Cornhuskers since he took the job ahead of the 2023 season. Rhule was viewed as the favorite for more than two weeks and even peaked above 50 percent, but signed an extension with Nebraska on Oct. 30, ending any speculation he would head to Penn State and thus seeing his price plummet.
Indiana’s Curt Cignetti was also mentioned for the job and was priced as high as 26.5 percent, but ended that by signing an extension at IU before Rhule signed his.
It was at this point that Drinkwitz rose to the top.
Once Drinkwitz was also off the board, James Madison coach Bob Chesney became the favorite. Chesney ended up taking the UCLA job.
Things got more interesting when reports came out that BYU coach Kalani Sitake was the focus of Penn State’s search. Sitake was off the board until these reports started coming out, but naturally became the favorite. However, BYU was able to keep Sitake and give him a new deal.
Finally, Campbell got the job and ended the madness.
Was the market correct to give Campbell a reasonable chance so early? Was it just following logical coaching candidates and knocking them off one at a time?
In all, there were “Yes” or “No” markets for 42 coaches during Penn State's search. Many of the names made some sense, while others were way out there. The likes of Urban Meyer, Jon Gruden, Nick Saban and Mike Tomlin were offered as possible hires.
Sometimes, information moves a market, like a report on a candidate interviewing for a job or being the focus of a school’s search. Sometimes, people are just going through a list of successful coaches who could make sense for a different job after other logical candidates have decided to stay at their current spots.
Kiffin mania also reached prediction markets
Kiffin mania didn’t get quite as crazy in the prediction markets as Penn State’s meandering search, but it still got plenty of attention.
In an amusing reflection of how the Kiffin saga went down, he was simultaneously the favorite to be the next Florida coach and the next LSU coach. Kiffin was the highest-priced candidate for the LSU job throughout, other than a random Saban spike, and was better than 71 percent to be Florida’s next head coach on Nov. 15. He wasn’t priced that high on the LSU market until Nov. 29, the day before he took the job.
Florida ended up hiring Jon Sumrall from Tulane. He was off the board for the first few days after Billy Napier was fired and was less than 10 percent until a few days before he was hired.
Sumrall was actually the favorite to take the Auburn job for most of the time it was open. The Tigers instead hired Alex Golesh, who was often second behind Sumrall in the market.
How close was Kalen DeBoer to landing at Michigan?
Michigan's coaching job opened up after most of the rest of the vacancies in the sport had been filled, which meant the candidate pool was relatively limited, even for a prestigious job like Michigan. Alabama's Kalen DeBoer opened as the favorite despite already sitting in a high-profile job and being in the College Football Playoff.
The prevailing wisdom, whether or not it was connected to reality, was that if Alabama lost in the first round to Oklahoma, DeBoer could be on the hot seat at Alabama and may want to jump ship for another big job while he had the chance.
Fast forward to the Crimson Tide's first-round CFP game at Oklahoma, and the roller-coaster ride of that game was reflected in the Michigan next coach market. DeBoer was priced around 25 percent heading into the game. Then, his Tide fell behind 17-0. Suddenly, it looked like Alabama would lose, putting in play the scenario where DeBoer could leave. His price reached 39 percent.
After Alabama came back and won, DeBoer was below 10 percent. All this within the space of a couple of hours.
We'll never know whether or not DeBoer would've gone to Michigan had Alabama lost at Oklahoma. That didn't stop the markets from believing that could be the case.
With Alabama's win and Kenny Dillingham's extension at Arizona State, Kyle Whittingham became the favorite on Dec. 21, five days before reports emerged that he would take the job.
What drives these predictions? And how easily are they manipulated?
Some coaching hires were more straightforward. Franklin landed at Virginia Tech in mid-November and was priced above 50 percent in October. Similarly, Ryan Silverfield was the favorite for the Arkansas job a month before taking it.
Meanwhile, Chesney was off the board to be UCLA's next coach until a week before he took that job.
Some markets were way off, but some seemed to be on the scent. Isaac Rose-Berman, a fellow at the American Institute for Boys and Men focused on gambling research and policy, said these markets are more predictive when there is more money in them, and most of these coaching markets had lower volume.
Kalshi claimed the Michigan coaching market got nearly $17 million in volume. The Penn State coaching market had over $6.6 million in it, but others had far less.
“You really have to make this distinction between markets that are highly liquid and markets that are not liquid,” Rose-Berman said. “Not to defend prediction markets too much, but prediction markets work when there's a lot of liquidity. So it's not fair to attack them for not being truth-seeking mechanisms when I, with one click of the button and not even caring about the money, could wipe out the entire order book.”
When there isn’t much money in a market, one significant bet could drastically shake up the prices, regardless of any connection to real information. They are also susceptible to false rumors manipulating the market.
“I can tweet out something that's going to move a market, and then, when people are going to buy in, because it's a prediction market, I can sell out,” Rose-Berman said. “I can buy a share at 10 cents and then drum up a bunch of hype about this guy, drive him to 20 cents and then just settle my position.”
It's easy to see the appeal and dangers of prediction markets in this context. Coach searches, free agency and trade rumors, all things you can bet on in prediction markets, all drive interest and clicks. Fans love thinking about what players and coaches their teams can add and want to see how likely certain scenarios are.
Prediction markets attempt to quantify internet rumors. However, it's difficult to know if it's actual insider information moving a market or pure speculation, as Drinkwitz called it.
Prediction markets are federally regulated and thus legal in places where sports betting is not, and it appears they aren't going anywhere in the short term. Companies are entering the space in growing numbers, and the current U.S. presidential administration has been looser in its regulation compared to the previous one. Kalshi even applied to list markets based on whether or not players would enter the transfer portal, although no such markets have been listed.
Being able to bet on coaching hires, and the wild rumors and swings that go with them, appears to be here to stay.