clock menu more-arrow no yes mobile

Filed under:

Stats Don’t Lie: Of Peach Baskets and Broken Bones

Time for some basketball math. No, wait, come back!

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

NCAA Basketball: Final Four-City Views Troy Taormina-USA TODAY Sports

Let’s talk arbitrary basketball statistics and broken bones. Bear with me here, and I’ll tell you which team will win the March Madness tourney this year. Probably.

On the last day of the 2021-2022 season, Illinois had a slim chance to take a share of the B1G title. They needed a bad Nebraska team to beat the Wisconsin Badgers, which had already clinched part of the conference crown. And if that unlikely event happened, Illinois still needed to take down their archrival Iowa at home. I was not optimistic, but it was a warm March day in northern Virginia. I threw on my decade-old Illinois cycling team jersey and rolled out for a bike ride.

Twenty miles later, a freak accident threw me off of my bike and damn near over the edge of a rocky embankment. I swore. It hurt. My leg was bleeding, my arm was in pain, and my bike was a mess. I’ve crashed many times before. This one was bad.

My wife picked me up and drove me to the ER. Broken collarbone confirmed, sling engaged, rest required. But the blood magic spell I had inadvertently cast was complete. Nebraska improbably took down Wisconsin, and while I self-medicated with a comically large pour of bourbon, Illinois held on to claim a share of the title. To all fellow traveler Illini fans, you’re welcome. I shattered bones for this team.

The next few weeks were impossibly frustrating and tedious. To keep my sanity, my best friend from my tiny Illinois hometown and I became basketball obsessives (shoutout to Dan, his 1-year-old Mattimeo, and his Villanova Wildcats, which might need some blood magic this year). We scoured articles and stats in the lead-in to the tourney. I eventually came across an Action Network article arguing the following: in the last 20 years, damn near every March Madness champ has had both a top-25 KenPom offensive efficiency rank and a combined KenPom offensive and defensive rank of 50 or fewer. I am a degenerate gambler, and I was intrigued.

We took the theory, collated better data, and played with the numbers. Ultimately, the formula was more predictive with this formulation: a March Madness champion must have a pre-Tournament top-20 KenPom-ranked offense AND a combined offense and defensive rank of 40 or less. Higher seeds are always more likely to win, but this primitive model removed one-dimensional teams from contention. To win a title, you need an elite offense first and foremost, but also secondarily need a damn good defense. Great defense alone won’t win it all, nor will phenomenal offenses with a sieve on the other end. Year to year, about 4-8 teams fall into that category, which we call the Peach Basket. If you’re in the Peach Basket before the tourney, you have a shot to win it all.

How did the formula do over the past two years? Out of five teams in the 2023 Peach Basket, Purdue shat the bed against Farleigh Dickinson, Houston and Alabama made the second weekend, and Texas made it to the Elite Eight. But more importantly, my celebratory Ennui Husky Homefield shirt and my gambling winnings confirm that UConn was a supremely undervalued squad last year, which the Peach Basket correctly identified.

In 2022, of the eight teams in the Peach Basket (Gonzaga, Kentucky, Villanova, Kansas, Arizona, Baylor, Houston, UCLA), seven made the Round of 32, six made the second weekend, three made the Elite Eight, two made the Final Four, and Kansas won the dang thing. With two years of live use of the Peach Basket, we were 2/2, albeit with a few hilarious flameouts (who can forget the St. Pete’s Peacocks taking down Kentucky?).

Since 2007, the Peach Basket model has picked the winner in 13 of 16 years. If a team is in the Basket, they have an 11.2% chance of winning it all and a 30% chance of making the Final 4. The numbers get marginally better if you remove Purdue from the data set. Just kidding, Boilerbots. It’s probably better if you just take out every Big Ten team.

As of February 3rd, nine teams fall into the Peach Basket, but this does tend to slim down by season’s end. The most underrated team is BYU, followed by Illinois and Arizona.

Is this lazy formula remotely helpful for gambling purposes? Probably not. The Peach Basket shows what we already knew: elite teams tend to win it all despite the perceived madness of the NCAA tourney. The Peach Basket won’t even help you win your bracket pool, given that Dan lost to my cat’s bracket last year. Make your bets accordingly.

But it’s worth remembering: Ball don’t lie. Neither do stats.


Click here to check out this week’s browser version of our brand-new weekly newsletter. And if you like it, subscribe here!