Creighton' Panic
1/22/23 - How college basketball's most confusing team has flummoxed humans and computers alike, proving us both wrong in the process.
Coming into the season, it was hard to find a more polarizing team than Creighton. The Bluejays represented the latest battleground in the ever-vicious debate over the utility of continuity in predicting college basketball success. The galvanized public saw Creighton’s young team push national champions Kansas to the limit in their Round of 32 game in the NCAA Tournament last season and assumed that the largely-unchanged roster this season would flourish to new heights with their newfound experience. Many preseason projections placed Creighton squarely in the title conversation, as my own rankings coming into the year saw the Bluejays place just outside the top ten, knocking on the door of college basketball’s elite club for the first time in decades.
Not everyone, or rather everything, was so convinced. Predictive algorithms and their followers dumped cold water on the sky-high expectations in Omaha, with the KenPom preseason rankings slotting the Bluejays at a respectable 22nd. A very good team by all means, but not the title contender envisioned by those who predicted a major leap from the young, talented squad that ranked 50th at previous season’s end.
Predicting these “leaps” from non-traditional powerhouses is notoriously unreliable, as illuminated by Three Man Weave’s Jim Root in his fantastic piece, Forecasting The Leap: How Good Are We At Predicting Team Jumps Into The Top 10? Per Root’s research, teams predicted to jump into the top 10 by preseason polls from outside the top 25 of the previous season are more likely to finish outside of the top 20 than inside the top 10. Put another way, if you weren’t that good last year, you probably won’t be that good this year. By this measure, Creighton’s achievement of the lofty expectations placed upon them this season would be a rather unlikely accomplishment.
There’s a host of reasons for why teams like Creighton receive so many plaudits in the preseason. They’re a non-traditional powerhouse, they “bring everybody back,” and they have a fun style of play that produces sensational highlights. Creighton, a relative newcomer to the historic Big East Conference, has never made the Final Four, while the Elite Eight has evaded the Bluejays since the outbreak of World War II. Everybody wants a big headline for their rankings drop, and the Creighton Bluejays are the perfect fodder.
For the naysayers and microchips, it came as no surprise that the Jays didn’t quite match their hype once the balls were rolled out. A strong showing in the Maui Invitational was followed by an unfathomable 6-game losing streak that reached its zenith with an appalling double digit home loss to perennial cellar-dwellers and local rival Nebraska. Creighton plummeted out of the top 25, and the national media that had once touted the arrival of college basketball’s new rockstars swiftly discarded their prior plaything in favor of Big East upstarts UConn and Xavier, as the destitute Bluejays currently sit with a ho-hum record of 11-8.
So the computers were right? Creighton didn’t make the leap? The team that was largely the same players as last season…have played largely the same as they did last season?
Here’s where it gets complicated. The computers love Creighton. In fact, according to both KenPom and BartTorvik, the Bluejays have actually overperformed this season. Coming into the year, KenPom had Creighton projected as the #22 team in the country. The current rankings slot Creighton in at #13. BartTorvik rankings projected Creighton at 23rd while the current T-Rank for Creighton stands at 17th. In a stunning role reversal, the algorithms that once decried the rise of the Bluejays have now become their greatest bulwark against a reactionary public quick to abandon their former infatuations.
How is this possible? How can Creighton be 11-8, 4th in the Big East, and yet squarely top 25 in analytical metrics? Computers simply don’t look at outcomes in the same way humans comprehend them. For metrics, Creighton’s nail-biting losses against Arizona and at Texas weren’t disappointments; these were affirmations that Creighton could play at the level of college basketball’s best. Two of the other 4 losses in their infamous losing streak were also by just a combined 5 points, lessening the blow to their analytical perception. Wins in Maui against Arkansas and Texas Tech, two teams the analytics rank higher than humans, more than make up for the rest of Creighton’s early season skid. The rest of the year, Creighton has followed the formula for analytical dominance: win big, lose little. Five of Creighton’s seven wins against major conference opposition have been by double digits, while another close loss against good competition, this time on the road at Xavier, avoided harming the Jays to any major extent. The sky hasn’t fallen in Omaha; according to the metrics, Creighton may have never even left the troposphere.
There’s one more wrinkle that makes this all the more interesting. It’s something that metrics like KenPom don’t account for: injuries. Creighton center Ryan Kalkbrenner looked like he had the worst hangover of his life during Creighton’s tournament finale in Maui, burying his face in a cold towel at every stoppage. It turns out Mr. Kalkbrenner must’ve been enjoying himself on the island, as he had come down with a nasty case of mononucleosis that would linger for months, knocking Kalkbrenner out of Creighton’s losing skid entirely. It’s possible that Creighton, with a healthy Kalkbrenner, could’ve flipped many of those close losses into close wins. While this would’ve moved Creighton slightly higher in the analytical rankings, it would’ve had a marked effect on public perception of Creighton, and the computers would once again be left as the sobering pessimists, perceiving close margins of victory as indication that Creighton had not transcended their opposition.
It's a bizarre chess game of media groupthink, human logic, and ones and zeroes. We were right that Creighton was good, but wrong that they are bad, while the computers were wrong that Creighton was bad, but most likely right that Creighton is good.
I’m talking myself into a pretzel.
Over half of the Big East schedule remains; the Bluejays currently sits 2.5 games back of league leaders Xavier. If Creighton is the team the media thought they were, and the team the computers think they are, they have plenty of time to prove it to both. Yet, no matter what they do at this point, the topsy-turvsy season in Omaha may only prove one thing:
We’re all just throwing darts in the dark.