Football is done, it is nearing the middle of February, and Selection Sunday is just less that five weeks away. While the godfather of Bracketology, Joe Lunardi, has been deep into forecasting the field since the final buzzer sounded last year, now is the time forecasting the field starts to get serious.
Based on my Elo model, which is updated through this past weekend’s games, here is my projected field.
Bubble Teams
Every year, there are a dozen or so teams biting their nails on Selection Sunday, waiting to see if they will be among the last in or first out. Here is the bubble picture as things currently stand.
John' Calipari’s Arkansas team started out rough, but after a recent win over then 12th-ranked Kentucky and tight loss to #3 Alabama, the Razorbacks find themselves (barely) in the field. SMU clings to the final spot in hopes of giving the pitiful ACC — which now has 18 members — just a 6th team in the dance. The Big East, which is also having a bit of a down year, hopes to see Xavier and ‘Nova make a push in the final month of the season and sneak into the tournament alongside St. John’s, UConn, Marquette, and Creighton.
Conference Summary
Not enough can be said about how ridiculous the SEC is this year, and these projections back that up. The SEC projects for 13 of its 16 teams to make the tournament, with just Vanderbilt (#52 in Elo), South Carolina (#84), and LSU (#89) missing out.
How Far Will They Go?
Using a linear regression model to predict how many rounds each team will go, with forecasted seed and various KenPom data as the inputs, then scaling the results to reflect their probabilities of reaching each round, I was able to give each team a probability of reaching the Elite 8, Final Four, National Championship, and winning it all.1
Auburn and Duke look to be in a tight battle to cut down the nets in April. However, A LOT can happen between now and even just Selection Sunday. So this model will be updated regularly over the next month and beyond.
Stay tuned!
Keep up with the College Basketball rankings and game predictions here:
Data accessed using the hoopR package in R: https://hoopr.sportsdataverse.org/