I’ve always been super interested in sports analytics as a student at Mizzou studying Data Science, and I had a little bit of free time on my hands. Thought I would start the journey of making my own metric to rank college basketball teams:
Bama 92.9%
Arkansas 78.1
UK 70.9
Tenn 70.3
A&M 68.2
Auburn 67.8
UF 51.1
Miss St 47.3
Vandy 45.7
MIZ 38.9
Ole Miss 31.4
UGA 24.0
LSU 14.4
SCAR 11.9
The values above are the predicted probability of an SEC win. The model only considers 2023 data from conference games and 5 statistics (in order of importance based on their relationship with SEC win probability):
Field goal percentage
Total rebounds
Turnovers
Steals
Assists
Obviously there are MANY other variables to consider, but this is a work in progress and model version 1.0. I plan to make it a lot more robust. But, I am in favor of model transparency and being able to easily comprehend what goes into a ranking. This is not, and I must stress, resume based.
Any suggestions on improvements or metrics to include are greatly appreciated!
Bama 92.9%
Arkansas 78.1
UK 70.9
Tenn 70.3
A&M 68.2
Auburn 67.8
UF 51.1
Miss St 47.3
Vandy 45.7
MIZ 38.9
Ole Miss 31.4
UGA 24.0
LSU 14.4
SCAR 11.9
The values above are the predicted probability of an SEC win. The model only considers 2023 data from conference games and 5 statistics (in order of importance based on their relationship with SEC win probability):
Field goal percentage
Total rebounds
Turnovers
Steals
Assists
Obviously there are MANY other variables to consider, but this is a work in progress and model version 1.0. I plan to make it a lot more robust. But, I am in favor of model transparency and being able to easily comprehend what goes into a ranking. This is not, and I must stress, resume based.
Any suggestions on improvements or metrics to include are greatly appreciated!
