Just playing with some numbers, I found that on-roster 3-star count was pretty strongly correlated with regular-season wins for SEC teams in 2019. I’ve done this in the past, and found that it is a pretty consistent relationship.

The correlation is 84%, R-squared is 70%, p-value is <.001, and SE is 1.76. Definitely something to this. The relationship between 3-stars and wins is strong than Blue-Chip count and wins, though it is close. This might just be a slightly different way of looking at the same thing, but I found it interesting.

Lifelong Florida Gator and Miami Dolphins fan, but love to research sports stuff.
View all posts by SORT14

Published

2 thoughts on “Number of 3-stars on roster negatively correlates to wins: SEC 2019”

This says a lot about a few teams, e.g., Florida does a great job of coaching up 3-star players while TA&M and Tenn do not do a good job.
It would be interesting to see how other star rating reflects to wins to see if the correlation is consistent. i.e. Number of 5-stars on roster correlating to wins: SEC 2019, Number of 4-stars on roster correlating to wins: SEC 2019, Number of 2-stars on roster correlating to wins: SEC 2019, Number of 1-stars on roster correlating to wins: SEC 2019, Number of unranked on roster correlating to wins: SEC 2019, Number of JUCO on roster correlating to wins: SEC 2019, Number of transfers on roster correlating to wins: SEC 2019. Just a few to consider

Yea, I’ve looked at some of those things. Since there are so few 5-stars the correlations are weak, but combined 4 and 5-stars (Blue chips) is a pretty good correlation. I don’t really have enough information about 2-stars and below, but I would think that having a bunch of those is probably bad for most P5 teams. I have been looking at the impact of transfers, but that is still underway. Thanks for the comment!

This says a lot about a few teams, e.g., Florida does a great job of coaching up 3-star players while TA&M and Tenn do not do a good job.

It would be interesting to see how other star rating reflects to wins to see if the correlation is consistent. i.e. Number of 5-stars on roster correlating to wins: SEC 2019, Number of 4-stars on roster correlating to wins: SEC 2019, Number of 2-stars on roster correlating to wins: SEC 2019, Number of 1-stars on roster correlating to wins: SEC 2019, Number of unranked on roster correlating to wins: SEC 2019, Number of JUCO on roster correlating to wins: SEC 2019, Number of transfers on roster correlating to wins: SEC 2019. Just a few to consider

LikeLike

Yea, I’ve looked at some of those things. Since there are so few 5-stars the correlations are weak, but combined 4 and 5-stars (Blue chips) is a pretty good correlation. I don’t really have enough information about 2-stars and below, but I would think that having a bunch of those is probably bad for most P5 teams. I have been looking at the impact of transfers, but that is still underway. Thanks for the comment!

LikeLike