"She’s happy and for her being so far away from home, that’s what me, as a mom, that’s what I worried about. Is she really going to be OK being this far away? She’s just fine."
— Maggie Vanoni (@maggie_vanoni) December 26, 2023
How freshman KK Arnold has adjusted to life with UConn women's basketball:https://t.co/AxgwPNKrzi
Sources are saying Strong is headed to UConn. Stay tuned! https://t.co/PGEzXwfVPV
— World Exposure Report Women’s Basketball (@WorldExposureWB) December 28, 2023
Just dropped Morgan Cheli’s bracket play highlights from @niketoc https://t.co/0pFZzeA6FT @MorganCheli @mitty_wbb pic.twitter.com/jzbpExKI92
— Hoopers4Dayz (@hoopers4dayz) December 28, 2023
Morgan Cheli is going to get solid minutes next season for the Huskies. She can play positions 1-4. I can't wait to see her in a UConn uniform. She is planning on wearing No. 23. Big legacy to live up to!!
- Massey 8
- NET 12
- Her Hoop stats 5
- AP 15
- RPI 1
- ESPN Power 9
- Creme 17, 18, 19 or 20
I thought it might be interesting to look at how ratings compare among five rating services:
AP rankings
NET
Massey
Coaches poll
Her Hoops Stats
I originally planned to look at the top 25, but it turned out to be a little more than I could manage so I decided to analyze whoever is in the top 10 in each of the five rating services, and wherever they fall in the other rating services.
One way of looking at this information is in the table as follows:
I tried to be clever and use the school colors to distinguish the teams. That may or may not work for you.
A second way of looking at the same information is to list the teams and then show the ranking for each of the five services.
Below the second table I put in the correlations; I’m a bit rusty so I’m not sure I created a standard correlation table but I think you can figure out what the numbers mean.
The first take away is unsurprising — the two human polls, AP & Coaches, are highly correlated with a 96% correlation coefficient. Not surprising as I’m sure the coaches look at what AP selects and vice versa and doesn’t depart too far in most cases. I expected them to be highly correlated although 96% is a little higher than I might’ve guessed.
A somewhat more surprising take away is that the human polls and the algorithm polls are not very highly correlated. For example, NET is only 32% correlated with the AP Poll. That’s not a huge correlation for two things that purport to measure the same thing. I’m fairly certain that alcohol sales and teacher salaries have a higher correlation.
I haven’t spent a lot of time looking at Her Hoop Stats, perhaps I should. The something to like about a rating service that thinks UConn is number five.
They are very uncorrelated with the AP poll, under 8%.
I also calculated the standard deviation of the ratings, the minimum and maximum ratings for each team and the range (the difference between highest and lowest). No one will be surprised that allating services have South Carolina first which means the standard deviation is zero and the minimum possible range.
I originally put this together a day or so ago, when Indiana was ranked 26 in the NET rankings. One day later after a decent win they moved all the way up to 17. I’m still surprised at the road to the jump, which is a cautionary tale that the NET rankings are still a bit suspect.
Kansas State and Colorado are now the major anomalies, with Colorado ranked as high as eight and as low as 24, in Kansas State ranked as low as eight and as high as 27.
.jpg)
No comments:
Post a Comment