Kenpom - MU has best defense in 2nd best conference

[Benny B]

JayBee said that he's in great pain and he wants to know if you can help him, doc.

[Henry Sugar]

What are you getting for eFG% & what is the source? I have found errors in ALL websites/sources without trying to... But in this one I've got the same figured as Kenpom.com.

I currently have 50.8% for offensive eFG% and 47.5% for defensive eFG%. I have the offense as slightly better than Pomeroy (48.6% in conference) and the defense as slightly worse (47% in conference).

The source is a spreadsheet that I created using Kalman Filter calculations. In other words, the spreadsheet takes a factor, creates predictions, and then modifies the factor based on performance vs the prediction.

[Dawson Rental]

Juan. He was a top 200 defender in the nation last season and a top 100 this season.

Thank you!

[Dawson Rental]

My two cents

Defense
1. I don't think the 3 pt% defense will stay this low. We are worst in the conference at allowing opponents to shoot threes. There is going to be a day where an opponent goes Michigan State or Ohio State on us.

2. The 2 pt% defense /is/ sustainable. MU has done a pretty good job with eFG% defense in general.

3. I don't know if the TO% defense is sustainable. Having said that, the team keeps outperforming expectations on TO%. How well will we do on the second run through opponents?

Offense
1. MU hasn't been bad at offensive eFG%. I track us as slightly better than Pomeroy shows, although declining a bit over the past few games.

2. I don't think that the offensive turnover rate is sustainable either. At least, I hope that the team learns to protect the damn ball. Soon. If that gets fixed, then the offense will go up.

In total, MU is still about a 0.500 team. Even if the defense gets worse, the offense should get better. At the end of the day, the most likely view is still somewhere between 14-17 wins.

I think that most of your posts should start with the following note.

"Spoiler Alert: If you do not want to know how the season will play out, do not read on."

[Henry Sugar]

I think that most of your posts should start with the following note.

"Spoiler Alert: If you do not want to know how the season will play out, do not read on."

I am wrong ALL THE TIME!

[Dawson Rental]

Shows ya defense is not the key to winnin', hey? Ya still gotta put it in the hole yourself, aina?

What's the saying about offense and defense?  Something about offense can come or go, but if you put in the effort, defense will always be there.

[Dawson Rental]

I am wrong ALL THE TIME!

I thought that I was wrong once, but it turned out I was mistaken...

[Jay Bee]

I currently have 50.8% for offensive eFG% and 47.5% for defensive eFG%. I have the offense as slightly better than Pomeroy (48.6% in conference) and the defense as slightly worse (47% in conference).

The source is a spreadsheet that I created using Kalman Filter calculations. In other words, the spreadsheet takes a factor, creates predictions, and then modifies the factor based on performance vs the prediction.

Gotcha, I think. You're calculating an eFG%-related metric that adjusts actual results based on whatever assumptions. Eg, predictive v's actual & so forth..... whereas Kenpom's eFG% is no diff than mine, Jim, Bob, dave's or anyone else's - it's just actual eFG%.

[MUDPT]

I ran a correlation coefficient from last year, using John Gasaway's defensive PPP on January 21, 2014 (http://johngasaway.com/2014/01/21/tuesday-truths-were-back-edition/) and the end of year Pomeroy defensive efficiency numbers.  r value= .415.  Somewhat correlated, but definite variance.

[Jay Bee]

I ran a correlation coefficient from last year, using John Gasaway's defensive PPP on January 21, 2014 (http://johngasaway.com/2014/01/21/tuesday-truths-were-back-edition/) and the end of year Pomeroy defensive efficiency numbers.  r value= .415.  Somewhat correlated, but definite variance.

Uhhh, what?

You looked at actual ppp in January vs. Pomeroys Adj DE (used for predictive purposes) as of year end? How come why?

[MUDPT]

Sorry, used the end of the year Conference DE numbers from Pomeroy.  Not the full season.  Was trying to see if there was correlation between after 5-6 games and end of the season. When I'm not at work later, I can run January 21 versus end of season.