Win Share / 40 All Games
1. Robinson*
2. McCloskey*
3. Thiemann
4. Kuany
5. Askew
6. Newell
7. Okafor
8. Alajiki
9. Brown
10. Bowser
11. Anywanu*
12. Roberson
13. Clayton*

*Less than 60 minutes played

Win Share per 40 Conference Games Only
1. Askew
2. Kuany
3. Thiemann
4. Newell
5. Okafor
6. Alajiki
7. Anywanwu
8. Bowser
9. Brown
10. Roberson
11. Clayton

The numbers suggest:
1. Lars over Okafor.
2. Kuany is our best option at PF
3. Askew is valuable, the fact he missed our two wins may be an anomaly.
4. Newell has been solid as a true freshman
5. Brown's recent improvement has not yet shown up in the numbers
6. Clayton has committed a lot of turnovers and missed a lot of shots. Hopefully that is attributable to rust.

Overall, Fox had been starting:
Lars
Kuany
Newell
Askew
Brown

That still appears to be our best lineup. Of the guys getting major minutes coming off the bench, only Okafor and Alajiki have positive numbers, though Bowser now does for conference games. Clayton and Brown are the wildcards. Hopefully they are on the upswing.

calumnus said:

---

Win Share / 40 All Games
1. Robinson*
2. McCloskey*
3. Thiemann
4. Kuany
5. Askew
6. Newell
7. Okafor
8. Alajiki
9. Brown
10. Bowser
11. Anywanu*
12. Roberson
13. Clayton*

*Less than 60 minutes played

Win Share per 40 Conference Games Only
1. Askew
2. Kuany
3. Thiemann
4. Newell
5. Okafor
6. Alajiki
7. Anywanwu
8. Bowser
9. Brown
10. Roberson
11. Clayton

The numbers suggest:
1. Lars over Okafor.
2. Kuany is our best option at PF
3. Askew is valuable, the fact he missed our two wins may be an anomaly.
4. Newell has been solid as a true freshman
5. Brown's recent improvement has not yet shown up in the numbers
6. Clayton has committed a lot of turnovers and missed a lot of shots. Hopefully that is attributable to rust.

Overall, Fox had been starting:
Lars
Kuany
Newell
Askew
Brown

That still appears to be our best lineup. Of the guys getting major minutes coming off the bench, only Okafor and Alajiki have positive numbers, though Bowser now does for conference games. Clayton and Brown are the wildcards. Hopefully they are on the upswing.

---

how do you calc win share?

I answered my own question. Here's the link to the calc:

https://www.sports-reference.com/cbb/about/ws.html

I see a few problems with win share, although I don't fully understand it so there could be a rebuttal. Here's some issues:

1. individual offensive ratings don't include assists. That seems off for two reasons: (1) it's an available stat and it directly contributes to a score (2) they count turnovers

2. individual defensive ratings don't include being scored against (it only counts blocked shots, steals and defensive rebounds)

And the problem with all these types of stats is they are limited by a recorded stat. On the NBA level they capture a lot of stats that aren't in the box score. I don't know any sources where these extra stats are available on the college level. I'm sure some individual schools record other stats for their own analytics purposes (I read Michigan pay a team of students to do this)

Regarding Lars vs Okafor, it doesn't take into account the double-triple teams that Lars draws. Coaches game plan their defensive strategy around the top scorers, and most have chosen to stop Lars. Okafor doesn't yet get that attention. If it's truly doesn't account for assists, it doesn't give credit to Okafor's good passing. It also doesn't account for the turnovers that aren't credited to Lars due to his poor hands (although I've argued a fair amount of blame goes to some of his passers)

But the biggest thing it misses is points scored against each player. This is the biggest problem in the stat and one of the most important things to assign a value to a player.

With that said, I think this is a reasonably good metric to help evaluate a player, and again I don't understand it well enough so am probably misunderstanding something

HoopDreams said:

---

I answered my own question. Here's the link to the calc:

https://www.sports-reference.com/cbb/about/ws.html

I see a few problems with win share, although I don't fully understand it so there could be a rebuttal. Here's some issues:

1. individual offensive ratings don't include assists. That seems off for two reasons: (1) it's an available stat and it directly contributes to a score (2) they count turnovers

2. individual defensive ratings don't include being scored against (it only counts blocked shots, steals and defensive rebounds)

And the problem with all these types of stats is they are limited by a recorded stat. On the NBA level they capture a lot of stats that aren't in the box score. I don't know any sources where these extra stats are available on the college level. I'm sure some individual schools record other stats for their own analytics purposes (I read Michigan pay a team of students to do this)

Regarding Lars vs Okafor, it doesn't take into account the double-triple teams that Lars draws. Coaches game plan their defensive strategy around the top scorers, and most have chosen to stop Lars. Okafor doesn't yet get that attention. If it's truly doesn't account for assists, it doesn't give credit to Okafor's good passing. It also doesn't account for the turnovers that aren't credited to Lars due to his poor hands (although I've argued a fair amount of blame goes to some of his passers)

But the biggest thing it misses is points scored against each player. This is the biggest problem in the stat and one of the most important things to assign a value to a player.

With that said, I think this is a reasonably good metric to help evaluate a player, and again I don't understand it well enough so am probably misunderstanding something

---

The offensive formula very clearly includes assists and both offensive and defensive formulas include essentially the box score, how much the team scored or prevent scoring (weighted by strength of opponent) when the player was in the game.

As you point out sample size is an issue and there are issues with co-variance: if you come in the game only with the top PG who sets you up for easy scores and plays great defense, your score will be better than your otherwise equal ability teammate who comes in with the point guard that commits turnovers and gets beat. Basically, basketball is not baseball, which is a series of discrete individual performances, it is much more an interdependent team sport making it more difficult to rate individual player performance. Still, I think the advanced statistics are interesting and informative.

"3. Askew is valuable, the fact he missed our two wins may be an anomaly."

Was it also an anomaly that in the last Askew-less win, 5 Bears were in double figures?

*My surmise is that Askew's heretofore hero/ball hog ball was pursuant to Fox marching orders. If Askew plays vs Furd methinks we will see him carrying out new orders. Askew has skills. We need him. I look forward to Friday. Haase is our 6th man. HAASE IN DA HAAS OF PAIN!

bearister said:

---

"3. Askew is valuable, the fact he missed our two wins may be an anomaly."

Was it also an anomaly that in the last Askew-less win, 5 Bears were in double figures?

*My surmise is that Askew's heretofore hero/ball hog ball was pursuant to Fox marching orders. If Askew plays vs Furd methinks we will see him carrying out new orders. Askew has skills. We need him. I look forward to Friday. Haase is our 6th man. HAASE IN DA HAAS OF PAIN!

---

Good point. I think KoreAm was the one that pointed out that with Askew out (and no one with the ability to play hero ball as time expires) Fox was forced to let the other players look for their shots earlier in the time clock so we ended up with better offensive efficiency.