SFCityBear said:
calumnus said:
Our best players in WIn Share/40 in conference games who have gotten significant minutes are:
1. Kelly .125
2. Bradley .095
3. Celestine .069
4. Foreman .053
5. Brown .042
6. Thiemann.039
7. Anticevich .033
8. Hyder .026
9. Thorpe .024
10. Betley -.017
Kelly needs to be on the court. Whether he plays C with Grant at PF or PF with Lars at C depends on matchups. Bradley and Celestine are now obvious.
My preferred lineup:
Kelly
Anticevich
Bradley
Celestine
Brown
With Foreman and Lars off the bench (or starting, again depending on matchups).
Betley, Hyder and Thorpe would see only very limited minutes much like Kuany, klonarias, Bowser...
Isn't it oversimplifying to use one statistic and conclude a player needs to be on the court?
And looking at the list, Anticevich is in the bottom half. I'd consider him at least our 3rd best player, and players like Celestine, Foreman, and even Thiemann are ranked ahead of him on the list.
I am assuming that your win shares/40 minutes are offensive win shares, right? If so, what about the defensive side?
I know you are passionate about Kelly, and I share your enthusiasm. But Kelly has some limitations, and against certain lineups, he does not perform as well as others. "Basketball is all about matchups" - ex-Warriors coach Don Nelson. Kelly had well below average games against teams with talented frontcourt players, UCLA, USC, Arizona, and some others.
Kelly came to camp in poor shape, way overweight. He was slow, and in early games he sometimes would come to the sidelines, gasping for air. His weight and his stamina have improved some since that time. Personal fouls were a problem for him his first two seasons, and a little less so this season, though he has fouled out of 4 games, twice as many as in the entire season last year. His minutes are up a little from last season, but he still plays only a little over half a game. He has played 30 minutes in only 6 games this season. That is largely due to matchups, and probably due to stamina, and maybe the foul situation. He is a valuable player when he is on the floor against the right matchup, but coach Fox needs to be judicious about when and how to use him. He is not yet a 30-35 minute player against all comers.
Win share is the combination of both offensive win share and defensive win share. It is an attempt to take everything into account (including opponent) and state it as a single statistic. How much does the player contribute (or detract) from winning games? Win share per 40 then normalizes it based on PT. How much does the player's offense and defense contribute to wins when they are on the court?
Saber-metrics was first applied to baseball, and if you've seen the movie Moneyball a major point is that "accepted wisdom" and the "eyeball test" are often wrong. Often pudgy players are underrated, for example.
Basketball is a little different from baseball, especially in that basketball performance is very codependent on who is on the court with you whereas baseball player performance is essentially independent of the performances of the rest of the team.
Still it is useful.
The other issue, and this applies to Grant especially, is it is based on what has actually happened when he was in the game versus what you think he is capable of going forward. Grant is 6th in conference scoring per 40 minutes (8th if you do not throw out Welle and Klonarias). He is 4th in rebounding of players getting major minutes. His defensive rating is 7th. His defensive win share is 5th. However, we know he can knock down threes (.350 in conference, .370 overall). The issue is he only shoots 3.1 per game in conference making only 1.1 while playing 29.1 minutes so his one three per game really doesn't overcome his other liabilities in his WS/40. Thus, I want Grant in my starting 5 AND I want him to shoot more than he has been.
Kelly is our best rated player. He is shooting .602 in conference, leads the team in rebounds in conference per 40 minutets (8.9) and importantly has the fewest turnovers per 40 on the entire team. Our team scores more points per possession when he is on the court (not just his scoring but probably because his threat of interior scoring opens up outside shots for others). He also leads the team in defensive win share (0.2) but that is weighted by his minutes. His defensive rating is second behind Lars. Yes, as much as we think Lars does not play good defense, our opponents actually score fewer points when he is in. Just having a 7 footer in there helps apparently. However, Kelly is #2 in defensive rating.
As for Kelly's "stamina" his minutes as a freshman in our first 7 games as a starter were 25. 30, 30, 26, 31,32, 32 His minutes played decreased when Vanover became the starter and Conor's minutes increased from 10 minutes a game to over 30 for the last third of the season.
Last year there were games Kelly would play 30 and games he would play 8. He only started 11 games. Lars started 18. Obviously Fox played Kelly or didn't based on matchups, not stamina. Or maybe he was in the doghouse. This year Kelly is third in total minutes played (547) behind only Betley (638) and Brown (581). Unlike those two, there are games where he plays 33 or 34 and games where he plays 15 or 17, due matchups or foul trouble but not stamina.
If Kelly can play 33 minutes at altitude in Utah and score 11, grab 9 rebounds get a steal and an assist with no turnovers and only 1 PF, then he can play 33 any night. It might not be a coincidence that was one of our few good wins.
I would play him at C or PF depending on matchups, but I would have him on the court more (and Betley less.).
