calumnus said:
So based on their last season, here is the roster in Win Shares per 40 min:
1. Camden PF .148
2. Ilic C .140
3. Dorsey SG .131
4. Petraitis SF .117
5. Dort C .096
6. Yeanay PF .089
7. Bell SF .077
8. Pippen CG .075
9. Ames PG .068
10. Campbell SG .028
11. Pavlovic PG -.107
12. Carr PG ?
13. Ruff SG ?
Here is our main players last year based on minutes played:
1. Stojakovic SG .117
2. Petraitis SF .117
3. Wilkinson SG .104
4. Sissoko C .146
5. Blacksher PG .063
6. Campbell SG .028
7. Dort C .096
8. Ola-Joseph F .102
9. Tucker G -.003
Team average was .090. Again .100 is average and a team of 5 average players would be expected to go .500 against average competition. With an average of .090 we would have been expected to go .450 or almost 15-18. We went 14-19.
So all kinds of caveats I know, treat this as just a "for amusement in the off-season exercise" but so far, indications are we could be improved. The average of our top 9 is .109 so we are looking at going .546 or 18-15 over last year's schedule. That does not include the possibility the freshmen make an impact or players improve.
Obviously wish we had Sissoko back, but we actually look talented up front, if a little light/small. The big question mark is at guard, especially PG. Madsen has brought in a lot of guards with potential, but we will need a couple to step up to being above average college players.
This is the problem I see with this. If you actually want to use win shares to predict improvement, I think given we have had nearly complete turnover the last 3 years, you need to look at how past classes fared against their previous year's win shares as well as the type of player you are talking about. If you did this analysis at this time for last year's team, this is what you would have seen:
Campbell: .145
Blacksher: .141
Ola-Joseph: .138
Petraitis: .132
Sissoko: .123
Christian Tucker: .104
Stojakovic: .031
Dort: .086 (freshman stats. played 7 minutes total the year before Cal)
Wilkinson: N/A
What that indicates to me is that using your methodology last year's team was an improvement over last year's team as they walked in with better win shares than they ended up getting. It also indicates to me that they were comparable but a smidge better than the new guys this year. If you are going to use win shares to predict improvement, comparing this year's team to last years at the same point, you would not expect an improvement. Although, I would say that it doesn't indicate any of those things to be because taken by itself win shares has little predictive value and is close to a useless stat
First of all, look at our stats last year. I don't want to pick on Petraitis because he is a good player. But being tied with Stojakovic and substantially higher than Wilkinson doesn't pass the smell test for me. Also, based on incoming win shares, Stojakovic would be by far our worst newcomer and no one believed that at the time nor did that end up being true.
I would note some trends:
Last year, Stojakovic's win shares went up considerably. He was a major conference player who nearly doubled his use rate at Cal.
Sissoko's win shares went up a good amount. He was a major conference player who nearly doubled his use rate at Cal.
Dort went up a little, but he hardly played at Vandy. However, major conference transfer.
Everyone else went down, usually significantly. The win shares of every player from a lesser conference went down dramatically. I just don't think your numbers are apples to apples when you are comparing guys coming from lesser conferences to Cal.
For instance, Dorsey trails only Sissoko in win shares last year. However, Blacksher's win shares were significantly higher than Dorsey's the year before and then plummeted this year. Petraitis' were about the same as Dorsey in a substantially better conference the year before he came to Cal and his came down. Campbell was better than Dorsey in a conference on the same level and his production and win shares dropped a ton. Point being that if you did this exercise for last year's roster, you'd be starting at about the same place as this year's roster in terms of win shares. Which doesn't mean that this year's roster isn't an improvement - we shall see - but it does mean that win shares isn't showing that.
It's pretty consistent that when Cal takes transfers from lower conferences, their win shares drop. They all did last year. The year before our small conference transfers also dropped - Jalen Cone from .107 to .066, Kennedy from .123 to .086.