Editor's Note: Just fantastic statistical analysis here. Say what you want about the validity of win shares, but BWoods clearly did his homework here and provides some great fodder for thought. I highly recommend this piece. -PM
I think we can all agree that the premise for 09-10 is that 07-08 and 08-09 just don't count because of missing Agent Zero and then Zero + BH. So, what happens if we take the time machine back to the 06-07 Wizards (umm, without that whole Gil injury/first round playoff sweep part) and account for (1) who has been lost and gained since then and (2) what kind of difference in production we might see over the minutes that are to be allocated, while holding constant the Big 3 and Mr. Haywood (though we'll be making some adjustments on his production given his likely increase in minutes versus the bad/good old days).
I'll use 06-07 stats for players lost and 08-09 for new players, except for a couple of small adjustments. (I think it is fair to give Mike Miller the benefit of the doubt on his production last season and give more significance to 07-08.) Stats are from basketball-reference.com or directly derived from their stats.
First, let's take a little tour through 06-07. The Wiz finished 41-41, finishing second in a fairly poor SE division. Their pythagorean record was actually 1 game worse than that, as they allowed .6 ppg more than they scored. The team finished 3rd in the league in O-Rating and 28th in D-Rating. The Big Three were the Biggest Three around; the modified Princeton was the hottest dance in town; Bulletproof was basically in his delayed rookie season; James Lang briefly seemed like he had potential, if not tremendous well-hidden upside potential (hey now, the spurs liked him); arvis Hayes was doing his arvis things; Mike Hall, Donell Taylor, and Roger Mason had some opportunities; Antonio Daniels was going hard to the hoop and never ever losing the ball; Etan was getting big undersized center minutes while some folks refused to use the letter n at the end of BH's name, and we were still in the reign of terror of both Mr. Booth and Mr. Ruffin.
From that team we still have The Big Three, Mr. Haywod, DeShawn, and 'dray. I think, given the direction Ernie has taken and the passage of time, it is fair to consider DeShawn essentially replaced and that the Wiz have a 'dray now that is a functionally different player (at least statistically) than they did back then. So, I'm putting both of them in the replaced category.
Gilbert played 74 games (let's not think about why it wasn't 80 or more), AJ had 70, Caron 63, and Haywood 77. I think those, in aggregate, are a safe working assumption for 09-10. At least for the Big 3, even if they are healthier and play more games they *should* see somewhat reduced minutes in the games they do play.
The team win shares worked out rather neatly to 41.3. Here are the total minutes and (cumulative) win shares for the Dearly Departed for the 06-07 season:
Daniels 1,761 4.4
Deshawn 2,419 3.1
ET 1,246 2.8
Arvis 1,626 1.6
D-Song 700 1.2
Booooth 380 0.6
Bulletproof 682 0.5
Lang 55 0.1
Ruffin 271 0.0 (it felt like more minutes than that)
Mike Hall 13 0.0
Mason 492 -0.1
D Taylor 369 -0.2
So, these folks were "responsible" for 14 wins and used up just over 10,000 minutes (208 full position-games), or about half of the total available minutes, to do it. AD was The Man on that bench.
Here are the "new" players with their 08-09 minutes and win shares, except for Miller, for whom we'll use 07-08.
Vale 1,143 2.0
Dr.dray 1,703 1.4
McGuire 2,072 1.7
M Miller 2,474 5.2 (note that he actually peaked in 04-05 and 05-06, when he was around 7 wins per season, back before the Griz broke his spirits)
N Young 1,837 1.2
Foye 2,494 3.1
'avaris 1,130 0.0 (ouch, didn't realize it was that bad)
M James 1,575 -0.1
Now, to give a very simplified baseline, if we just strip out Young, Javaris, and James, the others give us just under 10,000 minutes and 13.4 Wins, which, oh no, is actually worse than what they are replacing versus 06-07. That was not what I was hoping or even expecting to find.
Ok, let's try to instead think about how the minutes might really be allocated to see if/how much that helps.
Roughly breaking down the minutes from 06-07 gives us the following to work with (I'm over-simplifying a little by putting all AD's minutes at PG even though he saw time with Arenas, all DeShawn's minutes in the SG-SF group, even though he ran some PG, etc., but I think it works well enough):
Let's assume we get 08-09 Foye-level rate of production for all of the PG minutes.
Let's give Mike Miller 2,500 of those SG/SF minutes, but let's do it on his 04-06 level of production. We'll give another 1,000 each to McGuire and Young.
And, for the PF/C, let's be a little creative. Let's first give Haywood 400 more minutes (at 06-07 level, for consistency) to get him closer to starter-type minutes (he had only 1,740 in 06-07). For the other 2,900 minutes, let's do this: let's assume we can get the same rate of production for those minutes that McGee provided in his playing time last year. He produced, from a Win Shares perspective, at a much higher rate than Blatche. I think we need to assume that with the "palace of good play" a thing of the past and given the natural improvement one would hope to see from both McGee and Blatche, it should be reasonable to expect that McGee's rookie season productivity should set a baseline for the quality of those minutes.
SG/SF: 9.12 (Miller is 7.65, McGuire is 0.82, and Young is 0.65)
PF/C: 5.87 (Haywood is 0.8 and McGee-level is 5.07)
So, if you buy all of that, it amounts to an upgrade of just over 3.5 wins over the 06-07 replaced players, given their time. That takes it to a roughly 45-win team.
There may still be a couple more places we can reasonably find some more wins within the existing roster. First, 07-08 Haywood was much more productive than 06-07. If you take the 2,140 minutes we were awarding him and instead of using 06-07, use his 07-08 productivity, he goes from 4.3 win shares in that time to nearly 6.4. Given the legitimate change in his role and improvement in FT shooting, it seems realistic to think that improvement should stick. That buys 2 more wins. That brings it to a 47 win team. Not too bad, but not a championship contender.
The wins get harder to find after that. Caron Butler had 1.6 more win shares in 07-08 than he did in 06-07, in somewhat fewer minutes, but then saw his rate crash last year. If we are willing to assume that the improvement to 07-08 was real and last year was a fluke of injuries and carrying too much of the load, we're at 48.5.
If you want this roster to get that last 1.5 wins to get this to a 50-win team, it will have to come from some combination of Arenas and Jamison playing more than they did in 06-07 at a comparable level of production (to crowd out some of those lower-production minutes), which is actually possible, as Arenas was actually more productive in 04-05 and Jamison has been at least as productive in more minutes since 06-07, and from improvement out of the Foye/McGuire/Young/Blatche/McGee group. I think it would be dangerous to assume Arenas/Jamison can really provide much more, meaning the young bench needs to provide more of a boost to break through that 50-win barrier. There is definitely some statistical basis for that, aside from their ages. It shouldn't be very difficult for Young/McGuire to improve on the 1.5 Wins in 2,000 minutes we're assuming from them. 3.0 might not be unrealistic there. And, McGuire and McGee both showed real improvement over last season (though arguably in less-meaningful game situations). Of course, we've already baked in some improvement from Blatche, but it would be good to think we can get even more from him and McGee.
On the whole, however, I'm going to give the caveat that this has all assumed a number of near-best-case breaks and has not taken into account how different the Eastern Conference landscape is now versus 3 years ago. While, therefore, the win shares analysis gives some room for optimism, it looks like this team is going to need a lot to go its way.
What hasn't been addressed is what impact Flip and Sam-I-Am might have versus Eddie Jordan in how the collective works. Unless/until someone can prove to me otherwise, I think I'll assume that is statistically unknowable. But, for a sequel to this, I might try to compare your apparent 09-10 Wizards with the parts of Flip's 07-08 Eastern Conference Finals Pistons to see what he has to work with. I'll have to see whether there is really anything interesting there first, though.