An interesting evaluation of the NBA Central
The economics professor who writes "The Wages of Wins Journal" (who happens to be a Detroit Piston fan) has written an interesting post wherein he projects the win totals for the 2007-08 NBA Central Division based on each team's play thus far this season. He does so using the "Wins Produced" econometric formula he and his fellow authors came up with researching their book "The Wages of Wins", a formula I've referred to several times on this blog.
By his calculations, the Bucks are on pace to win 30 games. (The other day I tried the same calculations and pegged the Bucks as a 28 win team. His numbers, obviously, are far more reliable than mine).
That's kind of depressing, but seemingly accurate. Several interesting observations about the Bucks and the rest of the Central Division come from his analysis of the contributions to victory being made by individual players.
1. Yi Jianlian is doing worse than I realized.
I know I had Yi as a very below average power forward at this point in his career, but I didn't project him as costing the Bucks games. Professor Berri does. He ranks Yi near the bottom of all Central Division players. I think the professor uses a more precise prediction of playing time, whereas my playing time numbers were just dirty projections. When it comes to production, though, we arrive basically at the same numbers. If you express his numbers in the "+ or -" terms I like to use (I hate integers), he pegs Yi as a -3.55 power forward, whereas I believe I had him at -3.48, so the difference must be playing minutes.
2. Bogut is not doing as well as he did last season.
When it comes to producing the statistics that produce wins, the professor's numbers say Bogut is slightly above average among centers, but that he's producing below what he produced last season. That's a bad sign. He may be regressing slightly, or at least he isn't progressing as we would hope. If you go back to the link, the important number to look at in the professor's calculations is the 4th column. Anything above 0.100 is considered above average production for a player at that player's particular position. If you notice, Bogut is currently 0.112, which is slightly above average. However, this guy kept all the "Win Score" calculations last season, and according to his numbers Bogut was a 0.152 last season, a significant difference (that number is found in his last column). The way I express it on my blog, that translates into a +0.23 for this season compared to a +1.42 last season. Not huge, but you'd like to see it going the other way.
3. Villanueva is regressing as well.
I think Charlie V may be lost under the Crisco regime. His minutes are uneven, and I think its affecting his play. In his first two seasons, he was consistently a just slightly below average power forward (his numbers in Toronto were basically the same 0.091 that they were in his limited action last season). This season he has regressed into a way below average power forward. He is currently -0.001. That's terrible.
4. Redd and Williams are both up
I already established this, but its worth noting again. Michael Redd's Win Production is way up this season. Last season he was at .142 -- which actually put him behind Bogut for Most Productive Buck. He's at .222 this season. He deserves a shot at the All-Star game. Mo Williams is also up a few notches, from .122 to .127. However, when you compare him to the otherworldly Chauncey Billups (.310), you wonder if we shouldn't have made at run at the old Piston last summer. Billups is a big time winner.
5. Big drop-off at the 3 position as well
Notice the above average production we received last season from SF Ruben Patterson (0.157) compared to the below average performances we are getting from Desmond Mason (.028) and Bobby Simmons (.012). Interestingly, Patterson was waived by the Clippers yesterday despite the fact that he was outproducing (.089) both Mason and Simmons despite getting only sporadic playing time this season (although he was well below last year's career high production). He must have pissed someone off. Anyway, I still think the Bucks should go with either Villanueva and/or Yi at small forward, but that will never happen.
6. The Bulls should be playing Noah more
I HATE Joakim Noah, but the sad fact is he's been an above average NBA power forward this season... much better than Yi. He rates a .158 while Yi, as mentioned, is a lowly -0.016. Will that be the case down the road? Who knows? But it does highlight the way teams overvalue scorers and undervalue guys who can't score but who rebound well and convert garbage points. I'm still happy he isn't a Buck. He's not that far above average.
7. Check out the Pacers
The most stunning information in the professor's calculations is found in the Indiana Pacers column. Check out how well Mike Dunleavy is doing and how poorly Jermaine O'Neal is doing! O'Neal is actually creating losses while Dunleavy is so far above average for his position he is creating more wins than Michael Redd! Unbelievable. (Actually, I just checked it out, and those numbers are merely an exaggeration of what happened last season. O'Neal was just slightly above average whereas Dunleavy was pretty above average -- though not as above average as he has been this season).
8. Blueprint for the Bucks?
If you believe in the professor's regression analysis, and I have tested it and am a believer to a large extent, then the Dunleavy/O'Neal situation shows how the Bucks can rebuild themselves into a championship team. To rebuild, the Bucks have to identify undervalued "Win Producers"... in other words guys who's reputation rates far lower than their production, or who are undervalued because they "only rebound"... and then pick them up on the cheap. I think this is the far better route to take than waiting and hoping for the lottery to deliver, or accumulating middle class free agents who end up being overpaid.
9. Hypothetically Speaking
Let me illustrate what I mean with a "for example". Let's say Dunleavy is for real... a big assumption, but just work with me for a second. The Bucks could probably have gotten him this summer for a song... and maybe they could still get him next summer for a lower price than his production dictates. Then lets say the Bucks trade their almost certain lottery pick to the Knicks for PF David Lee and SF Renaldo Balkman -- two players I think the Knicks would gladly give up... clearly they don't value them as highly as their production would warrant. Ok, so now we've got Dunleavy (or someone like Dunleavy), Lee, and Balkman. Ok, let's further assume pairing Bogut with an aggressive power forward like Lee would bring out his very best play (which I believe would happen). Then let's say the Bucks pick up a "rebound only" backup center for Bogut, someone whose team might not mind getting rid of him... let's say Etan Thomas of the Bullets. Then you fill out the rest of the team with athleticly inclined minimum wage players.
Okay, now let's give reserve playing time to Balkman and Thomas (1500 minutes) and full playing time for the rest (2500 minutes) and a full season for all of the five I just mentioned. Now you've got this kind of potential win production:
1. Michael Redd...+3.49....11.5 wins
2. Mike Dunleavy...+4.38....14.7 wins
3. David Lee...+4.77....13.8 wins
4. Renaldo Balkman...+3.51....6.94 wins
5. Andrew Bogut...+1.90....8.75 wins
6. Mo Williams...+0.55....6.3 wins
7. Etan Thomas...+2.51....5.9 wins
Projected Win total for this team: 67
67 wins? Is that a reasonable projection? Almost certainly not. Its based on a ton of non-injury and other assumptions, and some of the player's production levels would surely suffer simply because they would be playing next to more productive players (eg Lee would probably take rebounds away from Redd). Besides, it completely ignores the Bucks existing salary structure.
But if you could put it together I'd be willing to bet that it would be at least a 50 win team. And, for a clever GM, it wouldn't be that hard to construct such a team... or something like it... overnight. All you have to find to do is identify and then acquire undervalued assets.
That's basically what the Pistons did a few years ago and it launched them to a World Championship.
posted by Ty Will @ Friday, December 14, 2007
10 Comments:
I like CV but it seems like he's not going to be able to get his chance/make the most of it here. If only he could drop 30+ in another game, but it seems like too much to hope for the Knicks to give us Lee for him.
To make the salaries match while keeping the trade as small as possible, it'd have to be something like Lee and Fred Jones (expiring about $2.somethingM) for Villanueva and a scrub like Storey or Noel. And maybe a pick. Might not seem too realistic, but getting a useful part for anyone outside our top 3 would be a big treat.
Dunleavy has four years, $36 million left on his deal, so I'm not sure how undervalued he is contract-wise. But perhaps his general rep of being vastly overpaid is unfair. Certainly the Bucks would rather have MD than Bobby Simmons at that price.
Is there a good site for season-by-season and current season WP48 and PAWS? I think that's a big problem with Berri's stuff...it's not super-easy to get the data. Until it is I don't think it's ever going to catch on in the mainstream.
As each and every one of my finance and econ profs in grad school have always maintained, they are still teaching and earning a salary.
None of them have been able to successfully translate the theories they teach into personal gain. Instead each of them frequently issue the same caution: Past performance is not a guarantee of future performance.
Let me know when you get the lock on such a formula that can guarantee future performance. I'd be happy to test it for you.
Here's a formula: Bucks - wins = sucks.
There's no way you'd get Lee AND Balkman with just any lottery pick. Maybe not even with the 1st pick. I actually think that Isiah likes those two guys especially Lee, so it would take Michael Redd to get them in a Bucks jersey.
An interesting read, but imo highlights why purely evaluating the game on stats often leads to misconception. Bogut has improved a large amount this year, particularly on defense, and his play has been vital to the Bucks. He looked good last year, he looks better this year, and the intangible of defensive presence, among other things, are impossible to measure.
I read this piece as well. I have a problem with his rankings. He has Jake Voshuhl and Ira Newble ranked as top 30 players in our division while putting Kirk Hinrich and Yi at the bottom.
I understand that "statistically" Yi and Hinrich haven't had great years. But watching those two guys on the court, I'd clearly rather have them on my team this year, right now, this minute than either Jake Voskuhl or Ira Newble.
Until Berri can explain to me how in his right mind he sees those two stiffs as contributing more to victories than Yi or Hinrich, I can't take the guy seriously.
Ruben Patterson is another case. He ranked really high last year in wins produced. Yet we let him go and now the Clippers cut him. Sure, he does some statistically efficient things, but most coaches say he is dumber than a box of rocks and doesn't know the plays, hence he won't work with a team concept.
I like Berri's stuff and this is fun to talk about. But I'm not sure I'd go after players based entirely on their rankings here. I'd want to correlate them with my own watching of all the games. Still, I do agree it could be useful to identify people you like from his formula. Balkman and Lee would fit well on this Bucks team.
The Clippers cut Patterson? I missed that news. I kind of liked how the Clippers were packed with Bucks rejects. Lets see; Thomas, Patterson, Knight, Cassel - am I missing anyone? Not to call Cassel a reject, I still like him.
here's a link to a great video on charlie Villanueva.
www.youtube.com/esnews
look up nba player with alopecia.
moshe
Those who can--do. Those who can't--teach.
Post a Comment
<< Home