Dave Berri is full of poo: Vol. 9,129

by rosewood on May 22, 2008 1:44 PM MDT Comment 4 comments

So I don't know how many of you are into Dave Berri and his Win Score/Wins Produced crusade, but I loved this little bit of analysis he came up with:

Apparently, because of his WP48 drop, Leon Powe is more responsible for the Celtics' decline in the playoffs than Kevin Garnett.

Of course, he doesn't quite explain how Powe, who's played only 10 mpg in the C's losses, contributes more to the losses than Garnett, who has played 40 mpg in those losses and has a worse win score per minute in those losses than he did in regular season losses (.269 vs .300).

Sweet analysis.

As per Pliny request: ; the link ;

0 recs | Comment 4 comments

Story-email Email Printer Print

Comments

link, por favor.

"True glory consists of doing what deserves to be written, and writing what deserves to be read".

by Pliny the Elder on May 22, 2008 2:31 PM MDT reply actions 0 recs

Sorry about that

It should be there now.

Sven to Benfica! Please, please, please!

by rosewood on May 22, 2008 2:48 PM MDT up reply actions 0 recs

First thoughts

I haven’t had a chance to look at it in depth yet, but I do remember another post by him where he states that his analysis and metrics show that all players all over throughout history decline in terms of productivity in the playoffs. He says this includes the greats (Jordan et al.) and the contemporaries. He states that the myth that players raise their play in the playoffs comes from the fact that many plays increase their scoring, but such players subsequently decrease their efforts in other areas (rebounding, blocking shots, stealing, assisting, shooting efficiency, etc.). Given that, I think the dropoff for Garnett was to be expected—a drop off for all players is to be expected in the postseason when using his metric. I think what he means is that the dropoff for Pierce/Powe has been more severe than the usual dropoffs for players.

Also, I watch the Celtics quite a bit and Leon Powe is absolutely critical to their success. His play really has not been up to par with the standards he set during the regular season. Then, he seemingly outhustled his way to multiple rebounds in his short time off the bench and usually converted these into free throw attempts or good looks at the basket. I realized this even before reading this post.

by RealTangiblesGuy on May 22, 2008 3:05 PM MDT reply actions 0 recs

Clarification

Berri has pointed out that a players average WP48 will be lower in the playoffs than it is in the regular season, which isn’t too surprising since the competition is better. So yeah, that KG is playing worse is losses now than in the regular season, I’m sure that he’d anticipate.

My problem is that he’s using the numbers from the Celtics playoff wins and losses to explain their decline. To me, the only relevant numbers are the numbers in the losses. And Leon Powe simply is not playing enough during the losses to make the detrimental impact that Berri wants to prove.

Basic numbers (from the first two rounds since Berri didn’t look at Gm. 1 of the Pistons):

In C’s 8 wins: Powe 17 mpg

In C’s 6 losses: Powe 10 mpg

So I would posit quite adamantly that 70% of Powe’s sample (the minutes in the wins) is irrelevant. What I want to see is that the accumulated Wins Produced from Powe’s 10 mpg has worse net differential than Garnett’s in 40 mpg.

Put another way: if KG is playing four times as much as Powe in the losses, then Powe’s net production differential has to be four times as bad as KG’s to even make their relative detriment equal. For it to be almost twice as bad, as Berri comes up with, he’d have to be playing almost eight times as poorly.

I’m doubting this until I see it in numbers. And then I still doubt it, because I think Berri’s system is theoretically wrong when applied to individual players. But that’s another post.

Sven to Benfica! Please, please, please!

by rosewood on May 22, 2008 3:34 PM MDT up reply actions 0 recs