Warriors and WP48

Dave Berri just posted an article where he lists the WP at the midpoint of every player and every team.

WP is one of those really useful stats, which appears to have an incredibly high R2. You can see a sense of its accuracy looking at his chart, because he lists the total WP for each team and their actual wins. It's rarely off by much. Here's the chart:

And a coupe of comments:

The Warriors are one of the few teams with a big gap between their WP48 and their actual wins, with a prediction of 16 wins but, of course, only having 13. This isn't too surprising given how bad a defensive team we are, and the fact that basketball stats aren't said to capture a lot of defense.

That being said, the overall accuracy of the WP numbers for teams should, I think, give people pause before they dismiss it. The largest discrepancies between predicted and actual wins are us (3.1 wins fewer than predicted), New Orleans (3.6 wins fewer than predicted) and Philly (3.3 wins fewer than predicted). Those are the only three teams where Berri's methodology misses by more than 3 wins. That's, quite frankly, amazing. 

Anyway, here's some interesting things that leaped out at me looking at those numbers:

Brandon Jennings vs. Stephen Curry. Remember early in the season when Jennings was putting up huge numbers, and everybody was saying he was the better player? Not so fast. Curry WP48 .099 Jennings .063. Curry is looking like the better pick. (This is no doubt due to Jenning's horrid fg%). A WP48 of .1 means that the player is average ... so neither of these guys are great, yet, but they're both young enough that we can expect to see their numbers rise. (Monta's WP numbers took a jump in his third season.)

That being said, Evans and DUJuan Blair are being more productive as rookies than our guy. Interesting that a second-round guy is leading the way here. 

Staying on the draft decisions: Oden wp48 .316, Durant .240. I think because of the scoring numbers Durant as perceived as a better player. Of course, he's played a lot more minutes, and health was one of the arguments for Durant at the draft (see Bill Simmons comments about watching Oden walk), but ... interesting.

Some Warriors specific stuff:

You get a sense of how big a problem our injuries have been 'Buike has a WP48 of .256, which would make him our most productive player (assuming Raja Bell and Acie Law's high numbers are functions of tiny sample sizes - but the sample size argument raises questions about 'Buike as well).

Biedrins is having a terrible year, right? No, it turns out: Despite having one of the worse years of his career, he's still at .2 WP48 - twice as good as an average NBA center. Who here would have guessed that Biedrins contributed twice as much, per minute, to our wins as Stephen Curry did? (Again, notice the overall accuracy of the WP48 predictions before you say that means the stat is broken.) My feeling is that this is an example of how most fans overvalue scoring and undervalue rebounds.

Stephen Jackson's WP48 as a warrior: .035. As a Hornet: .082. He's a lot better for them than he was for us (no surprise there) but he's actually still not very good. (no surprise there, either). Jamal Crawford:  WP48 .087 - don't be fooled by Atl's improvement, it's nt because of him. 

This one is a shocker: Monta Ellis WP48: .049. A lot of us lamented his decline in efficiency ... but I didn't think it was this dramatic. He's been much higher in the past. This is a huge warning shot. 

Turiaf almost as good as Biedrins? Turiaf -.07, Biedrins +.02. The difference of .027 WP48 is equivalent to adding Josh Smith or Chris Bosh for NOTHING.

Anthony Randolph: .117. Above average, but not by much. Considering how raw his game is, I find this encouraging. 

Looking beyond the Warriors:

The Lakers top 3 players in WP48 won't surprise anyone: Kobe, Gasol, and Odom. However, their ranking in WP48 will: Gasol .359, Odom .275, Kobe .184. Compare that to the amount of press each player gets. 

An analysis of WP48 would be complete if I didn't include the player who makes the least sense according to this stat. Troy Murphy, WP48 .315. Has he changed from who he was as a Warrior? The overall accuracy of these numbers at predicting wins makes Murphy's individual number hard to dismiss. I still don't want him on the team, but this is a weird one. It's far and away the biggest difference between my (and everyone else's) subjective opinion of a player and his WP48 value.

How goos is Lebron James? Compare the current Warriors to Cavs not named Lebron. We look like the better team, according to WP48. Man, superstars make a difference in this league. 

Anyway, I think the table is really interesting, and should provide room for fun discussion. What leaps out at YOU about this chart? 

This FanPost is a submission from a member of the mighty Golden State of Mind community. While we're all here to throw up that W, these words do not necessarily reflect the views of the GSoM Crew. Still, chances are the preceding post is Unstoppable Baby!

Log In Sign Up

Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Golden State Of Mind

You must be a member of Golden State Of Mind to participate.

We have our own Community Guidelines at Golden State Of Mind. You should read them.

Join Golden State Of Mind

You must be a member of Golden State Of Mind to participate.

We have our own Community Guidelines at Golden State Of Mind. You should read them.




Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.