This weekend - while watching the NFL draft (which doesn’t appear to be going so well for the Lions) - I finished my analysis of the 2007-08 NBA regular season. Over the next few weeks I will be offering comments on this past season (primarily focusing on each team). Before I start writing these comments, I thought people might be interested in looking at the Top 15 players - in terms of Wins Produced - at each position.
Table One: Top 15 Point Guards
Table Two: Top 15 Shooting Guards
Table Three: Top 15 Small Forwards
Table Four: Top 15 Power Forwards
For those interested in the MVP race, here is where the top candidates rank:
Chris Paul: #1 Point Guard (and #1 overall)
Kobe Bryant: #1 Shooting Guard
LeBron James: #1 Small Forward
Kevin Garnett: #2 Power Forward (would be #1 had he played more minutes)
Dwight Howard: #1 Center
As I have argued in the past, I still think Paul is the best choice. And of these, Kobe is the worst. That being said, Kobe did finish in the top ten in Wins Produced.
Okay, now back to the draft. The Lions took another wide receiver at the top of the fifth round. I am hoping they take a new general manager with the second choice in the fifth round. Of those left on the board, who does Mel Kiper rate as the highest GM?
- DJ
Our research on the NBA was summarized HERE.
The Technical Notes at wagesofwins.com provides substantially more information on the published research behind Wins Produced and Win Score
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Simple Models of Player Performance
What Wins Produced Says and What It Does Not Say
Introducing PAWSmin — and a Defense of Box Score Statistics
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.
23 responses so far ↓
ilikeflowers // April 27, 2008 at 11:20 am
Looks like Manu greatly outperformed Kobe. Almost the same wins produced in 2/3rds the minutes.
Animal // April 27, 2008 at 11:40 am
Kobe led his team to the best regular season record in a very tough West. For that I’d give him the edge over Paul.
Andrew G // April 27, 2008 at 12:54 pm
Thanks for posting this Dave.
Kobe wasn’t even the best WP48 player on his own team (Odom & Bynum). Bynum played far fewer minutes, of course.
Ronnie Brewer is listed at PF. He’s a SG (or maybe SF).
William // April 27, 2008 at 1:04 pm
Have you considered ranking players by WP over average:
WP - (Minutes/48*0.100)
Since 0.100 is league average WP48
That would control somewhat for varying minutes played (i.e. between Kobe and Manu). An even better ranking could be WP over replacement player (which is what you hear often in baseball):
WP - (Minutes/48*0.050)
Defining “replacement player” as half as good as “average player.”
Best yet would be WP over full season replacement player, to take away some of the credit given for missing games:
WP - (Minutes/48 *(82/GP)*0.050)
By these evaluations we find the following differences from Pf. Berri’s rankings…
WPOA - PG, Same 1-11
WPOA - SG, Manu #1, Kobe #2
WPOA - SF, Iguodala and Dunleavy fall, Miller Butler and Childress now 3-4-5
WPOA - PF, KG #1, Duncan #2
WPOA - C, STAT #3, Chandler #4
WPORP - PG, Same 1-11
WPORP - SG, Manu #1, Kobe #2
WPORP - SF, Miller #3, Iguodala #4, Butler #5
WPORP - PF, KG#1, Duncan #2
WPORP - C, STAT #3, Chandler #4
WPOFSRP - PG, Same 1-12
WPOFSRP - SG, Same 1-8
WPOFSRP - SF, Childress #6, Butler #7
WPOFSRP - PF, Same 1-12
WPOFSRP - C, STAT #3, Chandler #4
For what it’s worth…
Jacob Rosen // April 27, 2008 at 1:23 pm
Last year, there was a myriad of NBA players that were very talented in terms of efficiency, but did not receive significant minutes. Individuals that come to my mind are from last season are Ryan Gomes, Craig Smith, Manu Ginobili, Amare Stoudemire, Matt Barnes (for the first half,) Luke Walton to a certain extent, etc.
Who would you say are the sleepers from this season that have the potential to blossom next year with increased playing time? Who are the studs in terms of efficiency, that the media is just not paying enough attention to? Can it possibly be Rajon Rondo, or is that just a product of his superb teammates?
james // April 27, 2008 at 2:26 pm
Animal, Paul and the Hornets had the same record as the Lakers. And as you can see Paul was vastly more productive than Kobe.
William, in terms of the MVP race, total wins produced is the best way to evaluate players. Those who play less games (and minutes) are penalized in the race. When is the last time someone coming off the bench or missing a large part of the season won an MVP?
William // April 27, 2008 at 3:59 pm
james,
I wasn’t speaking specifically about the MVP race, but you raise a good point about bench/hurt players being rewarded for missing time in a WPOA or related evaluation. I think this is addressed in my last evaluation of this type: WP over full season of replacement player.
There’s no reward to be gained from missing more minutes since we’re taking away a full season of the replacement player (unless the player is of below replacement player ability). So long as they play at an above replacement player level they benefit from additional playing time.
And additionally, it’s most “valuable” player, so why shouldn’t a player’s value be considered in the context of replacing them? (Ideally we’d calculate a replacement player level for each position for each team, but I don’t have the data on available veteran free agents at hand.)
Animal // April 27, 2008 at 8:11 pm
John Hollinger is an idiot.
Owen // April 27, 2008 at 8:56 pm
William - See, “Introducing PAWSmin. “
MT Head // April 27, 2008 at 11:14 pm
Any system which rates Andris Biedrins as the 18th most productive player in the NBA is monumentally flawed. I don’t now what is wrong, but it does not pass the laugh test.
GV // April 28, 2008 at 5:25 am
Where is T-Mac?
Daniel // April 28, 2008 at 7:13 am
I thought you were an objective scientist, Mr. Berri.
Either you choose Garnett AND Ginobili or you choose Kobe AND Duncan.
You can’t choose the number two in one category and then ignore the nearly identical number two in another category . Manu produced nearly as many wins as Kobe in 72% of the minutes. No other shooting guard was close (or even resembling a .330 superstar in your words) to Ginobili, yet you still chose Bill Brasky…ahem… Kobe. Garnett produced nearly as many wins as Duncan in 88% of the minutes– If you choose him based on better per-minute performance with fewer wins produced then…
Ginobili is the logical choice.
Daniel // April 28, 2008 at 7:16 am
And Manu played 8/11 of the minutes Kobe did– not 2/3… for what it’s worth.
Bjorn // April 28, 2008 at 7:34 am
Two things. First, since when is Ronnie Brewer a power forward? And second, to Daniel, Prof. Berri wasn’t making his own picks for mvp, he was evaluating what the numbers said about various popular candidates for MVP. Since few people are talking about either Duncan or Ginobili as viable mvp candidates, neither were mentioned.
William // April 28, 2008 at 7:47 am
Owen,
In re-reading “Introducing PAWSmin” I see that you’re right that PAWSmin is calculated relative to position averages. I assume that’s what you’re referring to.
However, WP48 reintroduces a non-zero average because it is generated (or rather, 99% generated) by
WP48 = 0.104 + 1.621*PAWSmin
So an average player will have a 0 PAWSmin, but still a 0.104 WP48.
So my comment stands about (if we’re using WP48 to evaluate players) a player’s value being relative to their replacement.
dberri // April 28, 2008 at 8:30 am
Andrew,
Thanks for catching the problem with Ronnie Brewer. I think I fixed the rankings for SG and PF.
andrew g // April 28, 2008 at 9:39 am
I think Dave would agree this is the first time he’s been accused of having a bias FOR Kobe.
GV // April 28, 2008 at 4:52 pm
Still not seeing T-Mac . . .
Anon // April 29, 2008 at 3:14 am
I don’t think T-Mac is top 15 at his position
Some Guy // April 29, 2008 at 8:30 am
While interesting, this looks more like fantasy basketball rankings than real ones. If Jason Kidd is the 3rd best player in the league, I apparently haven’t learned anything in my 20+ years of fandom.
I’ll take my own lying eyes over lies, damn lies, and statistics.
GV // April 29, 2008 at 4:27 pm
T-Mac had a win score of 6.9, which puts him in the top 15 for both shooting guards and small forwards, although he would tie for 20th for small forwards. (But I think he should be considered a SG since he starts at SG and plays more SG than SF.)
GV // April 29, 2008 at 4:30 pm
Sorry — meant tie for 15th for small forwards.
Andrew G // April 29, 2008 at 5:17 pm
He missed 16 games, so he probably didn’t compile the total wins (what was used to sort) that the top 15 did.
Leave a Comment