Statistics and Dreamscapes: Fun with EWA
2010-03-07
LeBron’s PER is currently sitting at exactly 32. That number would be one of the best PER marks of all time. For most people, PER doesn’t mean too much. Since it’s a formula and doesn’t manage anything tangible like points or rebounds, a stunning PER doesn’t carry the kind of weight that it should in the mind of the public.
In a move that made his statistic more accessible if slightly less perfect, John Hollinger took PER and combined it with minutes played to create “value added” last season. The stat is a measure of how many points a player adds over a “replacement”-level player at his position. Hollinger then divides that statistic by 30 to get “wins created,” which estimates how many wins a player has given to his team that a replacement-level player would not.
Now, that statistic is obviously a rough estimate. It doesn’t account for defense or chemistry, and there is no statistical way to directly tie wins to the statistical contributions of one individual player. If you take the statistic at direct value, you’ll find all sorts of funny things, like that Brook Lopez has an EWA of 12 and the Nets have seven wins.
But just for a second, let’s use EWA to help illustrate just how statistically dominant LeBron James has been this season when compared to his peers.
LeBron James currently leads the league with 26.3 wins added. Dwyane Wade is second, with 18.3 wins added. That is an 8-win gap. The size of the gap between James and Wade is the size of the gap between Wade and Marc Gasol, who is currently 18th in Wins added. If the Cavs had lost eight more games, they would be behind the Magic for the lead in the East, and right in the mix with Boston and Atlanta.
James, Wade, and Durant are the only players with more than 15 wins added. Only nine players have half as many wins added as James does so far this season. The gap between James and Dirk Nowitzki, who is currently #8 in wins added, is the size of the gap between Dirk and Dante Cunningham, who is tied for 198th in wins added.
According to EWA, if James was replaced with a replacement-level small forward like Mickael Pietrus, the Cavaliers would be 23-41. That would make them a slightly worse team than the Clippers this season.
James has his 26.3 estimated wins in 63 games. This means he has produced .42 estimated wins per game by himself. Consider that for a second.
Here is a lineup of five players who do not have 26.3 wins combined and have played 55 or more games this season:
PG: Aaron Brooks
SG: O.J. Mayo
SF: Vince Carter
PF: Michael Beasley
C: Andrea Bargnani
That lineup would produce one less estimated win than a lineup of James and four replacement-level players.
Okay, that was some harmless statistical fun. I’m off to watch some Oscars. Go Hurt Locker. By the way, one more thing to love about EWA: The lowest EWA among qualifying players belongs to DeShawn Stevenson, who has -2.6 estimated wins at this point in the season. He’s such a great locker-room guy, though.
I have never understood all of the hate for stats. Stats essentially measure (almost) everything that a player does on the court. A stat like PER just compiles it all into one number and normalizes it, relative to everyone else. Why can’t it be said that Player A had a better season than Player B if he was more productive? As far as I’m concerned, it is how well you played, not how well you had the potential to play. Lebron is indeed having the best statistical season of all time, he’s producing more on the court per minute, per… Read more »
The reason Brook Lopez can have more wins than his team is because the other players on the team have negative wins.
Oh, and using PER as the means and the end to any argument is just dumb. Hollinger’s huge flaw is that he’s taken a formula that does an admirable job of grouping players and used it to argue specifics such as Player A is better than Player B or Season A was not as good as Season B because the number his personal formula spits out is greater or lesser.
Looking at numbers without accounting for the context in which they came (Lebron) is no different than ignoring the numbers in favor of gut feeling (Kobe). That’s the problem I personally have with arguments from both sides of the debate.
@Todd: I absolutely agree. I love the comments here, its always intelligent, reasoned comment. I think for me thats why sometimes things said about Kobe stand out as separate, and, just seem to me, to be less reasoned, and less intelligent. I think maybe I just dont get the rivalry. Things said about deshawn stevenson: hilarious, cos he’s such a terrible player and a legit cavs rival. But kobe and lebron seem to get on, never met in playoffs etc… I dont know it just really grates on me.
*** addendum to above ***
“any Cavs fans” is probably an exaggeration. I will say most Cavs fans, and almost everyone I’ve seen post here.
Correct me if I’m wrong, but this is a Cavs blog. If people want to poke fun at people that love Kobe, there’s nothing wrong with that. It’s not like the commenters are trolling on a Lakers blog. Half of the fun in sport is the ability to make fun of your rival, and I don’t really see anything other than good natured ribbing here. I don’t think anyone is perpetually posting “KOBE SUCKS!!!!!” because I don’t think any Cavs fans actually think Kobe sucks. I understand your concern for “sinking to the level of others,” but what’s typically said… Read more »
Right, again, I’m no massive Kobe fan (especially not after his play against matt barnes last night). But I really feel like we don’t need to descend into randomly bringing him into discussions about Lebron, just to slag him off. Thats for Kobe fans who have no answer to the relentless tide of statistics showing ‘bron to be awesome. I just don’t understand why its not possible to love Lebron for his amazing efficiency, and acknowledge his status as maybe even the future g.o.a.t. without simultaneously denigrating Kobe. Who is an unlovable ass, but a beautiful player to watch. and… Read more »
number of times top 2 in the NBA in PER / win shares:
Michael Jordan
10 / 11
Lebron James
4 / 5
Kobe Bryant
0 / 0
This article is propaganda trying to confuse people into thinking that Kobe’s not the best on the planet (Earth). Just say 4 championships, bro.
I would argue that Hollinger created EWA bc PER has a bigger problem – it has no correlation to the outcome of a game, which for economists, is a substantial issue. EWA actually at least attempts to create that correlation.
NBA TV former cavaliers moment of the night: after discussing and showing highlights of who has the best hands in the NBA past and present the announcer guy (forget his name) goes… “You know, I’ve noticed that people with good hands tend to be good shooters”… Eric snow pauses to think for awhile and goes. “….. I guess I didn’t have very good hands.” and then he laughs :)
The thing about per is that many peopl e dismiss it but then hang their hat on per game stats which tell us so much less. LeBron is a man among boys. He has all kinds of room to grow too if he starts limiting turnovers again and getting his ft % stable It seems comical to say the guy is underrated but only people like hollinger are comfortable comparing him seriously to Jordan and declaring that he should sweep the mvp. I don’t think people appreciate how much better he is than the rest of the future hall of… Read more »