It’s Christmas Eve (Merry Christmas to anyone who is reading this) and I would like to get this posted before Santa Clause comes. So here is the Pacific division, the last division to be reviewed (previously I reviewed the Atlantic, Central, Southeast, Northwest, and Southwest).
Wins and Efficiency Differential
As I have done with each division, we start with the current standings and each team’s efficiency differential (offensive efficiency minus defensive efficiency).
Phoenix Suns: 19-8, 5.2
LA Lakers: 17-10, 5.0
Golden State Warriors: 16-12, 1.8
Sacramento Kings: 11-15, -2.4
LA Clippers: 9-17, -5.1
The Suns and Lakers are quite close. In fact, one can argue that the top three teams in the West are now in San Antonio, Phoenix, and LA. In essence, Kobe Bryant is now playing on a championship contender. And one can argue – at least, I would argue – that this is all because the Lakers ignored Kobe’s request to trade Andrew Bynum for Jason Kidd.
Wins Produced and WP48 for the Pacific Division
We can see my argument when we look at the Wins Produced on the Lakers. This is reported – along with the Wins Produced for each other team in the division – in Table One.
Table One: A Team-by-Team Review of the Pacific Division
As I have done with each previous division review, here are a few brief notes on each team.
Los Angeles Lakers
The top two players on the Lakers are Andrew Bynum and Kobe Bryant. Bynum has developed rapidly into one of the game’s top centers. In fact, his production rivals the productivity of an old teammate of Kobe. My plan is to devote an entire post on this subject in the very near future.
Phoenix Suns
As was the case last year, the Suns are led by Steve Nash, Shawn Marion, and Amare Stoudemire. What’s different is the play of Boris Diaw. He has gone from good (2005-06), to bad (2006-07), to very, very bad (2007-08). If Diaw was replaced by an average NBA player, the Suns might be able to challenge the Spurs for best team in the West. He’s basically acting like a giant break on the Suns championship train.
Golden State Warriors
The Warriors are a hard team to figure. Who plays power forward for Don Nelson? When you look at this team’s big men you see Andris Biedrins, Patrick O’Bryant, Al Harrington, Didier Ilunga Mbenga, Austin Croshere, and Brandan Wright. The minutes of these players sum to 1,938. The minutes available at power forward and center, though, sum to 2,708. So there are 770 minutes being played by someone who is not thought of as a power forward.
For my calculations, I put Matt Barnes at the four spot. I also shifted some of Stephen Jackson’s minutes there as well. One could offer a different allocation, though.
Ultimately, though, it may not make much difference. The top two players on this team are Biedrins (clearly a center) and Baron Davis (clearly a point guard). The players who are taking minutes at forward and shooting guard are not the major wins producers, regardless of how they are classified.
Sacramento Kings
The Kings have a slightly above average center in Brad Miller. They also have two above average small forwards (Ron Artest and John Salmon) and an above average shooting guard (Kevin Martin). And then there is the power forward and point guard spots. Their best power forward is Mikki Moore. As detailed last summer, Moore is not a very productive player.
At the point is Beno Udrih. He’s better than what he was in San Antonio last year. But better than awful is still below average, so he’s not helping this team much.
Basically the Kings have a few good players, but no truly great player. Consequently, this team is below average.
Oh, and by the way. Just as most people seemed to predict last summer, Spencer Hawes hasn’t helped much either.
Los Angeles Clippers
And then there are the Clippers, once again staring up at the rest of the division. When Elton Brand was hurt we expected this team to struggle. Surprisingly, though, this team was the last one in the Association to lose a game. Unfortunately, once this team remembered how to lose, they got really good at it.
There’s one bright spot. The most improved player in the league might be Chris Kaman. He certainly is producing like a top center. Beyond Kaman, though, the Clippers only have Sam Cassell and Corey Maggette. Every other player on this roster is below average.
Ranking All Players
Before I close this post, here are all 72 players ranked in terms of Projected Wins Produced.
Table Two: The Players of the Pacific Division
Now that every division has been examined, we can now see the top players in the Association.
Atlantic: Kevin Garnett
Central: LeBron James
Southeast: Dwight Howard
Northwest: Marcus Camby
Southwest: Chris Paul
Pacific: Steve Nash
The current Most Productive Player (The M2P) in the Association is on this list. In my very next post, I will discuss the M2P (by the way, do you like the label?).
For now, though, I better end this post. I think I hear something on the roof, so I better get to bed. Or am I supposed to throw open the window to see what’s the matter?
– 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.
Wages of Wins Journal 
mrparker
December 25, 2007
please please please let truehoop link this soon. Kobe’s not in the top 5 in his own division. That should really get some good comments going.
As always thanks for keeping us constantly updated on the goings on in the Association, even on Christmas Eve.
good luck with your new book
Mr. Jennings
December 25, 2007
The system, which I generally like and find informative, is again befuddled by a high-energy guy like Stephen Jackson. Its a flat-out flaw in the ratings. He’s a better passer than his assist numbers, he’s a better defender than his block & steal numbers.
In fact, I’d go so far as to say the Warriors game is a chink in the entire Wages of Wins system, which does not account for some of the more dynamic aspects of team play, exemplified by the win score numbers of last year’s trade participants Dunleavy, Murphy, Jackson and Harrington.
merl
December 25, 2007
David, great stuff as usual.
I think that the warriors are a good example of a team that shows some of the flaws in the positional adjustment that your system has (Orlando being another).
Not every team plays a 1, 2, 3, 4 and 5.
I think that it would be more useful to class players as pointguards, wingmen and bigs.
The reason for that is that not every team puts two big men on the floor at any one time, and you shouldn’t be treating them like they are.
Also, I think that you should split the Warriors into a ‘with Stephen Jackon’ and ‘without Stephen Jackson’ sections. That might help you to appreciate SJax’s contribution more. In some crazy way he appears to be responsible for his whole team playing hard defense.
Hrmm, I see Mr. Jennings has posted something similar.
And with regard to Phoenix, I think this shows how they have mismanaged their cap. They should not have given big money to Diaw or Barbosa.
mrparker
December 25, 2007
I agree with the above two posts. The one “weakness”(i refuse to call it a flaw) is the need to adjust for position.
I agree that is the best way to go about the task using these numbers.
I however, love that we can look at the numbers and decide for ourself the impact of Steven Jackson. Its great to be able to isolate this player and decide for ourselves.
I am grateful that you have a found a grading system that tells the story 80 plus percent of the time.
KC
December 25, 2007
Another W fan here, floored by the Stephen Jackson numbers. I have to assume it is his shot count and efficiency, as well as middling rebounding numbers that make him compare badly to other 4s. But so what? He is not a 4. Would his WP change that much if you call him a 2 or a 3? What does this score tell us? We gotta hear more. Monta’s improvement over last year is another story to tell.
As for the ‘flaw’, I can see that you have to normalize consistently, and that saying the W’s don’t always play a 4 does not mean you can have a different norm for one team. Maybe it would be useful to have both a by-position normalized WP and a “flat rate” WP and see how that would play out with the Warriors.
dberri
December 25, 2007
Whether you call Stephen Jackson a 4, 3, or 2, his numbers are below average. Now he might be a great defensive player. But it is not just the WoW metrics that misses this, but the box score statistics in general.
Certainly it is possible that Jackson’s defense has a substantial impact on his teammates. But I would note that 1) this is not something people have generally observed with respect to Jackson throughout his career. 2) rebounds, which are supposed to be what defense impacts, do not vary much across a player’s career. Hence I tend to be skeptical when people talk about great defensive ability impacting other player’s productivity 3) the Warriors are not a great rebounding team anyways so I have trouble seeing how Jackson’s defensive ability are helping, and 4) we do not have great measures of defense (and if we do, please pass these along) so it is not clear to me how we “know” for sure that Jackson is great at defense.
I would repeat… the key to this team are Biedrins and Davis, and these players are not just good this year, but have been good historically.
mrparker
December 26, 2007
is there anything that Jackson does much better than whomever plays when jackson is out of the lineup….could it be that jackson is the missing link that provides something which might be below average yet still miles ahead of his replacement. If so this could explain alot.
dberri
December 26, 2007
mr parker,
I am not denying the possibility of this story. But I am pointing out that Jackson has not had this “impact” throughout his career. San Antonio survived his departure. Atlanta didn’t become a playoff contender with him. And Indiana got worse every year he was there.
Plus what he does in the box score is not very good. So he would have to be one amazing defender to overcome the stuff he is doing that we are actually tracking.
mrparker
December 26, 2007
maybe he’s the best boxer-outer ever
Mr. Jennings
December 26, 2007
One thing that Jackson does is guard a huge range of players, allowing the Warriors a lot of mismatches on the offensive end. During the Phoenix game he took everyone from Nash to Marion at one point or another, lets Nellie play some unorthodox lineups and cause havoc. He plays Boozer, Nowitsky, Kobe, McGrady, everyone. Just not much of a way to quantify this.
Jason
December 26, 2007
As a Warriors fan for more than two decades, what is happening post-Stephen Jackson’s return has been rather amazing. There are a couple of possibilities: one is that the 7 game sample at the beginning wasn’t substantial enough to really evaluate the team and the “improvement” is something of an illusion of sample size. Another is that Jackson really does provide something beyond the box score (whether or not this has anything to do with past performance is a separate but somewhat related issue).
Watching the games, the team appears to care more (entirely subjective) and looks to be more aggressive on defense (also subjective). Not so subjective (but still not quantified) is that Jackson does appear to shadow the other team’s most dangerous offensive players regardless of position.
The interesting thing about the Warriors is that they seem to be surviving by simply getting far more FG attempts than their opponents and taking (and making) a large number of 3 pointers. Since they’re not particularly adept at offensive rebounding (or rebounding in general) it appears that the differential in FG attempts comes from the turnover differential (effectively defensive stops before the opponent gets a chance to shoot) and the fact that fewer of the Warriors possessions find resolution at the FT line.
dberri
December 26, 2007
I agree with Jason. The key is additional turnovers.
It is possible that Stephen Jackson is responsible for this difference. Given this possibility I ran a little experiment. First, I moved Stephen entirely to small forward. This gave him a WP48 of 0.013, which is still below average. I then started allocating part of the turnovers by the opponent that are not steals to Jackson.
Currently Jackson has 31 steals and the team has 232 turnovers by the opponent that are not steals. If we take 45 of these turnovers and give them to Jackson, increasing his steals from 31 to 76, Jackson’s WP48 would be 0.099. In other words, if Jackson averaged 3.7 steals per game he would be an average NBA player. If he averaged 6.1 steals per game – a truly amazing achievement – then Jackson would have a WP48 of 0.200. This is still below the marks of Biedrins and Davis, but well above average.
What this little experiment tells us is that Jackson would have to be one hell of a defender to make up for all the negatives we see in the box score. And maybe it is true that he is this good. But I am somewhat skeptical.
Jason
December 26, 2007
To add to that, the turnover margin with and without Jackson is virtually identical. In the first 7 without Jackson the team still won the TO battle (save two games against Utah, a team that seems to have their number) and the FGA battle (again, save Utah). Jackson hasn’t really changed their turnovers. What has changed since he got back is that the team is shooting better and their opponents are missing more shots.
Mr. Jennings
December 26, 2007
I think we see the same situation with Bruce Bowen, who only registers a .001 WS/48 min., yet has logged major minutes on the most consistent team in the league over 8 seasons. These guys just don’t let you get a clean shot off, don’t let you get good position, things that no stat is reflecting. Maybe the WS disparity is offset by increased boards for teammates due to worse shots taken.
Westy
December 26, 2007
Jason notes that , “What has changed since [Jackson] got back is that the team is shooting better and their opponents are missing more shots.”
Now, as has been done, you can point at stats that don’t attribute this change to Jackson. But anybody watching this team (and Jason would be a better source than me on this) would say SJ’s made a difference.
How can it be that their own shooting has gone up and opponents down based only on the addition of one player?
Here’s how I think it could’ve happened, and why it’s not captured by WS. On offense, SJ’s now (outside of Baron Davis) their top offensive threat. That means that the opponent’s top wing defender is now guarding SJ. What that means also is that they’re no longer guarding Golden State’s other wings. I don’t know this, but I bet those are the guys whose shooting % has increased. Same thing would happen to whoever Bowen guards on the Lakers when Kobe’s not in the game, their shooting % goes down. Does SJ get credit for drawing the top defender from the opponent? Nope.
Then, on defense, as has been noted, where’s SJ? He’s guarding the other team’s top player. If said player misses a shot due to SJ’s defense, what’s SJ doing? You know what, very likely not getting the rebound. Just like with Bowen, good perimeter defense is not necessarily rewarded with a rebound. And if the player you’re guarding is shooting, it rarely results in steals either. So it seems likely if the opponents’ shooting % is slightly down, someone else on the team should see their rebounds slightly up. Defensive rebounds are the end result of a team’s good defensive stand and thus are a good indicator of a team’s performance at the team level. Awarding them solely to the player who grabs them is a mistake of great proportion at the individual evaluation level, however.
Dave, if it’s not SJ, what’s your explanation for the difference in the Warriors’ play since his return? For that matter, why on a team as well-run as the Spurs does a player like Bruce Bowen (.001 WP48 this year) continue to play?
Jason
December 26, 2007
The possibility of small sample size should not be discredited. While watching the Warriors this year (and last) has led me to believe that Jackson has value that doesn’t show up in WP, I also do not think that his presence alone is the difference between 1-6 and 15-6. If their record now accurately reflects their “true strength” as a team, a 6 game losing streak would be unusual, but by no means unheard of, especially when 5 of the 6 contests were against teams that made the playoffs last year.
Another factor that I think needs to be considered is the ‘streakiness’ of teams that rely heavily on the 3pt shot. The Warriors have a relatively efficient offense, but a significant part of that is due to their reliance on the 3 ball. While .333 from behind the arc is effectively the same as .500 from 2 point space, when ‘success’ still means failure more often than not, you can expect some runs where the offense sputters simply because of stochastic effects.
Jackson helped, but I don’t think he can be attributed as a single variable responsible for all differences before and after. I suspect he’s had a bigger part in the defensive FG%, but the Warriors ability to score seems more likely to be influenced largely by regular variation in performance.
magicmerl
December 27, 2007
I think that Bowen and SJax together score poorly for the same reason: their key contribution is how they disrupt the other team, and it comes in ways that aren’t captured by the box score.
If SJax is guarding the best offensive player on the other team regardless of position, then surely in some way this has to have a major effect?
You can’t measure the difference between how well the other team’s star would have dominated had they been guarded by a different player on the Warrior’s roster (which is what really matters), but perhaps you can measure how well the other team’s offensive star did relative to their average performance?