My latest for Huffington Post discusses what Jerry West recently said about the Lakers. For those who missed this story, the NBA’s Logo argued last week that the Lakers
- have a problem on defense
- and this is because the team is old
As I argued at Huffington Post, I think West is a little bit right and a little bit wrong. And I think this whole discussion can be linked back to what was being said earlier this month. Back on January 4th I argued that rumors of the Lakers decline was greatly exaggerated. This post appeared soon after the Lakers had lost four of six games. And at that time, people were arguing the Lakers were in trouble.
Starting on January 4th, though, the Lakers won seven consecutive games. But in the past week, the Lakers have lost two out of four. And now West is questioning this team.
One suspects that each time the Lakers lose, people are going to argue the Lakers are in trouble. And each time this team wins a few in a row, these arguments go away.
When we look at the entire season, though, this team seems to be about where they should be. Specifically, if we look at the Lakers’ Wins Produced this year, and what we would expect given the performance of the Lakers’ players last year, we don’t see much difference.
As the following table notes, given what the Lakers have done across the first 45 games we would expect this team to win between 59 and 60 games when the season ends. This is quite close to what we would expect given the player’s performance last year. Specifically, last year’s per-minute performance from these players suggests a team that would win about 62 games this year.
Now a team that is going to win 60 to 62 games is going to lose 20 or 22 times. That is simply how the math works. And when this team loses, it is going to look bad. It’s defense won’t look good and/or it’s offense will look bad. That’s what happens when you lose. But since those losses are expected, we shouldn’t be reading much into these losses.
In other words, we should look resist the urge to look at data from one game or a small subset of games and then leap to a conclusion. A better approach is to look at a much large set of data, say perhaps an entire season. When we look at this much larger data set we see that the Lakers are actually pretty good this season. Yes, this team is old. And yes this team will some day be too old to compete for a title. But I don’t think that some day is today.
– DJ
Wages of Wins Journal 
Chicago Tim
January 25, 2011
“But I don’t that that some day is today. ”
You may want to revise that last sentence.
I’ve been having some doubts about Wins Produced lately. I’m following Ronnie Brewer and Derrick Rose on the Bulls, and I know that the coaches are begging Rose to attempt to score more, and aren’t doing the same with Brewer. So Rose looks inefficient (although he has improved in other categories lately) and Brewer looks efficient.
Now, I know there are volume scorers who are also efficient. They are better than Rose at this time. But what about the players who have the luxury of waiting until they are wide, wide open to take a shot, or (like Rodman or Wallace) simply take no shots at all? I know this isn’t their intent, but could they be said to be padding their Wins Produced stats?
To take the most extreme example, shooting a three from half court or farther at the end of a period is a very inefficient shot. Yet we want someone to try it, because sometimes it goes in. Aren’t there other shots that are similar in an offensively-challenged team? Maybe not all volume scorers are ball hogs.
I’ve been looking at Win Shares on basketballreference.com. I know you like Win Shares better than PER because it is not dominated by scoring. Can you explain how Wins Produced are different from Win Shares?
dberri
January 25, 2011
Thanks, I fixed it.
As I note on the FAQ, I will say more about Win Shares later. For now I can say… like the WoW metrics, Win Shares rewards efficient scoring. So you don’t get credit (as far as I can tell) for just taking shots. The WoW metrics, though, are more consistent across time. Again, I will say more later, after I work through some stuff and make more progress on all the other projects I am working on.
I should note, I am part of more than a half dozen papers that will be presented this summer at the Western Economic Association. So those projects are going to take up more and more of my time. Maybe I will blog on this meeting soon.
ilikeflowers
January 26, 2011
Tim,
You’re talking about strategy and why. The boxscore doesn’t deal with strategy or why beyond assigning a standard position. It just describes what you did. In your case (if the Bulls have more efficient scoring options) the Bulls are likely employing poor strategy with Rose and Brewer. How you are used certainly impacts your wp48 i.e. Josh Smith should not be your three point shooter. The question of whether or not Rose’s ‘true’ wp48 value is better that what he’s showing this year (0.200 at 23 is quite good though) depends upon whether you believe his use makes him significantly worse than otherwise.
You have noted that there are drawbacks to being the designated scorer but do they outweigh the benefits? If you are the focus of the defense shouldn’t your wp48 (boxscore) benefit from the additional assists? If your teammates are trying to get you open shots and trying to get you the shots that you are best at shouldn’t this alleviate some (or all) of the negative scoring efficiency effects due to defensive scrutiny? Are the unplanned end of shot clock attempts for a given player so far above the typical number of attempts for his position that they make any significant difference?
I don’t have the answers to these questions in general, so unless I do have the answers for a specific case I just assume everything’s a wash or a minor effect and assume that the (full season) boxscore is a good first approximation of value.
Italian Stallion
January 26, 2011
The Lakers are playing better with Bynum back.
I’m starting to think that when Noah gets healthy the Bulls may have a chance to come out of the east. Their efficiency differential isn’t far from the top of the league and they haven’t been 100% all year. First it was Boozer and now it’s Noah. Their defense keeps improving and it’s going to be off the charts with Noah back.
They may not be the favorite, but I could easily see the Bulls giving teams like Miami, Boston, and Orlando all they want.
Italian Stallion
January 26, 2011
ilikeflowers
IMHO, every model breaks down when it comes to valuing scoring because the value varies depending on the circumstances and it’s not always clear what impact a change in usage will have on any individual player.
If you put 5 peak “Ben Wallace like” players together they are NOT going to be able to score enough to win even if they are all efficient at low usage levels and do lots of other things extremely well.
If you put 5 highly efficient great scorers together their individual usage level is going to decline even though they will clearly have some ability to raise it significantly and remain very efficient in different circumstances.
As time passes I keep thinking that the standard position definitions are too limited.
It’s not about a player being a C, PF, SF, SG, or PG.
It’s about a player being a scorer, rebounder, playmaker, shot blocker etc..
You need to create a balanced team. Part of that balance is having players that have the ability to score.
While very efficient scoring is always preferable, in some situations even average efficiency scoring can be very valuable to the team and in others it can be almost irrelevant.
Matt
January 26, 2011
“Now a team that is going to win 60 to 62 games is going to lose 20 or 22 times. That is simply how the math works. ”
Unless they lose 21 times. :)
dberri
January 26, 2011
Hate to do this again, but I think IS and I sort of agree :)
Seriously… WP is a measure of how productive a player is. The entire usage discussion is about why that player is productive. Some players — as IS notes — can’t take many more shots and be efficient. But just because a player takes only a few shots it doesn’t mean that more shots must make them less efficient. It depends on the player. It is up to the coach to figure why particular players are doing what they are doing, and whether or not there is some way to change that. WP is just a measure that tells you the value of what they are doing.
ilikeflowers
January 26, 2011
IS,
I agree. Each distance from the rim has biases towards all of the four fundamental factors: possessing, scoring, assisting possessing, and assisting scoring. Each of the five standard positions is going to have some aggregate playing distance/four factors bias attached to it that not all players are going to fit. Here I’m defining ‘true’ position as a player’s average distance from the rim and their four factors measurement profile (measuring how often and well that they’re performing each factor). Whatever the ‘real’ number of true positions is, I’m sure that it’s more than five and I suspect that there are multiple championship caliber groupings of true positions. Just from the math I can say with certainty that none of these proposed championship groupings can contain all of the true positions.
Tommy_Grand
January 26, 2011
Chicago Tim, no one is saying that Wins Produced is perfect. Personally, I think it’s the best measure available for free on the Internet. There may be a superior system that a genius Vegas bookmaker invented & keeps strictly confidential. I can think of lots of tiny little ways to improve the metric, but I’m not sure that the effort involved would be worthwhile. You could keep track of playas who take charges and count those turnovers as steals. You could index the value of a possession based on the amount of time remaining on the shot clock. We all know that possession with 1 second left is less useful than possession with 10 seconds; 1-second possessions rarely result in points. As you mentioned, players’ effeciency should not be dinged when they miss “inefficient” half-court shots w/ 2 seconds left, since it’s a “use it or lose it” situatiuon. You are right. But since this happens infrequently (less than 1% of all shots taken) and since it’s distributed among players, it’s not a significant flaw in the Professor’s measure.
You could count fouling an extremely poor freethrow shooter as half a steal (or maybe .4) Obviously, the hack-a-Shaq is effective at times – a perfect metric would capture this. You could also track players who induce opponents (and opposing coaches) to commit technical fouls. What about a terrific defender whose man is charged with traveling more than one standard deviation above the mean?
Some point guards attempt difficult shots on when they see that a teammate (due to postion & ability) seems likely to score the ball on a “putback” or follow up shot. Hence, some of these “misses” are effectively assists. But with no official stat, that data would be nigh impossible to track. Plus, it does not happen often enough to render WOW analysis suspect.
jbrett
January 27, 2011
WAY off topic, but speaking of Jerry West–
NBA-TV is showing the 1972 All-Star Game the other day, and I hear this sentence from Keith Jackson: “This year Jerry is averaging just under ten steals per game.” Ten? TEN?? Where do I start? Can that even be possible?
1. Bill Russell is the color guy in the booth; he doesn’t dispute the statement. Later he calls West and Walt Frazier the two best defensive guards in the game.
2. West had SIX steals in the All-Star Game, playing about half the time. In the fourth quarter he seems to take the ball away whenever he wants it.
3. Who was keeping track of his steals? It wasn’t the league, was it? If so, was it for their own amusement? Or a trade secret?
4. Halfway through the season, statistical averages may regress, but they’re not a fluke. If he’d just STOPPED stealing the ball at this point, he’d still finish the season near five per game.
Sorry to derail the theme here, but this was my WTF moment of the century to this point. Can anyone illuminate this for me?
reservoirgod
January 29, 2011
IS:
Heat haven’t been healthy all season, either. 1st no Miller, then no Haslem, then no Wade, LeBron or Bosh. They’ll be just as improved as CHI w/ Noah when all their players are healthy.