So you want to be “like Mike?” As noted by David Biderman of the Wall Street Journal, there is a downside to this objective. It appears that shooting guards – relative to the other four positions the NBA employs — are getting ‘short’-changed.
Biderman’s article draws upon a story we very briefly note in “Stumbling on Wins” (our next book). After controlling for a host of explanatory variables (player performance, team performance, games played, age, market size, draft position, etc…), we find that shooting guards are paid less (about $1.44 million less per season) than players at other positions.
This result appears to be driven by the “short supply of tall people.” As observed in The Wages of Wins (our first book), the labor market in the NBA is characterized by a limited supply of workers. There simply are not many tall people for basketball teams to consider hiring. Or on the flip side, there is a relatively large supply of little guys.
When we think about free agent salaries in the NBA the story that stands out is that scoring is over-emphasized. Points scored (as opposed to shooting efficiency, rebounds, steals, turnovers, etc…) have the largest impact on a player’s salary. So NBA teams do not get everything “right” when it comes to evaluating free agents (and a similar story can be told for draft position, minutes played, etc…). But when it comes to the evaluation of shooting guards, this decision seems to be essentially correct. The abundance of these players suggests teams should be able to acquire shooting guards for a discount.
Two more notes on this story….
- the abundance of shooting guards means it is harder for an elite guard to differentiate himself from his peers. This can be seen when we look at the top players – according to Wins Produced (a measure that evaluates each player relative to position played) — at each position. Only four shooting guards had more than ten Wins Produced last season (Dwyane Wade, Brandon Roy, Mike Miller, and Kobe Bryant). In contrast, at the other positions we see eight point guards, seven small forwards, seven power forwards, and nine centers who were in double digits in Wins Produced.
- the story we see with respect to salaries is also seen with respect to the NBA draft. Controlling for performance (and a host of other factors), shooting guards are taken later in the draft. As we get closer to the NBA draft I will offer more on this subject (it is also in the next book).
Let me close by noting that this is not the first time Biderman has referenced my work at the Wall Street Journal. Here are three more recent stories from Biderman that reference my research.
After Age 25, It’s All Downhill for NBA Players
In the NHL, More Dollars Equals More Wins
Few Starting Lineups Could Top These Celtics
One should note that both the shooting guard and aging story are briefly noted in Stumbling on Wins (the book has many, many more stories). And while you wait for our next book (just a few more weeks), I highly recommend the sports section at the Wall Street Journal. Here is a brief selection of recent stories:
We’re Picking the Jets to Win—Again (for insights into picking NFL games)
A Random Walk to the End Zone (a comment – from “Fooled by Randomness” author Nassim Nicholas Taleb – on fantasy football)
11 Minutes of Action (how much time is there really in a football game?)
In the NBA, 3 Is Cheaper Than 2 (check out Evan Eschmeyer’s – former NBA player — post in the comment section to this article)
What Price Vikings Fandom? Funny You Should Ask (the value of being a fan, from the research of Colorado College economist Aju Fenn)
The Count: Arenas’s On-Court Liability (Gilbert Arenes doesn’t help as much as his reputation suggests)
The Count: Hall of Fame Voters Snub Stats (what the stats say about who should be in baseball’s Hall of Fame)
What do these stories have in common? The Wall Street Journal is really emphasizing what the numbers say about sports. So if you haven’t already, take a look at the sports coverage offered by the Wall Street Journal. For those who like numbers and sports, the Wall Street Journal sports page is a place to go.
– DJ
The WoW Journal Comments Policy
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 
Italian Stallion
January 18, 2010
I totally agree on the supply and demand issue.
The question that remains unsettled (at least in my mind) is whether SGs are actually less productive in terms of WINs produced or whether the current available box score statistics and associated models are not capturing some of their value.
For example, while most NBA players score more efficiently around the basket than from the outside, it appears that some quality outside shooting is a prerequisite for a successful team.
Those outside shots get as much credit statistically as points scored around the basket (other than 3s).
So by definition, SGs and other perimeter players get penalized for their skill relative to paint players even though what they do is a requirement and actually more difficult.
There are a few things like that.
I realize that this model tries to capture some these things with positional adjustments, but there is something counter-intuitive about all this.
When I was a kid, I didn’t work very hard on making layups. I worked hard on my outside shot until I was good enough from some spots on the floor to be comfortable taking the shot. When I was finally comfortable, I considered myself a better player. Yet statistically I became a worse player as soon as I started taking them because they were still lower probability shots than the layups I was taking before. All this despite the fact that the team needed me to take some outside shots being that I was playing PG.
benamery21
January 18, 2010
Phoenix (and coach) Flailing.
Alvin Gentry must have thought he’d tried everything else to stem the Sun’s slide. Tonight he changed his starting lineup, substituting the significantly less productive (so far this season) Barbosa and Lopez for Richardson and Frye (who combined for a minus 3.5 WS coming off the bench). Phoenix lost again to Memphis (125-118).
Lopez produced–but he did it at the expense of Stoudemire–who (although productive) attempted just 9 shots and scored only 12 points. The Suns have tried the twin tower approach with Nash before (remember Shaq?); it doesn’t work with their game.
This is only the latest in a series of losses attributable to poor player time allocation by Gentry. Several were losses due to unwarranted use of Earl Clark, who (mercifully) made no appearance tonight.
While the Suns outrebounded Memphis, they also committed 7 more fouls, resulting in Memphis outscoring Phoenix at the line by 12. STAT and Lopez were responsible for 11 of Phoenix’ 22 fouls and missed 4 of Phoenix’ 5 missed free throws. Barbosa(2/10), Richardson (0/3) and Frye (0/1) negated Phoenix’ usual edge from 3-point range as Phoenix went 10 for 29. Nash contributed 6 TO’s playing with the new line-up.
todd2
January 19, 2010
Some SG’s are valuable because they score, others because they’re stoppers. The Oakland A’s made a big splash recently because they signed a player because of his glove. Are there metrics available to measure a basketball player’s defensive ability?
Sam Cohen
January 19, 2010
If the “short supply of tall people” explains why shooting guards are paid less, then why aren’t point guards also paid less? And shouldn’t it be even harder for a point guard to differentiate himself from his peers?
kevin
January 19, 2010
Sam, point guards have more extensive skills and responsibilities than shooting guards. Point guards can do everything a shooting guard can do, plus they can ball handle, set the offense and break presses.
It kind of reminds me of the old baseball saw; if you have an outfielder who can run but not throw, you put him in left. If he can throw but not run, he plays right. If he can do both, he plays center.
Point guards are like centerfielders.
Sam Cohen
January 19, 2010
Kevin- I understand your point, and agree that GM valuation of the relative importance of shooting guards vs. point guards could explain some of the difference in pay.
But how does that relate to the short supply of tall people? There are many more 6′ people in the world (i.e. potential point guards) than there are 6’5″ people (i.e. potential shooting guards). So point guards do more than shooting guards, but theoretically there should be many more point guards with all of the necessary skills. Thus, shouldn’t it be harder for point guards to differentiate themselves from their peers than it is for shooting guards? But according to Professor Berri’s post, that’s not true (since there are eight point guards with double digit Wins Produced). It’s possible that we happen to be going through a fluky time period where there are an above average number of all-time great point guards, but otherwise I would expect there to be fewer point guards with such high Wins Produced numbers.
Michael
January 19, 2010
How strong is the relationship (assuming there is one) between points scored and minutes played?
dberri
January 19, 2010
Michael,
There is a very strong link between scoring and minutes. But, in looking at the link between salary and scoring about I used points per minute. I have also used points per game and total points. The story is the same regardless. Scoring dominates player evaluation in basketball.
mrparker
January 19, 2010
Stallion,
Maybe your statistics would have gone up had you then learned how to use your new ability to shoot from further out to get your more shots closer to the basket or more trips to the free throw line or use the now wider passing lanes to get shots for your teammates.
I hope that doesn’t sound to assholish, I’m just pointing out what I think is the difference between average and good or good and great. Its not just certain abilities/attributes its what they are used for.
I can only think of shooting guard who was ever productive yet only was special at shooting and that was Reggie Miller.
Michael
January 19, 2010
“But, in looking at the link between salary and scoring about I used points per minute.”
Thanks that clears that up.
dberri
January 19, 2010
Michael,
This is a big story in the next book. In The Wages of Wins we only looked at per game stats and salary (and per game states and All-Rookie voting). In the next book we look at more decisions and we show that the story is the same on a per-minute basis.
khandor
January 19, 2010
re: Scoring dominates player evaluation in basketball.
Only in the minds of those who do not truly understand how the game of basketball actually works.
There are three main [distinct] phases in the game … i.e. Offense, Defense and Rebounding … and, no 1 phase is more important than the other 2.
khandor
January 19, 2010
re: Point guards can do everything a shooting guard can do, plus they can ball handle, set the offense and break presses.
This observation is simply not accurate.
PGs cannot do everything a SG can do; nor can SGs do everything a PG can do.
The best players at any given specific position are multi-dimensional performers/producers; while the less-than best players are something other than multi-dimensional performers/producers on the continuum which exists between:
Multi-dimensional [high point];
one dimensional [mid point]; and,
zero dimensional [low point].
khandor
January 19, 2010
re: Are there metrics available to measure a basketball player’s defensive ability?
Yes, there are.
In general, they are kept on a game-to-game and possession-by-possession basis by coaches who understand what’s involved with being a top flight basketball player/team, and include considerably more information than just steals, blocked shots and drawn charges.
kevin
January 20, 2010
“re: Point guards can do everything a shooting guard can do, plus they can ball handle, set the offense and break presses.
This observation is simply not accurate. ”
Only in the minds of those who truly do not understand basketball.
kevin
January 20, 2010
To be more accurate, shooting guards are bigger, in general, than points because they have to shoulder more of the defensive load than points because they cannot contribute as much on offense, in general. So they are expected to rebound more and be able to guard bigger players.
That’s why Iverson is such a problem. he’s too small to shoulder those other responsibilites,a nd never developed the mindset to play point properly. so just having him out there at a 2 weakens the team. it’s kind of like having a rightfielder who can’t throw or a secondbaseman who can hit but who has trouble turning the double play.
IMO, again in general, the 2 spot is the easiest spot on the floor to play. You don’t need a whole lot of size and you don’t need the passing and ballhandling skills of a 1. All you need to really do is play D and knock down jumpers.
khandor
January 20, 2010
Kevin,
Ray Allen is a future HOF Off Guard who can also perform adequately at the Point Guard Position, if need be.
Rajon Rondo, on the other hand, is a still-developing-but-already-very-good Point Guard who cannot function adequately at the Off Guard position … because he does not have the skill-set to fulfil this role properly for a High End Team in the NBA.
There are numerous differences between the Point Guard and Off Guard positions, and the personal attributes of those specific players who can excel at:
i. PG only;
ii. OG only; and,
iii. both, PG AND OG.
kevin
January 20, 2010
“Ray Allen is a future HOF Off Guard who can also perform adequately at the Point Guard Position, if need be.”
Actually, no, he can’t. Allen is a very good scorer but his ballhandling is inadequate to play the point. If he were forced to play point, he’d be a turnover machine.
I love Ray but a point guard he ain’t.
Italian Stallion
January 20, 2010
Mrparker,
I never said I wasn’t using my new found skills to the advantage of the team. I simply said that the outside shots I started taking were lower probability shots that the layups I was taking before. So my overall efficiency dropped.
However, the team was better off because I was a better outside shooter than the players that used to take those shots.
I had a new skill that I executed at satisfactory rates relative to the norm.
The team was better.
The stats said I was worse.
Was I a better or worse player?
I’m still having a hard time swallowing the idea that SGs are so much less productive.
The problem is this.
There are way more players of typical SG height than SF, PF and C. So with a greater supply to choose from you would expect a higher quality. Yet SGs are supposedly WAY less productive than those big guys also.
If all this is true, I want to make a prediction.
The SG position is eventually going to be occupied by players that could easily play PG, but they won’t because there just happens to be another great PG on the team. Might as well start training SGs to have PG skills and get extra production!
I don’t think that’s going to happen though.
I think PGs tend to be better at driving and getting to the basket because they are smaller and faster. That’s one way they both score efficiently and draw defenders to them, which in turn enables them to pass to open players and get assists.
I think the typical SG tends to be a better outside shooter than the typical PG. Since teams need some good outside shooting, that makes good SGs valuable.
However, because they spend more time on the perimeter, their efficiency stats tend to suffer while they are doing that required job.
Which takes me back to original point.
A layup is easier than a 20 footer, but they both count for 2 points. So the player taking more layups will have a higher efficiency and rate better statistically.
But if a team needs a guy that can knock down 20 footers at reasonable rates, that’s not so easy to do, but doing so is a requirement of successful basketball, then IMHO good outside shooters are underrated statistically.
Maybe there should be a 2 1/2 point line then the current models would reflect their value. ;)
khandor
January 20, 2010
IS,
You’re raising some solid basketball points in this thread.
When fans watch and attempt to analyse a specific basketball game it’s important to keep track of at least four different types of field goal attempts, in two distinct categories:
* 3PT shots [contested and uncontested]
* mid-range 2PT shots [contested and uncontested]
* interior 2PT shots [contested and uncontested]
* layups, tip-ins and dunks [countested and uncontested]
Simply keep a running tally for both teams in the next game you happen to watch.
If you do, you might be surprised at what the results will reveal to you about what was really going on in that specific basketball game.
mrparker
January 21, 2010
Stallion,
Dude, if you are taking outside shots instead of looking for layups by definition you are hurting your team. It’d be good to have an extra skill, but if you are taking 3 pointers when its possible to get a layup you are hurting your team unless you shoot some percentage that no player has ever shot.
Anon
January 21, 2010
I definitely think that pg and sg require different skills. most pgs need the ball in their hands a lot to maximize their ability to see the floor and all that stuff. while certainly many pgs could function as a sg (steve nash as shooter/scorer but not ballhandler, sure), i think many would be significantly less productive (chris paul? still good as a sg probably, but not nearly as much so as when he plays pg). andre miller and brandon roy don’t function well together because they’re both better off with the ball in their hands. that’s why roy likes to play with someone who is comfortable as a spot-up shooter.
regarding the 20 footer thing, sure it’s a good skill to have, but if you’re shooting a lot of 20 footers, that’s bad for your team, because it’s a low percentage shot and has almost no chance to draw a foul. It’s generally the type of shot you only want to take if it’s really wide open. if you’re a good shooter shooting more than 50% on a wide open 20 footer is reasonable, and thus this wouldn’t hurt your efficiency much. if you’re not open, you may as well step back a few feet and shoot a 3. even if you shoot a slightly lower percentage, you’re getting bonus points if it goes in. otherwise you’re better off attacking the basket. if you can’t get to the basket, and you can’t get an open shot inside the 3 point line, then shoot a 3. and i suppose the person who takes these desperation 3s will have his stats hurt a bit, but in my experience bad shots are normally split up relatively evenly (and many bad shots are simply the result of someone’s bad decision-making). most nba teams are capable of running an offense and getting a decent look most of the time unless the defense is playing really well, and that doesn’t happen too much, so it doesn’t matter.
I do think there just happens to be more good pgs than normal in the league right now, and maybe a bit of a shortage of elite sgs. seems perfectly plausible.
Italian Stallion
January 21, 2010
Mr Parker,
Perhaps I am not explaining myself well. I never stopped looking for layups. When those were available I took them.
There were times in a game I was open on the outside, but didn’t take the shot because I knew I wasn’t a good outside shooter. So I passed on what should have been a good opportunity except for my lack of skill.
As I developed more skill, I started taking some of those jumpers because they were good opportunities for the team relative to the typical alternative.
I was more skilled.
The team got better results.
Statistically I was worse.
I think once you accept the fact that every team is going to shoot from the outside between “X% and Y%” of the time and that it’s a requirement of successful basketball, that outside shots are tougher than dunks and layups, that not all players are equally skilled from the outside, and that those shots will not be distributed equally among the players, you must conclude that outside shooters are underrated for doing a disproportionate part of that required job.
You can argue that positional adjustments help capture some of that, but intuitively I don’t think they capture it all or correctly in all instances.
Italian Stallion
January 21, 2010
Anon,
I would not limit the discussion to 20 footers. That was just an example.
I agree that if a player is near 20 feet and wide open, he is probably better off stepping back and taking a 3 pointer.
What we are really taking about is anything other than layups/dunks getting progressively more difficult as you get further from the basket and some players taking more of them as part of their job.
No one is going to ask Shaq to suddenly develop a mid range or outside game. But if Lebron gets double teamed and Shaq is covered, Moe Williams is going to be expected to knock down jumpers at a reasonably rate even though it won’t be as high a rate as a Shaq dunk.
We know by looking at the average FG%, eFG% TS% and shot selection stats by position that difficult shots are not equally distributed. Of course there are some players that have poor shot selection, but some of the difficult shots are a requirement of the position or player and they take more of them.
Dan Barto
January 24, 2010
Though this is a little off the current debate, I have a couple points that really have me thinking and was curious for others thoughts.
1.) Did the shift in supply change when the hand checking rule was changed a couple years back? Especially was there a cause and effect on wins produced?
2.) As GM’s evaluate and place valuations on players, is the supply of shooting guards uniquely effected by the defensive differences between college and pro games. (In my initial thoughts, college shooting guards have their NBA upside over evaluated due to inflated offensive output do to hard helpside defense. Ex: Was Ty Lawson’s pro style production undervalued in regards to wins produced and Wayne Ellington’s wins produced overvalued due to hard help situations. )
3.) Since more and more pgs may have higher wins produced and shooting guards in high supply are coaches forced to produce systems and practice repetitions that continuely reinforce the stereotypes, therefore perpetuating the situation?
4.) What team will go or goes against the grain the most besides the Lakers and Cavs?
Westy
January 25, 2010
One skill worth noting that certainly varies between position is shot creating ability. (Although Mr. Berri disputes this exists, many other advanced statisticians, including Dean Oliver in his book Basketball on Paper, recognize a varying degree of ability for this.)
Quick tall players are obviously best at this and so it’s likely some shooting guards (who seem to exist at the crossroads between shooting, height, and quickness [to have all 3 at a high level is a rare thing indeed, which may explain why elite SGs are also rare?) are amongst the best shot creators in the league.