College seniors are the most overpaid group of players drafted by the NBA!!!
Okay, that’s the basic idea behind this post. Let’s start with the basic methods followed in the analysis supporting this conclusion.
This preliminary analysis was powered by NerdNumbers and conducted as follows:
- Using data from basketball-reference.com, nba.com and thedraftreview.com, NBA players drafted since 1978 were split into six groups: high school players, foreign players, freshmen, sophomores, juniors and seniors. Some errors have been identified in the copies of this data, so if you notice any please leave a note in the comments section.
- The NBA’s expenses for player salaries were divided by the total number of wins in the league to determine the average cost per win for each season since 1986. Player salaries for 1987 and 1990 were not available for this preliminary analysis. Currently, individual players’ season data were omitted if the player was traded after the start of the regular season. This will be corrected and the analysis will be updated in the future.
- The salaries for each group of players were subtracted from the product of the wins produced by that group and the average cost per win to identify the “cost efficiency” for each group. In other words…. Cost efficiency = (Wins Produced * Average Cost Per Win) – Player Salaries.
Here is how college seniors compare to players coming out high school:
From 1986 to the present, college seniors were paid over $700 million more than their production on the court warranted. Much of this over-payment, though, occurred in recent years.
The average cost per win produced in the NBA from 1986 to 2011 is $0.98 million. The average player produces 0.100 wins per 48 minutes (WP48), and in 1998, college seniors produced an average of 0.101 WP48. If we look at all years from 1986 to 1998, seniors were only overpaid three times.
Since the lockout-shortened season in 1999, however, the performance of college seniors has fallen off but their costs have not. The average college senior only offered a 0.069 WP48 in 2011. And seniors were collectively overpaid 12 times in the thirteen seasons since the lockout.
On the other hand, players who came directly out of high school have produced an above average 0.144 WP48. And these players have been underpaid in 13 of the last 16 seasons.
The results are listed below. Click the links for each group to see a spreadsheet of the analysis.
- Total Wins Produced: 1,475.4
- WP48: 0.144
- Total Salaries: $1.87 billion
- Average Cost Per Win Produced: $1.27 million
- Cost efficiency: Underpaid by $313.8 million
- Most productive player’s cost: Kevin Garnett with 287.3 wins produced and $270.1 million in salary
- Least productive player’s cost: Al Harrington with -12.8 wins produced and $68.6 million in salary
- Total Wins Produced: 1,585.5
- WP48: 0.107
- Total Salaries: $2.09 billion
- Average Cost Per Win Produced: $1.32 million
- Cost efficiency: Underpaid by $105.3 million
- Most productive player’s cost: Dirk Nowitzki with 159.2 wins produced and $141.3 million in salary
- Least productive player’s cost: Andrea Bargnani with -20.4 wins produced and $29.5 million in salary
- Total Wins Produced: 934.5
- WP48: 0.099
- Total Salaries: $1.24 billion
- Average Cost Per Win Produced: $1.33 million
- Cost efficiency: Underpaid by $93.6 million
- Most productive player’s cost: Shawn Kemp with 115.4 wins produced and $64 million in salary
- Least productive player’s cost: Dajuan Wagner with -7.4 wins produced and $8.2 million in salary
- Total Wins Produced: 2,634.3
- WP48: 0.117
- Total Salaries: $3.16 billion
- Average Cost Per Win Produced: $1.2 million
- Cost efficiency: Underpaid by $401.6 million
- Most productive player’s cost: Jason Kidd with 282.1 wins produced and $174.5 million in salary
- Least productive player’s cost: Ron Mercer with -9.0 wins produced and $33.5 million in salary
- Total Wins Produced: 4,878.0
- WP48: 0.123
- Total Salaries: $4.7 billion
- Average Cost Per Win Produced: $0.96 million
- Cost efficiency: Underpaid by $148.1 million
- Most productive player’s cost: Karl Malone with 282.2 wins produced and $104.1 million in salary
- Least productive player’s cost: Maurice Taylor with -23.4 wins produced and $39.4 million in salary
- Total Wins Produced: 14,510.8
- WP48: 0.093
- Total Salaries: $12.4 billion
- Average Cost Per Win Produced: $0.85 million
- Cost efficiency: Overpaid by $736.8 million
- Most productive player’s cost: John Stockton with 279.3 wins produced and $66.7 million in salary
- Least productive player’s cost: Clifford Robinson with -27.2 wins produced and $60.8 million in salary
While members of the National Basketball Players Association stand together during the lockout, maybe they should begin to question the value of higher education for their players.
The NBA owners don’t collectively lose money with bad contracts. This is because player salaries are a fixed cost as a designated percentage of basketball-related income. So when the Knicks overpay Eddy Curry, this doesn’t really cost the owners as a group.
Who does lose from bad contracts? Perhaps surprisingly, it is the player. The majority of players lose money with bad contracts because they reduce the pool of money available for the rest of the players. So while younger, productive players remain underpaid by salary restrictions (like “rookie-scale contracts” and “max deals”); the older, less productive players get to be overpaid by the fruits of their peers’ labor.
So maybe NBA players should be thinking about the benefits of rookie salary scales and the factors that drive bad contracts in basketball.
– Mosi Platt
If you enjoyed this article, then you may also like a similar analysis posted at the Miami Heat Index blog – Heat Check: Value of a College Education in the NBA.
Wages of Wins Journal 
motherwell
July 15, 2011
I’m not sure what conclusion to draw from this. Does that mean that the NBA is poor at projecting non-seniors future production, or that seniors, playing college closest to their athletic peak (24 years old) are overrated? Or is it that more seniors make it as 2nd round picks, skewing the data?
This is an interesting start, but without a “why”, I’m not sure what conclusion can be drawn from this, or what behavioral modification (like discount seniors) can be achieved.
Michael Weis
July 15, 2011
I wonder if it would make any difference to take out the best and worst out of the categories. Say, take out the top 5% and bottom 5%… kick the extremes to the curb for a minute.
Doesn’t D-Berri have some draft>>>>>production models? I think time-of-service-pre-nba (or age) was statistically insignificant, but I might be wrong.
mosiplatt
July 15, 2011
@motherwell:
I agree with you. I just started down this path to settle an argument w/ my dad (see the Miami Heat Index article linked at the end of the post) but found the data very interesting once I started pulling it together with the wins produced numbers Dre provided. Maybe we’ll discover something together, maybe it goes nowhere. Hopefully the posts along the way will be interesting, though.
Nicholas Yee
July 15, 2011
I would actually think that the biggest impact on this would be that the really dynamic prospects leave early. All it takes is 1 player a year, who is obviously NBA ready, to leave to skew these numbers greatly.
Jason Kidd alone is worth almost an extra $100 million by this metric.
In the new NBA the seniors tend to be the leftovers, and the guys who are late developers. Even someone like Dwyane Wade came out early, despite being a meddling prospect early in his NCAA career.
At any point if a prospect looks NBA ready, they tend to enter the draft immediately. Leaving players who are either middling prospects when they finally enter as Seniors, players who took 4 years to breakout (never a good sign, late bloomers are risky) and became good prospects, or Tim Duncan.
Of course this didn’t apply to years previous to the 1999 time frame, when prospects often stayed in the NCAA longer. The cultural change around entering the draft as an underclassman has changed, and I think this study shows that.
Conversely, does anyone actually believe staying in college longer can make you a better player? The benefits would be a higher usage rate, in actual games, which could help with certain skills, but just the sheer number of hours that an NBA player has, just seems to me like a no-brainer.
wiLQ
July 15, 2011
IMHO that’s just a really bad premise for three reasons:
1) Total Wins Produced by Seniors > Total Wins Produced by The Rest.
So isn’t it simply the issue of supply and demand?
Whole NBA can’t just decide “hey, let’s sign only High School Players!”.
2) shouldn’t you mention which players usually decide to skip college?
3) In the first 10 years of your sample underclassmen were very rare because of rules/stigma.
Frederic Bush
July 15, 2011
Can you separate out production on a rookie contract vs production over the course of the rest of a career? I think the former is more interesting, and more likely where the value difference comes in. Rookie contracts are massively advantageous for the teams that luck into a franchise player, and there are more of those leaving college early than sticking around.
Dre
July 15, 2011
Here’s a thought (I’d love to see verified) many High Schoolers were picked at top prospects early in the draft. Some panned out, some did not. Early in the draft teams take risks on hot prospects. You get some Dirks or Garnetts but you also get Darkos and Kwames.
Later in the draft teams are more conservative and draft/trade for need. In short teams with expensive picks are risky. Teams with cheap picks are frugal (at least that’s my hypothesis, now someone with more time than me prove it please) So it may be that teams have poor strategies for drafting. (Arturo’s done a ton on this)
fricktho
July 15, 2011
You can’t just say underclassmen are better than seniors. In order to be overpaid you have to under-produce your contract. It’s a question as to why are seniors paid more than their production would warrant in the NBA? It probably has to do with the fact they don’t have as much potential, so they don’t have the untapped ability to outperform their contracts. There is most likely a larger sample set of underclassmen and high school players that have significantly outperformed their contracts when they matured skewing the numbers.
This data doesn’t say much but it is interesting. (Editor’s Note: I’d say any data that’s interesting has to at least say something. . . )
Parker Flowers
July 15, 2011
It would be extremely useful to know the average draft position for each of these classes. I could see someone (like Malcolm Gladwell) completely misapplying this and saying “The only thing you need to know as a GM is that you should only draft high school players.”
mosiplatt
July 15, 2011
@Nichoals Yee:
There are people that believe staying in college longer makes NBA players more productive. That’s actually the reason I started looking at this data. People like Mike Wilbon from espn.com & my dad actually believe that LeBron struggled in the Finals because he didn’t go to college (see the Miami Heat Index article linked at the end of this post). My dad took it a step further and said that college seniors have been better in the NBA than high school players. That’s how this all began…
mosiplatt
July 15, 2011
@wiLQ:
Your point about the first 10 years is correct – very few underclassmen were drafted. But if I just look at players drafted in the last 10 years, college seniors were still overpaid as a group.
Cost efficiency by class from 2001-11:
312 Seniors overpaid $33 million
110 Juniors underpaid by $330.5 million
74 Sophomores underpaid by $500.7 million
46 Freshmen underpaid by $303.7 million
105 Foreigners underpaid by $124.5 million
31 High schoolers underpaid by $366.2 million
mosiplatt
July 15, 2011
@Frederic Bush:
It would be interesting to slice the data based on rookie scale contracts. I will look into it for a future post. With the options for rookie scale contracts, it may not be as simple as just running a query for the production of players drafted since 1995 for the first 4 years of their career.
mosiplatt
July 15, 2011
@fricktho:
I think you’re discounting the fact that salary restrictions in the NBA make it easier for younger players to outperform their contracts.
The 157 seniors drafted since 2005 were actually underpaid by $77.8 million for the first two years of their rookie scale contracts.
mosiplatt
July 15, 2011
@Parker Flowers:
Here are the average draft positions for each group of underclassmen, foreign and high school players since 1976…
Foreign players: 30th pick
High school players: 20th pick
Freshmen: 18th pick
Sophomores: 22nd pick
Juniors: 24th pick
wiLQ
July 16, 2011
@mosiplatt
Group with seniors from 2001-11 is still by far the biggest one… so my first two points still apply.
mosiplatt
July 16, 2011
@wiLQ:
I don’t think your first point applies from 2001-11. Seniors produced 1,461 wins and all other players combined to produce 4,557.4 wins.
I’m not sure I understand your second point.
Parker Flowers
July 16, 2011
Sorry ahead of the time for the long multi-faceted comment… I gotta kill time while doing mindless work on a saturday afternoon.
From your data, one could argue that seniors are the MOST underpaid as a win produced by an average senior costs less than any of the other groups…. though, you would have a pretty cheap below average team if it were comprised only of average seniors. Are you saying that the post-lockout data flips the trend in the other direction? If so, maybe it is that seniors are disproportionately represented among the “scrubs” of the league who are making around the veterans’ minimum and not producing. Could the problem be that the NBA overvalues bench warmers and bench warmers normally went to college for 4 years? Is the bulk of the money lost in bad contracts from a few grossly overpaid big contract-players or from many bad small contract players?
I have been pondering this stuff for a while and think it is really hard to answer this “Lebron would have been better if he had gone to college for 4 years” question, especially by comparing players’s salaries from each group. Basically, the LeBron question is unanswerable because you would have to find comparable players who did and didn’t go to college for 4 years and compare their choking frequency (which I have no idea how you could calculate… maybe you could calculate consistency, like the standard deviation of WP48 from game to game). So a statement like “Lebron would have been better if he had gone to college” is one of those armchair commentator statements that probably couldn’t be supported by a data crunching supercomputer, let alone the observations of a casual sports fan. The one thing that could be said with certainty is that Lebron would be (hundreds of?) millions of dollars poorer if he had gone to college for 4 years.
What if you throw out the salary data and make separate linear regressions for each group? You could graph draft position vs. productivity (maybe only over the first few years of their career, in WP48). Then you could see if a high schooler picked at number 10 is almost always better than a senior at number 10. I actually wanted to do this myself, but after a few minutes I realized I couldn’t find any of the data or figure out how to easily merge any of it together. Point being: I really appreciate the work you guys do… if only there were some sort of depository of all your raw spreadsheets…..
wiLQ
July 16, 2011
@mosiplatt
I meant this: “1) […] Whole NBA can’t just decide “hey, let’s sign only High School Players!”.”
I could add that NBA could do it but their productivity as a group would plummet.
> I’m not sure I understand your second point.
“shouldn’t you mention which players usually decide to skip college?”
If a player is good enough to be drafted out of high-school or after first year of college… basketball-wise what’s the point of staying in school for full ride? So group labeled “seniors” pretty much by definition contains only those players who either weren’t good enough to be drafted earlier or had some huge flaws in their game to work on or [in rare cases] cared more about post-playing career than actual playing career.
mosiplatt
July 17, 2011
@wiLQ:
To your 1st point, the issue isn’t choosing to sign seniors over other players. The issue is how much seniors get paid compared to other players. The owners didn’t have to overpay seniors after their rookie scale contracts from 2001-2011 but they did.
To your 2nd point, only 59.7% of the early entry candidates from 2001-10 were drafted. So it’s not necessarily the case that every player that decides to skip college is viewed as a much more talented player than a college senior. 189 early entry candidates went undrafted since 2001 while 305 seniors were picked. I could also offer the counter-argument that if the seniors stayed in school until they thought they were NBA-ready just like the underclassmen. Either way, we’re comparing NBA-ready talent to NBA-ready talent. It’s just one group has been getting paid more for their talent than the other.
mosiplatt
July 17, 2011
@Parker Flowers:
Maybe I will post a spreadsheet with the raw data after it’s all cleaned up. I wouldn’t want to post something with a lot of errors for people to use.
wiLQ
July 20, 2011
@mosiplatt
“To your 1st point, the issue isn’t choosing to sign seniors over other players.”
Wait a second, what’s their alternative? As a GM you have roster spots to fill right now, preferably with someone who could give you not-embarrassing minutes from the bench in case of injuries. IMHO seniors have been chosen because their supply was the highest [by a wide margin] and they were close to their peak from day 1.
If owners stop “overpaying” for seniors either productivity of all other groups will plummet [because many not-ready-to-contribute-right-now underclassmen/foreigners would join] and/or salaries of those other groups will sky-rocket [because of a competition for way smaller pool of players]. That’s why IMHO your analysis missed the point, seniors are not overpaid, they are necessary evil to keep going [at least until NBA creates true minor league!].
“To your 2nd point, only 59.7% of the early entry candidates from 2001-10 were drafted. So it’s not necessarily the case that every player that decides to skip college is viewed as a much more talented player than a college senior”
That’s a good point so I should have said “If a player THINKS he’s good enough…”.
Those who are good usually know about it but obviously some youngsters are wrong about their skills/potential.
mosiplatt
July 20, 2011
@wiLQ:
You seem to be assuming that signing a player & overpaying them are the same thing. They aren’t.