This morning it was reported at ESPN.com that the Boston Celtics have signed Rajon Rondo to a five year contract, worth at least $55 million. This is not the maximum salary the Celtics could have offered, but it does top the $45 million offer the Celtics were reportedly offering before.
Had Rondo not signed he would have been a restricted free agents when the 2009-10 season ended. In other words, Boston would have still had a chance to re-sign Rondo when the season ended; and at the moment, the Celtics were not bidding against anyone for Rondo’s services. So did the Celtics overpay for a point guard who averages less than 12 points per game?
As is often noted, an NBA player’s impact on wins extends beyond his per game scoring average. And when we look at the entire box score –and covert those numbers into wins – we see that Rondo is a very productive player. Last year he produced 17.2 wins for the Celtics. No player wearing Celtic green – as Table One reports – produced more last season.
Table One: The Boston Celtics in 2008-09
Among point guards, only Chris Paul (28.2 Wins Produced) and Jason Kidd (19.8 Wins Produced) offered more. And in the entire NBA, Rondo ranked 7th in Wins Produced. In sum, last season Rondo was the most productive player on the Celtics, and ranked in the top ten among all NBA players.
Rondo, though, is not going to be paid like a top 10 NBA player. To put his new contract in perspective, here is what some point guards are being paid who produced less than Rondo last year (contract details from the website of Patricia Bender and USA Today):
Chauncey Billups: 5 years at $60 million [12.8 Wins Produced in 2006-07, 10.0 Wins Produced in 2008-09]
Baron Davis: 5 years at $65 million [10.7 Wins Produced in 2007-08, 1.5 Wins Produced in 2008-09]
Tony Parker: 6 years at $66 million [5.3 Wins Produced in 2003-04, 8.5 Wins Produced in 2008-09]
Deron Williams: 4 years at $63.7 million [14.2 Wins Produced in 2007-08, 10.9 Wins Produced in 2008-09]
The production numbers are from the year before each of these players signed his contract (as well as what the player did last year). As one can see, each of these point guards is paid at least as much per season as Boston is reportedly going to pay Rondo. None of these guards, though, produced 17.2 wins in the season before their current contract was signed. In fact, these four guards have never produced that many wins in a single regular season.
So why don’t the Celtics raise their offer? The obvious answer is that player salaries in the NBA are not determined by Wins Produced. The primary determinant of salaries is scoring, and Rondo is not much of a scorer. Last year he only averaged 11.9 points per game; and on a per-minute basis, he was below average for a point guard.
If we examine what Rondo should be paid – according to the logic of the NBA’s free agent market [something we discussed in The Wages of Wins and in a published article in 2008] – then Boston’s original offer was closer to the mark. So by the logic of the NBA’s free agent market, the Celtics are over paying a bit (but not much).
When we look at this contract from both sides, one could argue both the team and player may not be happy with this contract. Boston is probably paying Rondo more than any other team would have offered. And Rondo – if we think about his contribution to wins – is underpaid.
So what can we do to make everyone happy? From the Celtics perspective, Rondo should focus more on scoring. If he averaged about 15 points per game – which he could if he simply took about three more shots per game (or less than one per quarter) – then Rondo’s pay would match what he should get according to the logic of the NBA’s free agent market. Of course, Rondo would still be producing more wins than his would pay would suggest. To put his production of wins in line with his paycheck, perhaps Rondo could reduce his production of wins by committing a few more turnovers or reducing his steals. NBA players aren’t paid for turnovers and steals, so changing these numbers won’t change people’s perceptions of his performance. Changing these numbers, though, will alter Rondo’s production of wins.
Assuming none of these changes occur, I will make the following predictions:
1. Rondo will continue to produce more than he is paid. Because he does not rely on scoring to produce wins, the NBA player market will continue to undervalue Rondo’s services.
2. Given the age of Kevin Garnett, Ray Allen, and Paul Pierce, we can expect the productivity of Rondo’s key teammates to decline in the near future. When this happens – and the Celtics start winning less often – perceptions of Rondo’s pay will change. Specifically, when the Celtics are not one of the top teams in the NBA, someone will argue that the Celtics overpaid for Rondo.
Let me close by noting that The Wages of Wins was published in 2006, and the free agent model reported has been updated since that time. As of 2008 – I have not added data from this past summer yet – the story told in The Wages of Wins still stands. Scoring dominates perceptions of value in the NBA (and you see the same story with respect to other decisions). And as long as this continues – and as long as Rondo maintains his production of wins — the money Rondo receives will not be consistent with his contribution to team wins.
- 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.
26 responses so far ↓
JoseMesaIsDead // November 2, 2009 at 10:47 am
I think this obsession with scoring is changing more rapidly than you’re saying – and I would argue that the contract the Cavs gave Anderson Varejao is a perfect example of that.
Plus, like you wrote when comparing LeBron and Kobe’s games at MSG last year – scoring at a super efficient rate is still the best way to help your team win (using WoW methodology). In other words, the tragedy is when PPG is emphasized over efficient per minute scoring. And then, of course, the usual debate over the value of shot creation and usage rate vs. efficiency unfurls…
PJ // November 2, 2009 at 11:21 am
Yeah, I’m JoseMesa here. We won’t know any time soon, but I think Rondo would have gotten a sizable offer from some team in the offseason — perhaps in the 12/13m per year range. Maybe that’s partly because he scored a lot in the playoffs last year, so he’s now seen as capable of doing that. Or maybe it’s because he’s built a reputation in the league for things other than scoring: assists, rebounds, steals. In his case, the usual WoW story may not entirely apply.
This is probably true, though: “when the Celtics are not one of the top teams in the NBA, someone will argue that the Celtics overpaid for Rondo. ” Here’s hoping (as a fan of both the Celtics and Rondo) that Perkins continues to improve and Ainge keeps adding talent to make up for age-related decline elsewhere on the roster.
Michael // November 2, 2009 at 11:36 am
Great post! There is a small error in the title though.
Michael // November 2, 2009 at 11:41 am
Also, I think it’s possible that as the ‘big three’ decline through age, Rondo may become more involved in the offense. So perhaps he will be able to average a few more ppg this year.
Hopefully if he can do that , and the Celtic’s continue to dominate as they have so far then he could get some all star consideration.
Tball // November 2, 2009 at 12:15 pm
The other side to this, of course, is that he would have been a restricted free agent next summer and not a free agent. As we see every year, offers to RFAs sit for a week or more and then the team chooses to match or let the player go. If you make a $12m/yr offer to a player like Rondo, you wait a week, and the Celtics match, there is no one of that caliber left to sign.
The Pistons made over the top offers knowing they wouldn’t be matched. Dallas made an offer to Gortat and, by the time it was matched, lost out on McDyess, ‘Sheed, and other big men that might have been had at the mid-level figure. Portland lost out on Millsap, but the timing on that deal was easier. David Lee couldn’t get an RFA offer. Childress couldn’t get a worthwhile RFA offer. I don’t remember if Josh Smith saw a legitimate offer. The Celtics could have used this CBA tool, along with the reluctance of GMs to pay for non-scorers, to sign Rondo longterm for less (not to mention, not having to commit until after the season – every game can end a career, especially one carried on such a wiry frame).
As a Celtics fan, I’d like to see the whole roster filled with underpaid, talented players.
Lior // November 2, 2009 at 1:04 pm
I think both the Celtics and Rondo should be happy with the outcome: the Celtics are getting 17.2 wins/season on the cheap, by paying the money to a player who produces wins in a non-scoring way. For Rondo the situation is reversed: the market is undervaluing hit talent, so he is likely to be underpaid no matter what. Unless other teams were willing to pay him based on his WP value, he has no reason to be upset with the Celtics not paying him much more than the other teams.
JAW // November 2, 2009 at 3:49 pm
This seems like a good Coasian bargain to me.
Italian Stallion // November 2, 2009 at 5:14 pm
Excellent entry.
This is more or less am example of “intrinsic value” vs. “relative value” with a twist.
Similar examples can be found looking at the stock market and business acquisitions.
If similar companies in an industry are selling for 15X earnings and someone decides to pay 25x earnings for one, it can still be a bargain if the value of its free cash flow eventually proves it was worth 30x earning. At the same time, one might question paying such a huge premium to the market value of similar companies.
todd2 // November 2, 2009 at 5:23 pm
Most productive player on the squad and the 5th highest salary? Ouch!
Zach K // November 2, 2009 at 9:57 pm
Dear Dr. Berri,
Thank you for singlehandedly convincing the Celtics to sign Rondo for $55 Million. The call you placed to Danny Ainge on Halloween night was crucial. Ainge especially liked how you sang the wins produced formula to his kids to the tune of Silent Night. And how you did not cuss.
Again, thank you.
Sincerely,
Celtics Fans everywhere
kevin // November 3, 2009 at 9:59 am
Dave, I’m wondering if your system overrates Rondo some. While I agree he’s a fine player, he can’t shoot at all. So much so that he’s often left wide open to take all the 15 footers he wants while his more skilled teammates are double-teamed. Yet he often still refuses to shoot, and that means that he is transferring his liabilities to his teammates, who are forced to shoot or pass while doubled up. That makes them look bad and him look good but, in reality, it’s his lack of shooting credibility that creates problems for his teammates on offense and hurts the team.
Italian Stallion // November 3, 2009 at 10:12 am
D. Berri,
Have you ever considered the possibility of weighing offensive and defensive fouls differently?
I never gave this issue much thought, but last night I saw a play that made an impression on me.
A guy committed a really dumb offensive foul at one end. Then on the other end he committed a defensive foul before the team was in the penalty that prevented what was probably going to be an easy bucket by the opposition.
So at one extreme we had a lost possession and perhaps a long term cost of 1 point.
At the other extreme we had a gain of at least a fraction of a point by preventing an almost certain 2 point basket and replacing it with an “average possession” and perhaps 1 long term point.
Granted, unless someone tracked this and made subjective judgments about whether a foul was smart or stupid, we could never get the exact costs and gains.
However, for our purposes it might make some sense to value offensive fouls differently than defensive fouls if the long term values suggest they are different (which they might be).
Tim // November 3, 2009 at 11:44 am
Hi, Professor. I’m just wondering whether you plan to read Bill Simmons new book, The Book of Basketball. I’m reading Chapter One, and I’m already squirming as Simmons goes on and on about how statistics don’t tell us how good the Isiah Thomas-era Pistons or Sean Battier are, because what they did just doesn’t show up in the box score.
I love Simmons’ wit and passion and his ability to recall details (or to spend hours researching details) of games that are for me only vague memories, but I do wonder — has he never read your stuff, or has he read it and deliberately decided to ignore it? He doesn’t even discuss it and reject it, he just ignores it. Maybe Malcolm Gladwell can get you two together some day. Sheesh!
dberri // November 3, 2009 at 11:58 am
Tim,
I do have his book, but haven’t had a chance to read it. Someone this summer wrote a wonderful essay comparing my approach to analysis to the approach offered by Bill Simmons. It was in a blog, but I don’t remember which blog (there is only a few million of these) so I can’t find it. I do remember that it was a wonderful attempt to explain how analysis can be done from very different perspectives.
Russell // November 3, 2009 at 12:18 pm
Assuming that Rondo’s fg% would remain the same if he took three more shots per game is a big assumption.
Michael // November 3, 2009 at 12:42 pm
Not really. His shots per game went up by 3.5 in 07/08 and his shooting percentage actually increased over the previous season. You could argue that he was young then, but he is still young now so…
Italian Stallion // November 3, 2009 at 5:41 pm
A small increase in the number of shots per 36 minutes at a mildly lower success rate doesn’t move the overall needle very much. In many cases you wouldn’t even be able to tell if it was a random fluctuation.
IMO, the impact would be more noticeable if you could look at incremental shots in isolation. That way you could see if the increase was coming with the addition of more difficult shots than before (on average). That might be possible if the location of every shot was tracked.
IMO, another problem with measuring the impact is that a player’s shot count tends to increase as that player improves his shooting range and other offensive skills. That makes it hard to isolate.
Intuitively and from personal experience, I suspect that a few shots either way don’t have a huge impact on efficiency, but at extremes it probably matters.
Tim // November 3, 2009 at 11:53 pm
Hi Professor, I think I found the article you remembered at http://notimetoread.wordpress.com/2009/09/07/44/
The article makes the point that Bill Simmons advocates the same attributes that you do, but Simmons (like many GMs and other professional observers) still gets seduced by the high scorers. On the other hand, Simmons and other writers do a wonderful job of describing the beauty of the game, something that is hard to capture when discussing numbers.
Still, for me the irony is that your theories support Simmons’ instinctive appreciation for team play, yet he ignores your theories and often ignores his own instincts when judging individual players.
Tball // November 4, 2009 at 7:47 am
IS,
Offensive fouls and defensive fouls do have slightly different values in db’s WP metric. Offensive fouls always result in a turnover, so the scorekeeper will award a foul and a TO to the offending player for an offensive foul, but only a foul to the offending player for a defensive foul. WP is sensitive to possession changes.
Russell // November 4, 2009 at 3:00 pm
Michael,
Yes really. They got Garnett in 07. Garnett and Pierce also become more efficient that year.
Italian Stallion // November 4, 2009 at 5:05 pm
Tball,
Thanks.
I didn’t even know that offensive fouls were recorded as turnovers. Doh!
Italian Stallion // November 4, 2009 at 5:18 pm
T-ball,
On second thought, that’s probably not exactly what should happen. If an offensive foul counts as both a personal foul and a turnover, that’s kind of double counting it. It should probably mostly count as just a turnover (with some slight value for the incremental foul that I assume counts toward the team eventually being in the penalty).
One of things that makes most defensive fouls so problematical is that the team gets free throws, which are very high probability shots.
But like I said, sometimes a defensive foul can be a smart play. That’s never true of an offensive foul.
Man of Steele // November 4, 2009 at 6:28 pm
Chiming in on the conversation about Rondo passing up open shots and hurting his teamates – I think what often gets lost in conversations on this blog is the simple fact that passing up a shot you can’t hit is a good choice. So if Rondo passes up a long jumper which he probably would have missed and makes a good pass to a teamate, that is a good play even though it is not aesthetically pleasing. Being a good player doesn’t mean you can do everything, it means you do some things and limit the damage you do in respect to the things you can’t do well.
Also, about Rondo’s teamates being “forced” to take bad shots when double-teamed … I don’t know about that. There don’t seem to be that many shot clock violations in the NBA, so presumably there are not that many times when teams are forced to take a bad shot at the end of the shot clock (What I mean is, if there are few shot clock violations, then presumably most possessions end in a “chosen” shot before the end of the shot clock). If Rondo passes to a heavily guarded teammate and that teammate chooses to take a difficult shot with time still remaining on the shot clock, that’s a bad decision. Having the confidence to take a shot is not a skill. Being able to discern which shots to take and making the shots you do take is what makes one a good shooter.
Westy // November 5, 2009 at 12:57 pm
Should the Celtics have paid more?
Not if Rondo was willing to sign for what he did. We’ll never know whether they were willing to pay more, as understandably, they agreed to pay him a value he agreed to.
6-0 and still not perfect: ways for the Celtics to play even better. « // November 6, 2009 at 2:01 pm
[...] do we want to pass up?). I’m not sure which situation is worse, but both are pretty undesirable. Rondo isn’t the only perpetrator here, so bring those fingers slowly down. We are throwing the ball [...]
A Note on Iverson and a Sunday Notebook » Boston Celtics Basketball – Celtics news, rumors and analysis – CelticsHub.com // November 8, 2009 at 10:27 am
[...] Another of the Web’s stat gurus, David Berri at Wages of Wins, likes the Rondo deal for the C’s and seems to think that it is the rare deal that compensates a non-scorer fairly [...]