It Really Is All About the Players

Economists are often asked – and perhaps, just as often just volunteer – to make predictions. This is odd, since – as the old joke goes – economists only seem to exist to make meteorologists look good.  In other words, economists often get their guesses about the future wrong.

Given this tendency, I always like to note when I get a prediction right (and it has actually happened before).  And prior to the Olympics, I did predict that the U.S. would win the gold medal in men’s basketball.  And on Sunday, that prediction came true.

Okay, that wasn’t much of a prediction (did anyone predict that wouldn’t happen?).  And despite the lack of challenge with respect to this prediction, I also heavily qualified my original forecast. Nevertheless, I did make something that could be called a prediction.  And it was right.  So that means something!

Now that the games are over, I wanted to take the opportunity to update the relationship that motivated the original story from two weeks ago.


Olympic Year


Win Produced per 48 Minutes

(data from NBA Season before games)

Margin of Victory

in Olympic Games




















The above table reports the average Wins Produced per 48 minutes for each player on the U.S. Men’s Olympic team the NBA season before the Olympic games.  And it also reports the margin of victory in the Olympic games.   As reported before the 2012 games – and this was actually the point of the original post a few weeks ago — there appears to be a strong link between these two sets of numbers.  From 1992 to 2004, the average NBA performance of the Olympic teams declined.  And the margin of victory in the Olympics also declined.  In the past two Olympics, the productivity of the players sent to the game has increased, and the margin of victory has also increased.

The correlation between these two sets of numbers is 0.96 (about what I found before we saw the 2012 results).  That suggests that the quality of players taken to the games is largely dictating the results.

Again, this result contradicts the story told by Coach K. after the 2008 Olympics. Looking back at the failure in 2004, Coach K. argued the problem wasn’t the players, it was the “system”:

“It was easy to point fingers and blame this guy or that guy for the way he acted or didn’t act in 2004, but Jerry (Colangelo) and I both believed that it was our current system that was flawed, not the players.  This system was no longer conducive to winning.”

A similar point was actually made by Doug Collins during the broadcast of the 2012 gold medal game.  Both coaches seemed to ignore how the quality of players sent to the games changed over time and focus instead on issues of team chemistry, coaching, and systems.

Again, the data suggests that Coach Collins and Coach K. are not exactly right.  It really does look like the key to the success of Team USA is simply the productivity of the players on the team.

With that in mind, let’s take a stab at who can win the gold medal for Team USA in 2016.   In picking this team I am only going to consider:

  • players who will be 30 years old or younger in August of 2016; and
  • players who played at least 1,500 minutes this past year in the NBA

Given these rules, I am not going to consider any player who hasn’t played in the NBA yet.  And I am going to leave off the 2016 team Kobe Bryant (he will be 97 years old in 2016), Tyson Chandler, Deron Williams, Andre Iguodala, and Carmelo Anthony.   Each of these players were on the 2012 team but will be older than 30 in 2016. 

I am also going to break my rules and include LeBron James and Chris Paul (so much for rules!).   Both players will be 31 in 2016.  And although players tend to begin to get noticeably worse in their 30s (and tend to stop improving in their mid-20s), I think James and Paul will still be quite productive at 31. 


Age: August 2016


Wins Produced per 48 minutes

(2011-12 NBA season)

Dwight Howard




Andrew Bynum




Kevin Love


Power Forward


Blake Griffin


Power Forward


LeBron James


Power Forward


Kawhi Leonard


Small Forward


Kevin Durant


Small Forward


James Harden


Shooting Guard


Paul George


Shooting Guard


Chris Paul


Point Guard


Ty Lawson


Point Guard


Mike Conley


Point Guard




Average WP48


Joining this duo will be Kevin Durant, Kevin Love, and James Harden; three more veterans of the 2012 squad.  That means there are seven players who did not play this year.  These include…

  • Dwight Howard and Andrew Bynum at center. Notice that in terms of WP48, these players were not that far apart this past season.  One should note that Howard posted a WP48 of 0.301 in 2010-11 and a mark of 0.293 in 2009-10.  Then again, Howard is also older than Bynum and coming off an injury. So it’s not entirely clear that Howard completely transforms the Lakers (although Steve Nash definitely helps as well). 
  • One rookie from this past year – Kawhi Leonard – posted WP48 marks in 2011-12 that suggests he deserves a look (Kenneth Faried was also very good this past year, but didn’t play 1,500 minutes).  Leonard is not a major scorer.  But a player’s contribution to wins is not just defined by his scoring totals.  And on a team with an abundance of people who like to take shots, it might be a good idea to have at least one player who is happy contributing in other areas.
  • There appears to be a shortage of shooting guards that are both young and productive in the NBA.  Harden was on the team this year.  And George – who also plays small forward – might also be able to play some at the off-guard spot.
  • In contrast, there is an abundance of point guards.  Derrick Rose (who did not play 1,500 minutes this past year), Russell Westbrook, Kyrie Irving, and John Wall are all young point guards who could be considered.  If we look at what players have done in 2011-12, though, Ty Lawson and Mike Conley out-produced all of these players.  Of course, that may not be true in 2015-16.

Again, I am ignoring players who have yet to play.  So this roster should not be seen as a prediction of who will go or even who should go by the time we get to 2016.  What this exercise does show, though, is that it is possible to construct a team that is almost as productive – in terms of WP48 – as the original Dream Team in 1992.  That team had an average WP48 of 0.247.  This team comes close to this mark.

One last note… before these games, there was a bit of controversy as both Kobe and LeBron thought this current team “could” defeat the original Dream Team.  When the games ended, this current edition posted the second largest margin of victory.  When we consider that the international competition has improved in the past 20 years, maybe Kobe and LeBron were on to something. 

Certainly I think it is possible to construct a team that is as productive as the original Dream Team.  Of course, if decision-makers think that chemistry and coaching are really the important issues, that simply may not happen.


"And I am going to leave off the 2016 team Kobe Bryant (he will be 97 years old in 2016)"

He looks great for a 93 year old


"And I am going to leave off the 2016 team Kobe Bryant (he will be 97 years old in 2016)"

It should say "37 years old".


RE: Kobe Bryant's age

I think you meant to say that he would be 37 in 2016, but 97 is pretty funny.

Tim F

Patty Mills scored 26 points on only 20 shots vs. the US. He would never do that in an NBA game. I think coaching/chemistry has some room for improvement.

Tommy Grand

On 26 April 2012, Mills scored 34 points & dished out 12 assists in the Spurs' 107-101 win over Golden State ......

Tim F

Damn you!

I still stand by my point though, he shouldn't be doing that against a team better than an all-star team.

Dave Berri

Just to be clear... saying Kobe is going to be 97 in 2016 was meant to be a joke. For a basketball player, 37 is really very old. There are some exceptions, but in general, at 37 a player tends not to be very good.


Fascinating. I have two questions:

1. I like thinking that the value of a team is their WP48 average in that it gives us a basis for comparing 2012 to 1992, but I'm wondering if it could get even more specific if the team's WP48 was informed by each player's WP48 factoring in the number of minutes they played. The problem with a WP48 number for an entire team with each player's WP48 contributing equally to that number is that not all players play the same number of minutes. Would this yield a more significant figure, or would it basically be the same number?

2. Does Coach K's team-culture-driven take on why 2004 performed poorly relate to the fact that he's a college coach, that there are no trades in college, that players rarely (if ever?) transfer between big D1 programs, and that because of the way NCAA ball is setup each season he's forced to make the best team from whatever players he has, and doesn't make a practice of blaming anything other than programs and team chemistry--the things he controls as coach--on poor performance?



Your story implies that the average level of the competition in the olympics has held constant. I think most everyone would agree that in fact the average level of olympic competition has improved, so the almost perfect relationship between WP48 and scoring margin needs to be re-examined.
(All this also assumes that WP48 is a good and accurate statistic of which I am still not convinced)


Spain is certainly better, but the USA beat Tunisia 110-63 , and Nigeria 156 -73, which represents the vast majority of their winning margin.

Andrew B. Lee

Prof. Berri, it seems to me that everyone simply assumes that international competition has improved. I am not necessarily denying the statement, because it certainly SEEMS that way - but the spirit of Wages of Wins demands that we look at the numbers for the competition as well.

Would you be able to take a look at the quality of production from the other teams both in 1992 and 2012?

Andrew B. Lee

I believe that Arturo said that he built a database for European team production, so perhaps this analysis of Dream Team opponents is now possible?


Why can't it be a both/and? The 2004 was clearly inferior from a talent standpoint from other incarnations of NBA-filled Olympic Teams. But it is still entirely possible that the team would have performed better, possibly even winning the gold, if they had been coached better and/or utilized a better system. They wouldn't have suddenly become a better team than the other ones, but they might well have performed better, which seems to be precisely Coach K and Collins' point.

Additionally, looking at average WP/48 fails to account for the different amounts of playing time of the athletes. LeBron James and Chris Paul each played over 25 minutes a game. James Harden and Tyson Chandler each played under 12 (fewest amount for NBA players). If we through all their WP/48 into the same pot and weighed them equally, we would get a skewed sense of the actual strength of the team as it was structured.



Seems like a false dichotomy.

Perhaps chemistry and the system are what attracts the best players to the team.

Perhaps also chemistry and the system is a euphemism for avoiding inefficient shoot-first players who do not contribute much else, which Wins Produced doesn't value highly and which the 2004 team was built around when Shaq, Garnett, McGrady, and others couldn't be convinced to play and the coaches limited the on court time of LeBron James, Dwyane Wade, and Carmelo Anthony in favor of Allen Iverson and Stephon Marbury.


where's Rondo?


I tend to agree that it's both/and, but Berri makes good use of his W/48 formula here. Interestingly, the average age of the players on the 1992 Dream Team were 29, just approaching Berri's cut-off point for picking players. Only Larry Bird and Magic Johnson were older than 30, though, so I'm sure he would have made exceptions for these two legends.