Woulda, Coulda, and the Real Story Behind the Redeem Team

(Photo: Håkan Dahlström)

ESPN.com recently offered a somewhat confusing article comparing the 2012 U.S. Men’s Olympic basketball team to the 1992 Dream Team.  The headline of the article – “LeBron: We Would Beat Dream Team” – makes it clear that LeBron James believes the 2012 team would defeat the 1992 Dream Team. 

The first line of the story, though, makes a somewhat different claim: “LeBron James has joined Kobe Bryant in saying that he believes this year’s Team USA Olympic men’s basketball team could beat the 1992 Dream Team.”

And then further in the article, we see…

James’s comments echoed those of Bryant, who two weeks ago made a similar proclamation.

“It would be a tough one, but I think we would pull it out,” Bryant said at a news conference. “People who think we can’t beat that team for one game, they are crazy. To sit there and say we can’t, it’s ludicrous. We can beat them one time.”

Bryant appeared to soften those comments a bit Friday, telling reporters, “I didn’t say we were a better team. But if you think we can’t beat that team one time? Like I’m going to say no, that we’d never beat them.

“They are a better team. The question was ‘Can we beat them?’ Yes we can. Of course we can.”

So are LeBron and Kobe saying the current team would win, as the title of the article suggests?  Or are they saying the current team could win, which seems to be the argument made within the article? 

At Wired.com, Jason Turbow has written an article that helps answer these questions.  Quick preview: The numbers don’t suggest the word should be would.  But if you want to say could… well, we know from the NCAA Tournament that upsets in a single game of basketball definitely happen.  So the 2012 team could – as Kobe argues – win a game against the original Dream Team. 

In the process of comparing the latest Olympic team with the Dream Team of twenty years ago, Turbow touches upon a team that many U.S. fans would like to forget.  The story of how the team from 2004 only won the bronze – and the Redeem Team from 2008 – seems to be somewhat misunderstood. But I think the very same analysis used in the Wired.com article can help shed some light on why the 2004 team failed so miserably and the 2008 team was so successful. 

Before I get to this analysis, though, let’s discuss another point of view.  In The Gold Standard, the coach of the 2008 team, Mike Krzyzewski, had this to say about the 2004 team’s problems:

“I, too, had watched as the United States had lost its competitive edge in international basketball.  This is not to place blame on those involved with the 2004 Olympic basketball team, a team that was comprised of some of the most talented players and some of the most knowledgeable coaches in the game.  In my perspective, the system failed them.  The team was sent into competition ill-prepared.  It was not a lack of talent or basketball know-how; it was simply a lack of proper time and competition.”

Coach K. went on to note…

”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.”

Coach K. appears to make a compelling argument.  After all, the 2004 team employed such “stars” as Allen Iverson, Stephon Marbury, and Carmelo Anthony.  Surely such “stars” couldn’t have been the problem.

Then again, maybe every “star” in the NBA isn’t equally productive.  As noted at Wired.com, players who score in abundance tend to be considered “stars.” But wins in the NBA – again, as noted at Wired.com — are primarily about shooting efficiency, gaining possession of the ball (i.e. grabbing defensive rebounds and steals), and keeping possession of the ball (i.e. grabbing offensive rebounds and avoiding turnovers).  Just chucking the ball at the basket might help a player increase his scoring totals (and consequently his salary and fan base); but if the shooting isn’t very efficient, then that scoring doesn’t really help a team win. 

We can see this point when we use all the box score statistics the NBA tracks to measure how many wins each player produces. When we take that step, we can develop a very simple model to explain the failure in 2004 and the success observed in 1992, 1996, 2000, and 2008.

Let’s start with the 2004 team.  Before we look at the performance of the players on this team, we must remember that an average NBA player will produce 0.100 wins per 48 minutes.  And as one can see, the 2004 had a few above average players.  But it also employed four NBA veteran players – LeBron James, Amare Stoudemire, Carmelo Anthony, and Allen Iverson – who were below average in 2003-04.  LeBron and Melo were just rookies that season.  And Stoudemire was only in his second season.  Iverson, though, was an NBA veteran who had already demonstrated an inability to hit shots efficiently (a trait that carried over to the Olympics).  

2004 Team

Wins Produced per 48 minutes

(2003-04 NBA season)

Tim Duncan


Carlos Boozer


Shawn Marion


Richard Jefferson


Stephon Marbury


Dwyane Wade


Lamar Odom


Emeka Okafor*


LeBron James


Amare Stoudemire


Carmelo Anthony


Allen Iverson


Average WP48


** – Okafor’s WP48 is from 2004-05, his first NBA season 

The article at Wired.com repeats the same analysis for the 1992 and 2012 teams.  And we can repeat this analysis for the 1996, 2000, and 2008 teams.  When we take this step, an interesting pattern emerges.

Olympic Year


Win Produced per 48 Minutes

(data from NBA Season before games)

Margin of Victory

in Olympic Games
















From 1992 to 2004, the average productivity of players employed – as measured by the average WP48 on the team – declined.  And the average margin of victory in the Olympic games also declined. Then in 2008, the quality of players sent to the games improved.  And suddenly, Team USA was once again trouncing their opponents. 

The correlation between Average WP48 and Margin of Victory in Games is 0.96.   That seems to suggest a very simple model to explain what we see in the Olympics.  The key issue is the quality of players that Team USA is sending to the Olympics.   This doesn’t mean that the effort made to change the system used to prepare the players didn’t have any impact.  And it doesn’t mean that changes in the level of competition didn’t matter.  But it suggests that the 2004 outcome is primarily about the quality of players the U.S. sent to these games.  In other words, contrary to the argument offered by Coach K, I think we can blame the players for what we saw in 2004 (and yes, I am aware that a number of productive players chose not to go in 2004 – so it is not simply the case that people in 2004 chose the wrong players).

Furthermore, rather than attempt to credit the coaching, or system, or team chemistry, I think the evidence strongly suggests the success of the 2008 Redeem Team was mostly about the players who played four years ago.

Given all this, what can we expect in 2012?  As noted, an upset could definitely happen.  So there are no guarantees.  Still, it seems likely – given the players the U.S. has sent to London – that the U.S. would be expected to win the gold medal. And given what we have seen so far, the margin of victory should be similar to what we saw in 2008.  In other words, the U.S. winning by 27 points on Sunday doesn’t seem far from this team’s expected margin of victory in this Olympics. 


I didn't read in your article where is mentions the talent level of the USA team's opponents. There are 29 players who played in the NBA last year on participating country's rosters in 2012. In 1992 I think there was 1. You need to form some sort of algorithm to show how that effect outcomes. You have your fancy Wins Produced per 48 minutes sabermetric stat, but fail to see that the level of competition is significantly greater now. Essentially it was like the 1992 team was playing a bunch of Division 3 basketball players and now, the team is playing a very solid Division 1 team with a few NBA guys mixed in.


There are also 18 former NBA players playing in this year's games


Since the nba players compete against the foreigners in the nba the adjustment is already in there


Well that doesn't make sense. The 1992 players accrued their WP48 stats against NBA players, as did the 2012 team. But the 1992 team then went and played weak competition in the Olympics (1 NBA player, according to a previous poster), and the 2012 team is playing less weak competition.

On a side note, I think the 2012 team would, not could, beat the 1992 team. Sure, the 1992 may have been more dominant in their own era than the current team, but as we all know players get stronger, faster and better as time goes by. I think people greatly underestimate the increased level of physicality among players in the last 20 years. The 1992 team would just not be strong and fast enough to cover players like Lebron, Melo, Durant, etc.

The one mark 2012 has against it is the lack of big men, but that's not because there aren't any in the NBA. If you actually got a roster of the best current NBA players, all healthy, so you're now including Dwight Howard, Derrick Rose and others, they would be even more athletically dominant.



The 'would' or 'could' debate regarding the 2012 team and the 1992 team is a semantics argument.

"But it suggests that the 2004 outcome is primarily about the quality of players the U.S. sent to these games."

No it doesn't. It just suggests that the 2004 team is the least productive team the US sent since 1992... it says nothing about what the primary reason for the outcome. At best it only suggests it could be a reason for the outcome.

If one wanted to suggest the quality of the team was the primary reason for the outcome, one would have to show the quality of the team was less than the quality of its opposition. (ofcourse WP doesn't touch international leagues which I would expect is where a significant portion of the players come from).

I'd also expect some evidence that the 'system' or 'ill preparedness' had a smaller impact than the 'quality' of the team. Something else that isn't offered here.



96% correlation.

If you understand statistics, you'll understand that it's very hard for any other reason to be the primary reason for the outcome in 2004. Yes, the quality of the opposition matters as well, but if the increasing quality of opposition was a big factor, the margin of victory in 2008 should have been substantially down from what it was back when the US was sending dominant teams to the Olympics. It wasn't. The US proceeded to blow out teams just like before when they started sending cream-of-the-crop NBA players again.


"the quality of the opposition matters as well, but if the increasing quality of opposition was a big factor"

why are we only considering the quality of opponent increasing? Why could the quality of opponent not have been better in 2004 than 2008 and 2000? How do we define the quality of the opponent?

where does Berri look consider Coach K's explanation of lack of preparedness? How can we say coach K's explanation is not right by looking at something completely different?

The reality is Berri is only looking at one variable (the quality of the USA team) and drawing a 'primary' conclusion from that. If you understand good science, you'll understand its very hard to draw an accurate conclusion from such limited information.