As we near the 2014 FIBA World Cup, it’s time to look back at what team USA has done in recent years.
With only a handful of games and the fact that teams are split into groups before the single game elimination tournament, it’s a bit difficult to ascertain how strong certain teams have been. And it’s important to note which teams were the best and for which reasons. Thus, with a method to solve for a team strength rating based on point differential and a proxy for estimated value via a player’s impact in the NBA, we can quantify how strong every team has been in FIBA tournaments and compare team USA to how good they should have been based on the available talent.
Methodology
Using Microsoft Excel’s solver, you can calculate team ratings by minimizing the sum of squared errors. This method, numerical analysis, is used when you don’t have a direct computation. The errors are simply the difference between the point differential for a single game and the estimated point differential given the strength of the two teams. For example, if team USA is +20 and Puerto Rico is +5, but USA only beats them by 3 points, then the error is 3 – (20-5) = -12 points.
(For more technical notes, I set the initial values as the raw, unadjusted point differential for every team. And I constrain the average of all the team ratings to 0. This is using the GRG nonlinear solver, so be careful with your initial estimates. The solver can only find *local* values, not the “real” global values.)
I don’t know of any other similar system available besides one from basketball-reference, but they do not center the team ratings around 0.
Team Strength: 2002 to 2012
Let’s start in 2002, working forwards, and see what we can learn. (I’m focusing on 2002 to 2012 for a couple specific reasons: I want to use a good player metric from NBA seasons and real plus-minus, aka xRAPM, only goes back to 2001, and international competition significantly increased from the first Dream Team to the 00’s.)
2002: This year, the 2002 FIBA World Championship, is the origin of the Redeem Teeam model. The US lost for the first time with professional players in a tournament. In the last game before single game elimination stage, Argentina, led by their golden generation of Manu, Scola, et al, beat them by seven points. Team USA’s preceding games in 2002 were all blow-outs in their favor, but that included only one strong team, Germany with a young Dirk. The halo disappeared: they were mortal. The U.S. subsequently lost in the first round in the elimination stage to Yugoslavia, the last year the former country was intact in a basketball tournament. After defeating a tough Puerto Rico team, they lost to Spain in the loser’s bracket to claim a shocking sixth place. New Zealand placed higher, and Argentina took the gold medal.
However, by adjusted point differential (team ratings), the U.S. was still a very strong team and ran into some bad luck and some poor timing. They lost their games by an average of 5.3 points, and the point margin was never less than 10 in their wins. Consequently, they have the highest team rating in 2002. Yes, the top team will not always win a single game tournament: NCAA fans know this.
In fact, look at the circumstances with the U.S. and New Zealand. The kiwis were lucky to play Puerto Rico in the first round and won by a mere two points. The U.S., unfortunately, ran into the Yugoslavia powerhouse and lost by three points in the first round. New Zealand then lost to Yugoslavia in round two by 11 points and were blown out in the bronze medal game to Germany. Since they had the lucky placement against Puerto Rico and squeaked by with a win, they tripped into fourth place. The U.S., meanwhile, played New Zealand’s opponents better and the only game against one another was a ridiculous blow-out of 48 points in the U.S.’ favor. Team USA lost its dominance and a handful of other teams caught up, but they were still an excellent team and arguably the best.
In a poetic touch, the sixth place showing happened on US soil in America’s basketball heartland, Indiana, and shocked the media, setting the stage for the redemption.
| Team | Rating |
|---|---|
| United States | 21.3 |
| Yugoslavia | 19.9 |
| Argentina | 19.1 |
| Spain | 17.3 |
| Germany | 13.2 |
| Puerto Rico | 5.1 |
| Brazil | 1.2 |
| Turkey | 1.0 |
| Russia | -0.3 |
| New Zealand | -3.5 |
| Angola | -9.2 |
| Canada | -9.6 |
| China | -11.1 |
| Venezuela | -13.1 |
| Algeria | -20.3 |
| Lebanon | -30.9 |
2004: The 2002 team is typically cited as the worst incarnation, but the 2004 Olympics team was undeniably worse. Tim Duncan was at his peak, but his fellow superstars failed to show up and FIBA officials were unkind to him, sending him into early foul trouble often. The team was haphazardly constructed and the designers failed to adjust to international basketball, bringing in highly marketed slashers in Allen Iverson and Stephon Marbury, and surrounding them with a horde of untested kids who were clearly not prepared for international basketball. They had no shooters, too many athletic slashers, no good distributors, and they tried to rely on Duncan posting up. It failed, and in retrospect it’s not surprising, but what’s shocking is how bad they were.
At least the 2002 team fought the other elite teams to a standstill and destroyed weaker teams. This was a strange tournament where no one dominated, but Spain and Argentina appeared to be significantly better. The U.S. was thrown into the mix with the mid-tier strong clubs like Italy and Lithuania. But to a neutral observer, Argentina was the international powerhouse of the mid-00’s, having repeated as gold medalists with strong point differential ratings in both years.
Americans don’t care about the basketball world championships, but they do care about the Olympics. This loss couldn’t be written off. Set among the historic Athens scenery in Greece, this was meaningful. Something happened.
| Team | Rating |
|---|---|
| Spain | 7.9 |
| Argentina | 7.7 |
| Greece | 6.5 |
| Italy | 5.9 |
| United States | 5.9 |
| Lithuania | 5.8 |
| Serbia and Montenegro | -0.2 |
| Puerto Rico | -2.1 |
| Australia | -2.8 |
| New Zealand | -4.2 |
| China | -12.0 |
| Angola | -18.3 |
*Note: since there were fewer teams in the Olympics, the ratings are deflated relative to the world championships. Thus, a +10 rating in the Olympic is more impressive than a +10 rating in the world championships. Here’s a simple conversion: 6.2 points going from 12 teams to 24, and 2.1 points going from 12 to 16.
2006: With the embarrassment on the world’s biggest sports stage, the U.S. took international competition seriously, and installed continuity and accountability in its system. At first, it was working, as they destroyed the weaker teams and handled the stronger ones reasonably well, with the exception of a pesky, hot-shooting Italian team who kept the game close for far too long. Unfortunately, the U.S. couldn’t stop Greece, who outshot the US from inside and out. There were severe problems trying to stop Sofoklis “Baby Shaq” Schortsanitis, a 6′ 10″, 350 lb behemoth underneath the basket. Team USA, however, recovered and beat a very strong Argentina squad to claim a bronze medal.
Despite the loss to Greece, the U.S. finished with a strong 22.8 rating, but they were outclassed by a peaking Spanish team. Spain actually had a higher raw point differential in group play (pre-single game elimination) and overall. The U.S. was indeed mortal, even with better preparation and participation. Spain, meanwhile, had crested with its own golden generation built around Pau Gasol and a cahche of high-level talent suited to FIBA from Calderon to Garbajosa to Rudy Fernandez.
| Team | Rating |
|---|---|
| Spain | 23.7 |
| United States | 22.8 |
| Argentina | 18.8 |
| Greece | 12.8 |
| Lithuania | 8.9 |
| Serbia and Montenegro | 8.9 |
| Turkey | 5.2 |
| Italy | 4.8 |
| Angola | 4.5 |
| Brazil | 2.5 |
| France | 2.5 |
| Germany | 2.1 |
| Australia | 1.7 |
| Slovenia | 1.2 |
| Puerto Rico | 0.1 |
| Nigeria | -3.9 |
| China | -5.8 |
| New Zealand | -8.0 |
| Japan | -13.5 |
| Senegal | -14.6 |
| Venezuela | -16.1 |
| Lebanon | -16.8 |
| Panama | -18.8 |
| Qatar | -23.0 |
2008: In what is perhaps a meditation on the power of inertia in group settings, it took a long time for the US to react, adjust, and adapt to their fallen status. Team USA brought in its big guns, procuring the services of 4 of the top 5 players by MVP votes in 2008 — Kobe Bryant, Chris Paul, LeBron James, and Dwight Howard — and five out of the five top players by MVP votes in 2009 — LeBron, Kobe, Dwyane Wade, Howard, and Chris Paul. They were joined by other high level-players Chris Bosh, Carmelo Anthony, and Deron Williams along with a veteran distributor in Jason Kidd and two role players with the sharp-shooting Michael Redd and the defender Tayshaun Prince, with Carlos Boozer as the backup center.
The Redeem Team wasn’t all about star power either. It was finally a team that made sense in international ball with two pure point guards, more outside shooting, and someone to rebound. As a result, the team finished with an absurd 28.4 rating, which is unfathomable in normal NBA terms. Keep in mind this is with 40 minute games. With a 48 minute game, this rating would approach 34, and we haven’t even made an adjustment for how teams relax with big leads. If you looked at the rankings in the previous seasons, you would have noticed a handful of teams near the +20 mark, but that was only with the wide-open world championship tournaments. With more teams invited, the ratings are inflated. A 21 rating in the 2006 tournament is more like a 15 rating here. The U.S. in 2008 was playing at an exalted level. They never won by less than 11 points and only won by less than 20 points once.
Spain and Argentina both played well, and in a different year one of those two would have been a worthy champion. It was Spain’s first year with Marc Gasol and Ricky Rubio with their core, and Argentina had its golden generation intact. But both were defeated by a properly prepared US team loaded with superstars.
| Team | Rating |
|---|---|
| United States | 28.4 |
| Spain | 10.7 |
| Argentina | 10.4 |
| Greece | 7.7 |
| Australia | 6.3 |
| Lithuania | 4.7 |
| Croatia | 1.6 |
| Russia | -4.1 |
| China | -7.1 |
| Germany | -9.1 |
| Iran | -24.4 |
| Angola | -25.1 |
2010: This forgotten Redeem Team sent to Turkey in between Olympics is the best proxy for the 2014 squad. They were exceedingly young and were led by breakout star Kevin Durant, with a small number of veterans and rising star rookies and sophomores. To strengthen an argument to play for your national team, a few players used the summer as a springboard for a successful season: Derrick Rose had a meteoric rise and won the MVP, Tyson Chandler turned in his best season ever (at the time) and won a title, Stephen Curry improved on a decent rookie season to turn in one of the best shooting performances ever in 2011, and Kevin Love won a full-time starting job and had the first 20-15 season since Moses Malone in 1983.
Team USA was indisputably the best team that summer, but Turkey, surprisingly, had one of the best ratings ever even adjusted for the softer world championship field. The host country lost only once (the gold medal game) and blew out a few teams by huge margins. Turkey has been an underrated basketball player factory, and they sent out an NBA-tested core of Turkoglu, Asik, and Ilyasova.
As a last note, the U.S. was perhaps a bit lucky that summer. Spain was upset in the second round, and Ginobili was absent sending the typically strong Argentina team to a mediocre rating.
| Team | Rating |
|---|---|
| United States | |
| Turkey | |
| Lithuania | |
| Spain | |
| Serbia | |
| Brazil | |
| Greece | |
| Russia | |
| Slovenia | |
| Argentina | |
| Croatia | |
| France | |
| Iran | |
| New Zealand | |
| Puerto Rico | |
| Australia | |
| China | |
| Canada | |
| Ivory Coast | |
| Germany | |
| Tunisia | |
| Lebanon | |
| Angola | |
| Jordan |
2012: Team USA followed its 2008 domination with a sequel. It was a noticeably stronger team than 2010, which is no surprise because Americans value the Olympics more. They brought back the big guns of LeBron, Paul, and Kobe and replaced Howard with a duo of Chandler and Love. Opposing coaches noted the U.S.’ ability to play with five guys on the perimeter made them virtually impossible to guard. What’s remarkable is how much better they were than their competition: a full 22 points over Brazil and silver medalist Spain.
This may have been the strongest US team, and strongest FIBA team in general, since the 90’s. The numbers beat it out. However, the 83 point rout of Nigeria warps the adjusted point differential, but it’s still a sign of dominance. (If they had only beaten Nigeria by 40 points, their strength rating would have been reduced to 25.8.) And we must remind ourselves with humility that one of the greatest teams ever still only beat Lithuania by 5 points and Spain by 7 points. To expect perfection leads to inevitable disappointment.
Brazil, surprisingly, was the second strongest team even though they lost in the first round. They lost only once in group play, and it was by one point to a strong Russian team. They beat Spain in their only match-up. With a Nene-Splitter-Varejao frontcourt, they remain a contender as a medalist in 2014.
| Team | Rating |
|---|---|
| United States | 30.1 |
| Brazil | 8.0 |
| Spain | 7.5 |
| Russia | 6.7 |
| Australia | 5.4 |
| Argentina | 4.5 |
| France | -0.7 |
| Lithuania | -1.4 |
| Great Britain | -3.5 |
| Tunisia | -15.8 |
| Nigeria | -20.3 |
| China | -20.4 |
From the Parts to the Whole: Estimating Team Strength
One of the most interesting discussions about Team USA is how star power is translated to a team setting. Guys who are used to be the leading scorers are thrust into secondary roles, and other high scorers become mere role players. There’s also a discussion of fit and players who are suited to the international game.
One straightforward way to estimate team strength is to calculate it from individual player ratings using a metric. The metric I used here is real plus/minus (or xRAPM), as it’s the best publicly tested metric that mixes box score stats and plus-minus to measure impact. (I weigh the metrics by how much a player played, and I used an average player rating from the two adjacent NBA seasons.)
The two greatest teams by this measure were the 2008 and 2012 squads, of course. The non-Olympic teams were not close at all. In the table below, I’ve included the estimate from real plus-minus as well as the team rating adjusted for the number of teams in the tournament to put everything in terms of the Olympics. (This was done with regression using the number of teams and the average rating of the top 5 teams since the top 5 teams should be common among every tournament.) I even threw in a couple years from the talented Spain and Argentina teams, which are rough proxies because a few of their players never played in the NBA.
| Team | Year | Rating | Est. strength (RPM) | Difference |
|---|---|---|---|---|
| USA | 2012 | 30.1 | 21.1 | 9.0 |
| USA | 2010 | 20.0 | 11.1 | 8.9 |
| USA | 2008 | 28.4 | 26.2 | 7.5 |
| USA | 2006 | 16.6 | 20.9 | 3.0 |
| USA | 2004 | 5.9 | 13.6 | -6.3 |
| USA | 2002 | 19.2 | 12.2 | 6.8 |
| Spain | 2010 | 10.9 | -1.1 | 12.0 |
| Spain | 2008 | 10.7 | 0.0 | 10.7 |
| Argentina | 2010 | 2.9 | -4.1 | 7.1 |
| Argentina | 2008 | 11.1 | -0.6 | 10.9 |
*Note: the RPM estimate is pace adjusted too. Since there were not complete stats available for every year, I used the average pace of Team USA in 2006, 2008, and 2010: 80.
One observation is that the death knell of the 2004 team, and the pre-Redeem teams in general, was not a lack of talent, per se. Their estimated strength from real plus/minus was in fact greater than the young 2010 team and on par with the teams in 2002 and 2006. They just severely underperformed.
Also, given that the average difference between the real plus-minus estimate and the tournament rating is about 5 points, one is tempted to say that a net zero team in the Olympics would be a poor NBA team but not completely over-matched. But international players significantly perform better in these tournaments, muddying the analysis. This is their game, after all.
And that leads to the inescapable question: how good will the 2014 team be? With the final cuts made, that analysis will be available shortly.
