netwins.substack.com

The question of which USA Olympic basketball team was the best is usually answered by reputation. The 1992 Dream Team wins by default because it was the first, because it had Jordan and Magic and Bird on the same bench, because the world had never seen anything like it.

But reputation is not a number. So I ran each roster through the Net Wins formula, using each player’s performance from the NBA season immediately before the Olympics. The 1992 team gets their 1990-91 numbers. The 2004 team gets their 2003-04 numbers. The idea is to capture each player at the specific moment they represented the country, not their career peak or their legacy.

Net Wins is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The results are not what you would expect.

The Method

Net Wins measures how much a player’s statistical contributions converted into actual wins and losses, normalized against what their team was doing that season. A positive number means the player was a net contributor to winning. A negative number means the opposite.

For each Olympic roster, I summed the individual Net Wins of every player in the database from the qualifying season, then calculated a team average. I also include each team’s average winning margin in actual Olympic competition, which is the real-world test of the formula’s predictions.

One methodological note on 1992: Magic Johnson had retired in November 1991 after his HIV diagnosis and played only five games that season. I used his 1990-91 season (+8.01 net wins), which is the standard approach for capturing his contribution. He was the same player in the summer of 1992 that he had been in the spring of 1991.

The Roster Spot Nobody Talks About

Before getting to the numbers, there is one piece of 1992 history worth addressing.

Isiah Thomas was not on the Dream Team. Clyde Drexler was chosen over him for the final professional roster spot on May 12, 1992. Christian Laettner of Duke was added as the token collegiate player.

Thomas in 1990-91 posted +3.36 net wins, a solid season on a 50-win Pistons team. Drexler that same year posted +7.49. By the formula, Drexler was the right choice on merit. But the basketball world at the time knew, and subsequent reporting has confirmed, that Jordan and Pippen and others made clear they would not be comfortable playing with Thomas. The Bad Boys era was recent, the rivalry was real, and Thomas paid for it.

Whether he deserved to be there is a separate conversation. What we do know is that Laettner, who had zero NBA games to his name, was chosen over a two-time champion and future Hall of Famer for reasons that had nothing to do with basketball.

And here is what nobody mentions: the practice sessions from that summer in Monte Carlo were frequently described by coaches and players who witnessed them as some of the best basketball ever played. The Dream Team scrimmaging against itself. Jordan and Magic and Bird and Pippen and Barkley at full speed against each other with nothing to prove to anyone but themselves. Those sessions were never filmed for public broadcast. The world never got to see them. That remains one of the genuine losses in basketball history.

1992 Barcelona, Gold Medal

Season used: 1990-91 | Avg winning margin: +43.8 points

Player

Net Wins

Michael Jordan +9.38

Magic Johnson +8.01

David Robinson +7.54

Clyde Drexler +7.49

Scottie Pippen +6.90

Karl Malone +6.38

John Stockton +6.23

Larry Bird +4.92

Chris Mullin +2.95

Charles Barkley +1.92

Patrick Ewing +0.55

Christian Laettner college

Team total (11 NBA players): +62.3 | Team average: +5.66

Chuck Daly never called a single timeout during the entire tournament. The team averaged 117.3 points per game and won every game by at least 30 points. Daly described traveling with them as “like Elvis and the Beatles put together.” Opposing players asked for autographs before tip-off. Other Olympic athletes mobbed the hotel lobby just to be near them. There was nothing in sports history to compare it to and there has been nothing since.

One detail that surprises people: Charles Barkley was the team’s leading scorer at 18.0 points per game, ahead of Jordan, Magic, and everyone else. The formula had him at +1.92 the prior season, the second-lowest mark on the NBA roster. What it cannot measure is Barkley completely unleashed, playing with no defensive assignment pressure, no coaching system constraints, and twelve Hall of Famers around him making every possession easier. He was the most fun player on the most fun team in sports history and he played like it every night.

The formula sees a roster where eight of eleven players produced more than five net wins the prior season. Bird at +4.92 was already diminished by injury. Barkley at +1.92 was in a down year. Ewing at +0.55 was the weakest link by a significant margin. But the top of the roster was historically concentrated talent.

1996 Atlanta, Gold Medal

Season used: 1995-96 | Avg winning margin: +31.8 points

Player

Net Wins

Scottie Pippen +10.63

Shaquille O’Neal +9.64

David Robinson +7.60

Gary Payton +7.22

Karl Malone +6.47

Anfernee Hardaway +6.45

John Stockton +6.10

Hakeem Olajuwon +3.72

Grant Hill +2.94

Reggie Miller +2.91

Charles Barkley +1.66

Mitch Richmond -0.49

Team total (12 players): +64.9 | Team average: +5.40

Here is where the historical consensus gets complicated.

The 1996 team had a higher team total than 1992 across 12 fully active NBA players. Seven of its twelve posted more than five net wins, compared to five of eleven for 1992. Five players carried over directly from 1992, bringing institutional knowledge of how to play together. Pippen without Jordan’s shadow was the best individual performer at +10.63. Shaq at 24 was approaching the dominance that would define the next decade.

The winning margin dropped from 43.8 to 31.8 points. The common interpretation is that 1996 was a weaker team. The more accurate interpretation is that the rest of the world had watched 1992 on television and spent four years improving. The gap closed not because the Americans were weaker but because the competition had been studying.

A hypothetical 1992 vs 1996 matchup is genuinely interesting. The 1996 team’s depth and collective average would make it competitive with anyone. Most basketball observers would take 1992 on the strength of Jordan alone, and they would probably be right. But the formula says the margin would be smaller than most people expect.

2000 Sydney, Gold Medal

Season used: 1999-00 | Avg winning margin: +9.7 points

Player

Net Wins

Jason Kidd +5.45

Kevin Garnett +5.42

Alonzo Mourning +4.20

Steve Smith +3.92

Antonio McDyess +3.52

Allan Houston +3.25

Gary Payton +2.57

Vince Carter +1.42

Vin Baker +1.36

Tim Hardaway +0.57

Ray Allen +0.12

Shareef Abdur-Rahim -4.27

Team total (12 players): +27.5 | Team average: +2.29

The numbers explain the margins. A 9.7-point average winning margin was the smallest of any gold medal team and included two genuinely close games. The formula predicted a good but not dominant team, and the results confirmed it. Vince Carter’s dunk over Frederic Weis was the defining moment of the tournament, which tells you something about how much the Americans needed individual brilliance to carry games that the 1992 or 1996 teams would have won by 30.

2004 Athens, Bronze Medal

Season used: 2003-04 | Record: 5-3

Player

Net Wins

Tim Duncan +7.38

Richard Jefferson +2.20

Lamar Odom +1.01

Dwyane Wade +0.57

Stephon Marbury +0.47

Carlos Boozer -0.17

Carmelo Anthony -0.25

Shawn Marion -1.05

Amare Stoudemire -1.92

LeBron James -2.04

Allen Iverson -3.59

Emeka Okafor n/a (rookie, college 2003-04)

Team total (11 NBA players): +2.6 | Team average: +0.24

A team average of +0.30 is essentially neutral. Duncan was an island at +7.38. LeBron at 19 in his first NBA season was -2.04. Five of eleven players were negative.

Larry Brown installed Duncan and Allen Iverson as co-captains of the team. That framing is worth sitting with for a moment. Duncan at +7.38 was the most valuable player on the roster by a wide margin. Iverson at -3.59 was the furthest below zero.

A word of honesty about that Iverson number: this formula has a known relationship with his style of play. Iverson was one of the most electrifying players in the history of the sport. He also led the league in missed shots most years, drew enormous defensive attention that created opportunities for teammates the formula cannot fully credit him for, and played his best basketball in a specific offensive system built around his isolation creation. The Net Wins formula sees his high shot volume and turnover rate without seeing what his presence unlocked for others. His true value to a team was always more complex than the raw numbers suggest.

That said, even accounting for the formula’s blind spots on Iverson, a team average of +0.30 is hard to explain away. Five of eleven players were negative. The roster was young, inexperienced in international play, and assembled on short notice without the kind of continuity the 1996 team had built over years. The formula would have predicted the bronze medal before the plane left for Athens. It did not predict the specific path of three losses, but it understood the aggregate talent level clearly enough.

2008 Beijing, Gold Medal

Season used: 2007-08 | Avg winning margin: +27.9 points

Player

Net Wins

Kobe Bryant +6.93

Chris Paul +6.46

Dwight Howard +5.65

Carlos Boozer +5.09

Deron Williams +4.91

Tayshaun Prince +3.85

LeBron James +2.95

Carmelo Anthony +2.86

Jason Kidd +2.77

Chris Bosh +1.42

Michael Redd -3.11

Dwyane Wade -7.02

Team total (12 players): +32.8 | Team average: +2.73

Wade at -7.02 on a season disrupted by injury and Redd at -3.11 dragged the average down considerably. The real core was Kobe, CP3, and Howard. The 27.9-point winning margin reflects a well-prepared team facing a field that still wasn’t at NBA level, not historical dominance. The Redeem Team narrative was real. The formula saw something more modest.

2012 London, Gold Medal

Season used: 2011-12 (lockout-shortened) | Avg winning margin: +32.1 points

Player

Net Wins

Kevin Durant +6.07

LeBron James +5.72

Russell Westbrook +4.94

James Harden +3.69

Kobe Bryant +3.05

Chris Paul +2.85

Tyson Chandler +2.35

Andre Iguodala +1.80

Carmelo Anthony +0.80

Kevin Love -0.04

Anthony Davis -0.47

Deron Williams -1.40

Team total (12 players): +29.3 | Team average: +2.45

These numbers come from a lockout-shortened 66-game season, which compresses totals. Anthony Davis at -0.47 was a 19-year-old rookie in his first NBA season. Durant and LeBron at the top of the roster were both at or near their primes. The winning margin of 32.1 points and the formula’s average of +2.45 track each other closely. The gold medal game against Spain went to seven lead changes and required Durant’s 30 points to close.

2016 Rio, Gold Medal

Season used: 2015-16 | Avg winning margin: approx +20 points

Player

Net Wins

Klay Thompson +8.80

Draymond Green +8.43

Kevin Durant +5.71

Kyle Lowry +5.37

Kyrie Irving +4.76

Harrison Barnes +4.73

DeMar DeRozan +4.58

DeAndre Jordan +4.23

Paul George +1.52

Jimmy Butler +1.23

DeMarcus Cousins -0.68

Carmelo Anthony -2.11

Team total (12 players): +46.6 | Team average: +3.88

The 2016 team is the most underrated roster in this analysis. Thompson and Green came off the 73-win Warriors season at +8.80 and +8.43 respectively. The depth from positions three through eight is better than any other team in the database. The team average of +3.88 is second only to 1992 and 1996. And yet because the roster didn’t have a single narrative player, because it was built on Warriors depth and complementary excellence rather than one dominant star, it gets undervalued in the historical conversation.

The lower winning margin reflects how much better the world had become by 2016. Against 1992 competition, this team might have won by 35 points a game.

2020 Tokyo, Gold Medal (held 2021)

Season used: 2020-21 | Record: 5-1

Player

Net Wins

Devin Booker +4.77

Kevin Durant +3.10

Khris Middleton +3.08

Bam Adebayo +2.88

Jrue Holiday +2.68

Damian Lillard +2.61

Draymond Green +1.45

Jayson Tatum +0.63

Keldon Johnson +0.17

Zach LaVine +0.06

JaVale McGee -0.36

Jerami Grant -2.77

Team total (12 players): +18.3 | Team average: +1.53

The formula, with an average of +1.53, would have flagged this as a team where a single bad game was possible. France caught them in the group stage. Jrue Holiday’s defensive brilliance in the semifinal rematch saved the tournament. The formula can measure what he produced statistically in the regular season. It cannot fully account for what he did to Evan Fournier’s shooting percentage in a single elimination game in Tokyo.

2024 Paris, Gold Medal

Season used: 2023-24 | Avg winning margin: approx +16 points

Player

Net Wins

Jayson Tatum +7.92

Anthony Edwards +5.29

Derrick White +4.85

Joel Embiid +4.17

Anthony Davis +4.17

Tyrese Haliburton +4.13

LeBron James +4.00

Kevin Durant +3.82

Jrue Holiday +3.75

Bam Adebayo +3.30

Devin Booker +2.57

Stephen Curry +2.55

Team total (12 players): +50.5 | Team average: +4.21

The 2024 team has the third-highest average of any roster in this analysis and is the deepest gold medal team outside of 1992 and 1996. Nobody was a drag on the average. Twelve players, all positive.

One number worth sitting with: Jaylen Brown posted +7.00 net wins in 2023-24, the same season he won Finals MVP. He was left off this roster entirely. Had he replaced any of the lower contributors, the team average climbs to roughly +4.58. The three Celtics teammates on this roster, Tatum, Holiday, and White, all made the team. The one who was named the best player in the NBA Finals did not.

The defining moment of the tournament had nothing to do with the formula. France in the gold medal game, on home turf in Paris, with the crowd louder than anything the Americans had heard all summer. Stephen Curry, who had been quiet for stretches of the tournament, caught fire in the second half and buried three after three in front of a stunned home crowd. It was one of the great individual performances in Olympic basketball history and it happened on the biggest possible stage against the most hostile possible crowd.

Tatum at +7.92 was the highest individual mark on the roster coming in. He played almost no meaningful minutes in the knockout rounds. He arrived in Paris as a fresh NBA champion, having just led the Boston Celtics to the 2024 title weeks earlier alongside Jaylen Brown, who was controversially left off the Olympic roster entirely despite being the NBA Finals MVP. Steve Kerr, whose Golden State Warriors had defeated those same Celtics in the 2022 Finals, effectively benched Tatum as the tournament progressed. Whether that history played any role in the decision, or whether it was purely tactical, is a question only Kerr can answer. What the formula can say is that the player who entered Paris with the best individual numbers on the roster spent the gold medal game on the bench while his team won without him. That tension went largely unspoken during the celebration but it was there.

The 1992 team wins. But 1996 is closer than its reputation suggests, with a higher team total and comparable depth. The 2004 bronze was foreseeable. The 2016 team deserves more credit than it gets. And the winning margins tell a story of their own: as the Americans declined in the early 2000s, the margins narrowed. As they rebuilt from 2008 onward, the margins recovered. But they never returned to 1992 levels. The world watched Barcelona, went home, and got better.

The scrimmages in Monte Carlo that summer, the Dream Team against itself, were reportedly some of the best basketball ever played. No cameras. No broadcast. Twelve of the greatest players who ever lived competing at full speed with nobody watching.

The formula counts everything it can find. It cannot count what was never recorded.

The full Net Wins database, 295 NBA players and every season from 1946 to 2026, is free at willf123.github.io/nba-net-wins.

Next: Shaquille O’Neal at #6. Subscribe to get it when it drops.

© 2026 Will Fiore. Net Wins formula and all written content are original works. All rights reserved.

Get full access to NET WINS at netwins.substack.com/subscribe

The full Net Wins database, Every season from 1946 to 2026, is free at willf123.github.io/nba-net-wins.

© 2026 Will Fiore. Net Wins formula and all written content are original works. All rights reserved.

Hosted on Acast. See acast.com/privacy for more information.