1992 Dream Team vs. 2012 Dream Team

With the 2016 Summer Olympics opening soon, we’re all going to hear plenty about the mythic 1992 Dream Team and how the current generation cannot compare — even the zenith of the Coach K-USA Basketball era, the 2012 Summer Olympics roster. But what would happen if those two teams clashed? Who would win? The answer is complicated, clouded by tall tales and few trustworthy numbers, but the question will never die.

An eye on match-ups

At the first level of analysis, you can look at the appropriate matchups of the two squads. Looking at the two rosters below, there’s a glaring difference — the 1992 team is much larger with two traditional centers, two traditional power forwards, Christian Laettner, and even some larger perimeter players.

A size advantage isn’t key for the Olympics, as the Americans have routinely run through other teams in the last few tournaments with smaller roster. But the smaller 2012 team won’t have the advantage of overwhelming athleticism, which they relied on versus larger, more immobile international opponents in the last Olympics.

I imagine “old-school” purists will note the defensive power of Michael Jordan, Scottie Pippen, and David Robinson (or Patrick Ewing) together, but the team isn’t perfect there. They also had Magic Johnson, Charles Barkley, Chris Mullin, and Larry Bird. And to take advantage of their size, they’d need to slow the pace more, which was not how they destroyed opponents. Both sides have weaknesses, relative to each other. Also, remember that not everyone was in his prime in 1992 — e.g. Larry Bird.

I do want to strongly state that a gut test, or eye-test, for these two teams has a difficult problem to overcome: how much has basketball progressed since then, and, consequently, will the 1992 squad be ever-so-slightly outdated to the extent that they could lose this fictional match-up? I don’t know how LeBron James will fare against Jordan or Pippen, no one does, so let’s not pretend like we know with certainly. I just know that the two teams are fairly evenly matched except for the frontcourt, where the 2012 team could be out-rebounded and out-muscled. But, again, most teams have weaknesses, and the 2012 roster can take advantage of the shorter three-point line significantly more than the original Dream Team.

There’s no solid conclusion here.

Point differential

Since both Team USA squads were undefeated in the Olympics, one can look to another measure of team strength: point differential. The 2012 team clobbered opponents, beating them by an average of 32 points per game. The 73-win Golden State Warriors team, for example, beat opponents by a little under 11 points per game — and that’s from arguably the most successful regular season ever. However, the 1992 team is even more impressive, with a point differential of nearly 44 points per game. They won every game by at least 30 points. The 2012 “Regime Team” is outclassed there.

Raw point differential, of course, is pretty basic, so are we missing any key information? You can adjust for schedule strength, which is pretty important for a tournament because, naturally, teams don’t have identical schedules. Basketball-Reference’s SRS is popular for dealing with this, and it’s available for the Olympic tournaments both for 1992 and 2012. The older team wins here again with an SRS of 43.5 versus 28.0.

You can use a more advanced method too. I calculated similar team ratings but with ridge regression. You come to the same conclusion — the 1992 Dream Team “wins” with a rating of 12.9 compared to 9.2 for the 2012 squad. Why are the ratings so much smaller than the SRS ones, and what’s ridge regression? Both questions are actually important here. Ridge regression penalizes numbers for better predictions, and there are so few games that the penalization is heavy. In fact, using a different but valid penalization technique, the penalization was even more severe. With only a handful of games, we can’t truly state how good every team was. Then there’s the issue of international team strength: the game has expanded in other countries over the years. We need a different method.

Projected rating

You may see a handful of other articles using some advanced individual stats to determine how great certain teams are. The general approach is to use a metric, weigh by the number of minutes everyone plays (or is projected to play), and calculate a weighted mean. With the right metric, that can be surprisingly accurate, but there’s one issue here: Olympic tournament basketball is a different beast. The metric that says one player is a superstar has little bearing on how he’ll actually play in the Olympics — Michael Jordan doesn’t even lead his own team in scoring, Hakeem Olajuwon is a third-string center, and Kobe Bryant becomes a role player. It’s a strange universe.

Thus, you need some adjustments, and I tried a tested one: adjusting for usage like Nathan Walker has done in his NBA win predictions, which have done supremely well. Basically, usage percentage denotes the percentage of time a player has used a basketball possession. When a team throws a bunch of stars together, their usage rates go down because they can’t sum to more than 100% and those stars are then less valuable. Normally, the adjustments are fairly modest, but here they were gargantuan. And there was a surprising result: the 2012 team had the larger usage adjustment.

Source: Basketball-Reference

1992 starters: Magic Johnson, Michael Jordan, Scottie Pippen, Karl Malone, David Robinson

2012 starters: Chris Paul, Kobe Bryant, LeBron James, Kevin Durant, Tyson Chandler

There were still a few issues with this method, of course. I used the total minutes from the entire tournament, which was all that was available, so those ratings are not accurate reflections of how good those teams could be when needed. Since both squads were ridiculously deep with a similar college star at the end of the benches, it may not be a significant issue, but you can also compare two likely and strong starting lineups. The Dream Team’s lineup was more malleable, so it depends on which star you choose — Patrick Ewing downgraded their BPM compared to Robinson, for example — but in many iterations, they at least have a slim lead there.

Another issue I see with many other articles is that people use career stats to show impressive the original Dream Team was, which has no bearing on how good they were at a specific time. Accordingly, the projected rating stats above are from the two adjacent NBA seasons to the summer games.

Final adjustments

There’s a fallacy here, however, in comparing the stats. The baselines are assumed to be the same for both eras — basically, that the league has the same strength despite being 20 years apart, and the competition is assumed to be the same as well. But that’s all unlikely. There have been a few attempts at quantifying league strength, year-by-year. Here’s one, for example, which shows that the competition in 1992 was slightly tougher than 2012. From what I’ve seen, the results are fairly similar from study to study.

Unfortunately, I think the methodologies have a fatal error: there are very few true absolute measures. For instance, if field goal percentage falls from one decade to another, is that because of better defense or worse offense? But whether you want to use a mental adjustment or a quantitative one, it’s clear league strength should be considered.

Ultimately, I don’t have the answer here, and no one should. This is a basketball fantasy, but the quest for the answer leads to some enlightenment about the game.

For more Olympics news, check out our hub page. For more NBA coverage and analysis, visit the FanSided NBA hub page.