Debunking “The Superstar Theory”

On Realgm.com, a writer Elrod Enchilada posted a well-researched, multi-part article about “The Superstar Theory”. Popular online, this theory shows the domination of superstar-led teams winning titles in the NBA’s 58 year history. The idea thus that if a team wants a title, they need a superstar, or they’re hoping for an aberration.

While the Superstar Theory looks good and has merit, I have reasons to retort it

My first argument against it, is that the champions in the 1950s and 1960s are not relevant evidence. With 8 teams in 1960 growing to 12-13 by 1968 and 1969, the probability of a team having a superstar like Bill Russell, Wilt Chamberlain, Jerry West, Oscar Robertson, Elgin Baylor, Bob Pettit, Walt Bellamy, etc. is much higher. The ratio of superstars to teams in the 60s is similar to the ratio of all-stars to teams in modern day and that all the champions had superstars, is probably no more meaningful than every modern champion having an all-star talent. This is before considering the other time gap differences between the 50s and 60s now – such as the vastly higher pace of the games, spacing-less games because of the lack of a 3 point shot/perimeter skill and no 3 in the key rule and general strategic deficiencies. I’d argue the 50s and 60s titles should be thrown out entirely of this research. It’s a different league.

In the 70s the league expanded, which along with the ABA diluted the talent level in the league and made it easier for the best teams to dominate. Despite this, the 70s is by far the best decade for ‘ensemble’ teams in NBA history. The 1978 Washington Bullets led by Wes Unseld, Elvin Hayes, Bob Dandridge and 1979 Seattle Supersonics led by Gus Williams, Dennis Johnson, Jack Sikma are held up as premium examples of ensemble champions. The New York Knicks in 1970 and 1973 and Boston Celtics in 1974 and 1976 are also within the ballpark of ensemble teams. While Walt Frazier, Willis Reed, John Havlicek, Dave Cowens along with the Bullets and Sonics players are stars, there is a line in the sand between those stars and others in the era like Kareem Abdul-Jabbar and Jerry West. The comparison would be if Gary Payton’s Supersonics won 1996, or one of Jason Kidd’s Nets/Chris Webber’s Kings won in 2002. Teams with stars, but they wouldn’t fit the model of generational stars that most other champions have had the last 34 years.

Between the smaller amount of teams, changes in the rules and style of play and success of ensemble teams in the 1970s, the Superstar Theory using seasons before 1980 isn’t necessary. And this doesn’t hurt the theory, since most like pointing out that since 1980, 33 of the 34 champions have had one of these players: Kareem Abdul-Jabbar, Magic Johnson, Larry Bird, Moses Malone, Isiah Thomas, Michael Jordan, Hakeem Olajuwon, Shaquille O’Neal, Kobe Bryant, Tim Duncan, Dwyane Wade, Lebron James, Kevin Garnett, Dirk Nowitzki. 33 of 34 is hard to get past, right?

To start with disputing this, I’d throw some doubt towards one of those names: Isiah Thomas. Isiah at one point of his career, fit the profile of a mega-star. From 1984 to 1987, Thomas averaged at least 20 points and 10 assists a game, finished in the top 10 in MVP voting each year and had 3 1st team All-NBA appearances, 1 2nd-team. However when the Pistons won their titles in 1989 and 1990, Isiah was neither the same player in accolades or statistics. In ’89 Isiah averaged 18.2 points, 8.3 assists, 3.7 TOVs, .528 TS%, 106 ORTG, in ’90 he averaged 18.4 points, 9.4 assists, 4.0 TOVs, .501 TS%, 104 ORTG. Neither exemplary lines. He made the all-star game both seasons, but didn’t make an All-NBA team and jarringly, finished 17th and 13th in MVP voting both years despite the Pistons winning 63 and 59 games those seasons. Being the best player on a dominant team is usually a free pass to the top 5 in MVP voting – which is how players like Jermaine O’Neal, Peja Stojakovic, and Chauncey Billups have finished top 5 in MVP voting in seasons the last decade. In either statistics or accolades, there’s little reason to believe the Pistons are less of an ensemble team by 1989 and 1990 than the 2004 Pistons, or 1979 Supersonics and 1978 Bullets. A team with stars, but not a superstar like the rest of those 33.

Adding those 2 title teams, makes the odds a little more fair. That makes 3 of the last 34 as ensemble teams. Or if one takes a more favorable selection by counting 78 to 04, that’s a healthy 5 in 27 years.

However, that’s not the only reason to doubt the 80s and 90s as evidence. An important distinction is looking at exactly what built the Lakers, Celtics and Bulls dynasties in the 80s and 90s, combining for 14 titles in 19 seasons between 1980 and 1998. After the Jazz signed Gail Goodrich in 1976, they were forced to give the Lakers a future unprotected pick as free agent compensation. That’s how the Lakers had the #1 overall pick in 1979 to take Magic Johnson, despite having a good 47 W, 2nd round team built around Kareem Abdul-Jabbar the previous season. In 1980 the Cavaliers owner Ted Stepien began a binge of trading future 1sts, to the point where the league had to create a rule named after Stepien to prevent trading consecutive 1sts. One of those picks went to the Lakers, which in 1982 ended up #1 overall, allowing the Lakers to take James Worthy. In 1979 the Celtics traded Bob McAdoo to the Pistons for an unprotected future 1st, which allowed the Celtics to pick #1 overall in 1980 despite winning over 60 Gs Larry Bird’s rookie season. They infamously dealt this #1 pick to the Warriors for #3 (Kevin McHale) and Robert Parish. In 1987 the Bulls’ average season gave them the 10th overall pick, but they also had 8th, Denver’s draft pick, which the Knicks had originally acquired and then dealt to the Bulls. The Bulls took Horace Grant 10th, then moved from 8 to 5th to draft Scottie Pippen.

Obviously in modern day, teams with superstars like Kareem, Magic and Bird having the #1 overall pick, is almost impossible to replicate as a strategy. Teams value draft picks for more, draft picks are protected and free agent compensation is defunct. How the Bulls got the 8th pick is more replicable, but it’s likely either Denver or New York protects that pick in modern day.

Another major difference is the CBA. The introduction of the rookie salary scale and max contracts, among other differences, changed the strategical environment for NBA teams. The last 15-20 years is also when the league stylistically started to resemble modern conditions.

While an argument can be made for allowing the Magic, Bird and Jordan era into the equation, between antiquated way those teams were built and the growth of modern strategy and CBA, I see it more reasonable to use the 15 seasons after the Jordan era as the real comparison for the environment teams have to build a champion now.

Now, the Superstar Theory still holds up OK the last 15 seasons. 14 of the last 15 champions have had Shaquille O’Neal, Tim Duncan, Kobe Bryant, Lebron James, Dwayne Wade, Kevin Garnett or Dirk Nowitzki, true all-time greats. The main retort against this is sample size. 15 years, or even 30 years, is well within reach of what could be a hot streak for superstars, not a true trend. While it’s not totally fair to compare it to 15-30 ABs for an MLB player or 15-30 FGAs for an NBA player, that’s an example of a statistical selection that’s not reflective of a player’s ability.

Furthermore, the last 15 years have had a number of extremely close calls for ensemble teams. Consider the 2000 Blazers, who had an all-time collapse against the Lakers in the Western Conference Finals, preventing an ensemble Blazers-Pacers finals. The 2002 Kings had a heartbreaking loss to the Lakers and were the victim of an acknowledged crooked game a la the Tim Donaghy scandal. It could be argued that Webber or Kidd qualify for a star led title team, but like some teams in the 70s, there’s a line in the sand between them and a Shaq or Lebron. The 2005 Pistons’ bid to repeat ended in G7 of the Finals against the Spurs. The 2010 Celtics had a 3-2 series lead and were up in the 3rd quarter of Game 7 of the Finals and lost to the Lakers. Of course, we just saw the 2013 Spurs lost after the trophy was rolled out in Game 6 for them. Even if 2 or 3 of those 5 teams had won the title, there’d be 3 or 4 ‘ensemble’ teams in 15 seasons since the Jordan era ended. That works out to a 1 in 3.75-4 chance, far friendlier than 1 in 15. While superstar teams would still have an advantage, that’s a healthy enough ratio that we wouldn’t be talking about how hopeless it is for teams without a superstar. One can make the argument that superstars are responsible for why 2 or 3 of those close calls teams didn’t win, but the point is that there’s reasonable doubt. It’s not that it’s a guarantee that of the next 15 titles, 3 or 4 will be ensemble teams, to dispute the “Superstar Theory” only requires a chance. Since it was within reason for 3 or 4 out of the last 15 titles to have gone to ensemble teams, it’s within reason for 3 or 4 the next 15 too. On the outskirts, it’s within reason 6 or 7 do, even if unlikely. This is especially true considering the league is always changing. The CBA is intended to increase parity, with the NFL as a desired model. The development of advanced metrics teams is changing the environment of the league. In NBA history previous generation’s versions of the Memphis Grizzlies fell short at the hands of superstar teams, but John Hollinger’s response to that may be that those teams didn’t have a strategy resembling his, or perhaps they’d have broken through.

There’s no question that the game is tilted towards superstars and the best players winning titles. However, the conclusion that it’s hopeless for everyone else, is founded on a stretch the last 15 years of superstar teams dominating, when it may have been simply bad luck for ensemble teams more than anything else. When considering this and that the league is consistently moving in new directions that will favor different types of franchises, I don’t find The Superstar Theory to be conclusive evidence.