Just a thought I had while reading this – back when Portland had that famous meltdown in game 7 of the Western Conference finals against the Lakers – those games were billed as “Superstar Team vs. Good Players Team”. I can remember a bunch of pundits concluding that having two superstars trumped having a whole bunch of good players on a team. Now, I wonder two things – Were the Lakers really just a 2-horse team(I doubt it) and were there any major producers on that Portland team who weren’t recognized as stars.
Let’s me start my answer with the efficiency differentials – or offensive efficiency minus defensive efficiency – for each team. The Blazers mark was 6.889, which makes this edition of the Blazers the third best Portland team since 1973-74 (the 1990-91 and 1991-92 teams were better).
As for the Lakers, only the 1986-87 team (again since 1973-74) had a higher differential. And the difference between the 86-87 and 99-00 was exceedingly small. The earlier team had a differential of 8.846 while the later team’s differential was 8.841.
So both teams were quite good, although the Lakers were better. When we look at Wins Produced – which simply takes what we know about efficiency differential and uses that to evaluate individual players – we see which players were responsible for the success of each team.
Table One indicates – and this should not surprise – that Shaquille O’Neal and Kobe Bryant led the Lakers. Shaq’s season was easily the best in his career. When we look at the combined output of Shaq and Kobe we see that 64% of this team’s wins are linked to its two stars.
To make our analysis consistent with the discussion of the Pareto Principle, though, we need to look at the percentage of wins produced by the top three on the team. When we look at Shaq, Kobe, and Ron Harper, we see that 74% of wins were produced by the top three players. This means that 26% was produced by the supporting cast. This cast did included three above average performers: Robert Horry, A.C. Green, and Brian Shaw. So although Shaq and Kobe were huge, this was not strictly a two-man team.
Turning to the Blazers, we see the top three on the Blazers – Scottie Pippen, Arvydas Sabonis, Steve Smith – produced 31.4 of the team’s 58.4 Wins Produced. So the top three on the Blazers only accounted for 54% of the team’s wins, while the supporting cast offered 46% of the team’s victories.
Therefore we have to conclude that Ryan’s description is essentially correct. The Blazers were a collection of “good players.” Yes, Pippen and Sabonis – who each posted a WP48 in excess of 0.200 – were better than good. But neither came close to what Shaq offered. Still, the Blazers were still able to field a very good team because it had so many above average producers. Beyond the top three the following players had a WP48 in excess of 0.100 (the average mark): Rasheed Wallace, Detlef Schrempf, Bonzi Wells, Brian Grant, and Stacey Augmon. Plus Damon Stoudamire, Greg Anthony, and Jermaine O’Neal came close to posting a 0.100 mark.
So to answer Ryan’s questions: The Lakers were more than just two players, although clearly Shaq and Kobe were the major stars on this team. And the Blazers had a number of above average performers who were probably not recognized as “good” players at the time.
Now about the bigger picture: Do you want a collection of “good” players or a team dominated by one or two “stars”? The answer is that it’s not about team composition. It really is all about wins.
That being said, in general most wins come from just a few players on your team. So the “star” model is the usual – and probably easier — approach. But the Blazers show that it’s possible to assemble a very good team with many “good” players.
Is this good news for the Wizards (the subject of yesterday’s post)? Not exactly. Remember, the best player on the Wizards is Gilbert Arenas, and his WP48 last year was only 0.180. Had both Pippen and Sabonis offered a WP48 that low, Portland’s Wins Produced would have only been 52.6. Yes, even with all these “good” players, if the top player is below 0.200, you barely clear the 50 win mark.
One last note before I end the post. Yesterday Ryan offered a forecast of the Hornets for 2007-08. According to Ryan – and he was using some goofy stat called Wins Produced – the Hornets can expect to win 50 games this year. This may seem way out there when you look the where the Hornets finished last year, but I thought last season that had Peja Stojakovic and Chris Paul been healthy that the Hornets would have been pretty good. Of course, I tend to think Wins Produced gives you a pretty good assessment of a team’s talent (although it ain’t a crystal ball), so I am a bit biased in my assessment of this forecast.
Our research on the NBA was summarized HERE.
The equation connecting wins to offensive/defensive efficiency is given HERE
Wins Produced and Win Score are discussed in the following posts