In Part II of this series, I developed a method of estimating a team's probability of winning the NBA Championship based on the allocation of their possessions among their top 5 players. The idea is that, assuming 2 teams are championship-caliber, the one who follows the time-tested pattern of Star 1a + Star 1b + 3 role players will be more likely to win a championship. Today, I'm going to apply this to all regular-season teams in NBA history, and see which teams were theoretically built for postseason success, then look at what actually happened to them.

First, we need to define what it means to be a "championship-caliber team". Historically, the average regular-season SRS of all NBA champions is 6.07, and the median SRS is 6.059. Obviously, teams have won with SRS scores of under 6 (the 2006 Heat were the last team to do so), but as a general rule, if you post an SRS of 6 or greater during the regular-season, you have established yourself as either the odds-on favorite or at least one of the leading candidates to win the NBA title, which is what we're going for here.

Now, we need to look at the possession usage for each championship-caliber team's top 5 players by MP during the regular season. Since possessions are impossible to calculate for seasons prior to 1977-78, I'll be approximating them with what's called "Modified Shot Attempts", a possessions-like stat that doesn't include turnovers. I estimated the % of the team's MSA used by each player while on the court, and sorted the teams' top 5 minute-getters by that result. I also forced all teams' "5-man units" to have a combined %MSA of 100%; we didn't have to do this in the playoffs (because players don't change teams mid-postseason), but in the regular-season players do move around, and before 1965 there was no real accounting for split-season stats. Because of this, if no adjustment is made, some early teams will have combined %MSA of 110% or more for their top 5 players, and we need everything to be scaled to 100% for consistency's sake.

Anyway, once we have the top 5 players' %MSA figures, we can use the following equation to estimate their championship probability based on the usage mix of their top players:

p(Championship) ~ 1 / (1 + exp(3.23 - 0.01*MSA#1 - 0.10*MSA#2 + 0.10*MSA#3 + 0.01*MSA#4 - 0.03*MSA#5))

Here are the championship-caliber regular-season teams with the most ideal usage patterns since 1952:

Are the 1965 Celtics underrated in "Greatest Team of All Time" discussions? Yes, they barely outlasted the 76ers en route to the Finals, but Red Auerbach's team construction was nothing short of beautiful: One outstanding offensive go-to guy (Sam Jones) in the backcourt, a pair of solid two-way players (Tom Sanders & Tom Heinsohn) at forward, and two of the best defensive players of all time (Bill Russell, K.C. Jones) rounding out the starting lineup, plus a gifted scorer off the bench in John Havlicek, and another decent all-around reserve in Willie Naulls. The result was 62 wins (the most of the Russell era), one of the top defensive teams ever, and an offense that was constructed to maximize its potential in the playoffs. It was also a team that was more than the sum of its parts, which is what this series has been all about -- achieving great team results by finding the perfect combination of players that "fit" best together.

On the flip side, though, you have to feel bad for the Cleveland Cavaliers, who followed the championship usage formula to a T in 2009 & 2010, but have no rings to show for it. And the Boston Celtics of recent vintage, more balanced now than ever with the emergence of Rajon Rondo, are the antithesis of the formula with the offensive load spread almost evenly across 4 starters, but have sustained success and are 2 wins away from their 2nd Finals berth in 3 years. All of which goes to say that the ideal championship usage pattern isn't set in stone, and it's not a given that if you build a team according to it, you'll automatically win. But it is a nice guideline based on the most successful teams of the past and what it has historically taken to win a title from a team-building perspective.

This entry was posted on Friday, May 21st, 2010 at 11:27 am and is filed under Analysis, Playoffs, Statgeekery. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.