Sunday, May 23, 2010

NBA Champions: Pareto


Yes, a pareto chart may not be the right way to plot NBA champs over the years because it is generally used to solve problems. Example: if you have a car giving you problems then you can use the pareto to organize thoughts after identifying a main problem. To start you can use an Affinity Diagram. This is where you just list out all problems with the car and group similar issues together.

The Pareto Principle postulates that 80 percent of the trouble comes from 20 percent of the problems. It separates the "vital few" from the trivial many. This principle is remarkably effective.

So, is it fair to look at who is winning the NBA championship as a problem? Probably not. However, it could be symptomatic of other problems the NBA has like the number of teams in the playoffs or the way they organize the playoffs by seed. You could even look at the length of the series in the playoffs. The length is particularly interesting to me because it eliminates "luck". While this may be the fair way to do things, it is not the most exciting. Football, both college and pro is exciting because of "luck". The best team may not always win. If the NFL were to play playoff series then the Colts would probably have about four more titles.

The pareto at the top of this post shows NBA champions. Is it a problem that half off all championships go to just a select few? Debatable and it can be argued. Particularly when you consider that the size of the league has changed. I do believe that it shows that fairness on some level in the NBA is not clear. The remarkably few teams that win multiple titles could be due to a whole host of things. What I see is the 80-20 (20% of teams winning 80% of the time) rule being broken. It is muddled by teams moving and league expansion. However; that said: in the NBA it is more like the 94-6 rule. Two franchises winning 50% of the time. Were one to apply the 80-20 Bradford Distribution law to the 29 teams making the finals you should have 6 teams winning 20% of the time. Laws have been broken here!

Population density, team management, wealth, wealth generated by winning, the teams the NBA advertises most, and yes... even corruption should be considered. Especially when half the league makes the playoffs. It's also important to note that 29 different teams (not franchises) have made the finals with 19 different winners. If you were to stop there then it seems fair, but when you take it a step farther to the frequency of which team wins over the 60 year history it just plain looks like somethings amiss.

Those are things maybe the NBA should look at. In the meantime, it's good to be a Lakers or Celtics fan because the averages show that if you just wait 3 years - your chances of winning a championship are nearly guaranteed.

If you are a one time winner and not one of those teams? Savor it; because it will not happen again in your lifetime. If you are lucky enough to have a dynasty (San Antonio, Chicago)? Savor that as well.

If you are a fan of a team that has never won one? Hang in there chances are you'll get one within the next 60 years.

Here are the teams that have made the finals along with the number of times they have been there. Of all these teams only five won titles in two or more decades. The Celtics have won at least once in every decade. The Lakers took a nap in the 90s. Detroit has won at least one in the last three decades. No other team has won in three different decades.

24:Los Angeles Lakers
20:Boston Celtics
8:New York Knicks
6:Chicago Bulls
6:Philadelphia 76ers
5:Minneapolis Lakers
5:Detroit Pistons
4:San Antonio Spurs
4:St. Louis Hawks
4:Houston Rockets
3:Washington Bullets
3:Seattle SuperSonics
3:Syracuse Nationals
3:Portland Trail Blazers
3:Utah Jazz
2:Milwaukee Bucks
2:Orlando Magic
2:San Francisco Warriors
2:Ft. Wayne Pistons
2:Phoenix Suns
2:New Jersey Nets
1:Cleveland Cavaliers
1:Philadelphia Warriors
1:Miami Heat
1:Baltimore Bullets
1:Golden State Warriors
1:Indiana Pacers
1:Rochester Royals
1:Dallas Mavericks

As with any data, one can draw what ever conclusions they wish. Especially when dealing with statistics. After playing around with NBA finals data I still believe that if you are not a fan of either the Celtics, Lakers or maybe the Pistons you are out of luck. Fans of other teams will have to settle with just getting there once or, if you are lucky, twice a generation.

Suck it Vilfredo Pareto!!

-----------
For a brief comparison: the NFL has had 20 different Super Bowl winners in 44 years. The NBA has 16 year head start on the NFL and only has 19 different teams. This; while the NFL has had a smaller league size (in the past). It takes 5 teams in the NFL to make 50% of the winners... it takes the NBA 2 teams to get 50%.

The NFL's most winning Super Bowl team (Steelers) has won 14% of the time. NBA's Celtics have won 28% of the time in a league that has existed longer with more teams.

Now the MLB in its 105 history? Well... the Yankees have won a remarkable 26% of the World Series. They are one of the teams in the World Series 38% of the time. For another team just to win 27, as the Yanks have (using WS wins to years), it will take over 300 years. That team would be the Cardinals. I don't need to go any farther then that. Baseball's season length reduces variability, which makes it more probable that teams with the best players (i.e. most money), win. While season length is great for looking at statistics, it may be hurting it in terms of the variety of WS winners.

I know that just merely looking at championships doesn't mean everything. Just interesting to look at. Variability is at work here. The NFL has a short season, single elimination playoff and has not been around as long. The NBA is in the middle with a longer season and longer playoff format. The grandaddy, MLB, has been around longer and has an epic season length in terms of games and its playoff format only allows for a small % of teams as well. It has almost all but eliminated "luck".

You win the WS in baseball... you're the best team that year. Almost no doubt. Win in the NFL? Seems like it is more like playing the lottery.

Now, the NBA? The best playoff team wins. It's season length is long enough to tell you which teams are good but the number of teams entering the playoffs diminishes the importance of the regular season and in doing so makes determining the best team before the playoffs start difficult. It's one of the reasons games are played at three quarters speed. The NBA playoffs might as well be another season, heck, even a different sport.

Which brings me back to why I'm writing this long post today. The NBA is a head scratcher in a way. But I know that if one were to treat the fact that Boston and the Lakers winning most of the time as a problem (as the car example) then you put the pareto back into balance at 80-20 and any year going by with out the Yankees winning... well that's almost like betting the odds.

2 comments:

Jim said...

how does ohio state, michigan, and the big 10 fit into the pareto theorem?

Larry W Johnson II said...

heh heh. I posted it.