When you find data that suggests conventional wisdom may be false, it's always a good idea to do further statistical analysis rather than jumping to rash conclusions. It's true if you're doing Lean Six Sigma or, if you're like me, even when you're just talking about sports with your friends.
I've already used ANOVA to show that, contrary to common belief, the NFL may actually have the least amount of parity of any major professional sports league. Now I’ll use Minitab Statistical Software to look at the distribution of winning percentages for each franchise to see if there is more data to back up my allegation.
I took the average winning percentage of every NFL, MLB, NHL, and NBA team over the last 15 years and made histograms of the winning percentages for each league. Leagues with more parity will have a narrow histogram with a high bell curve. Leagues with less parity will have a wide histogram with a low bell curve.
Note: Because both teams are awarded points in an NHL overtime game, the average winning percentage for that league was not centered around 0.500. I adjusted the numbers so that it was centered around 0.500. Although this changed the winning percentages of each team, it did not change the distribution of winning percentages, which is what we want to compare.
These statistics give us further evidence that baseball has the most parity and the NFL has the least. The Patriots, Colts, Packers, and Steelers all have better winning percentages over the last 15 years than the Yankees. The Lions, Browns, Texans, Cardinals, Bengals, and Raiders (not pictured) all have worse winning percentages over the last 15 years than the Royals. And look at the Atlanta Braves: They have the second best record in baseball over the last 15 years and haven’t won a single championship. That’s parity!
So the data analysis has told us that baseball has the most amount of parity, hockey has a little less, and football and basketball have the least amount of parity. So where did the conventional wisdom of the NFL having the most parity come from? There has to be some explanation...and there is. So far I’ve simply been looking at data from the regular season. Next, I'll look at what happens in the postseason.