Whether we're talking about process improvement or sports, it's always a good idea to see if data analysis supports the ideas you assume to be true. Sometimes the data confirm what you expected to see. But sometimes the data suggest that the "common wisdom" isn't all that wise after all.
For instance, sportscasters talk about “parity” all the time. The NFL is frequently assumed to be the professional league with the most parity. I decided to use data analysis to see if this is true. I'll start by using Minitab’s One-Way ANOVA to see which league has the widest discrepancy between the best and worst team.
We’ll stick with the 4-major American professional sports leagues, the NFL, the NBA, the NHL, and MLB. When parity exists in a league, every team has roughly equivalent levels of talent. The "best" team is not significantly better than the "worst". Let’s compare the difference in winning percentages between the best team and the worst team in each league. The smaller the difference between the best and worst team, the more parity the league has.
I took the final regular season standings from Sports-Reference.com and calculated the difference between the best and worst winning percentages for the last 15 years. Then I used Minitab’s One-Way ANOVA to compare them at a 0.05 significance level.
Because the p-value is lower than our selected significance level of 0.05, we can conclude that at least one sport is different than the others. To determine which sports are different, look at the 95% confidence intervals. None of the confidence intervals overlap, which means there is a statistically significant difference between the averages for all four sports.
Football has the largest difference, with a mean of about 0.75. This means that over the last 15 years, the best NFL team wins an average of 75% more of their games than the worst. So the NFL has the least amount of parity because the best teams have significantly better records than the worst. The NBA is next, with the best team winning an average of 58% more of their games. Hockey comes in third and baseball is last. So by this metric, baseball is the sport with the most parity.
But wait, that can’t be! The Yankees are great and the Pirates are terrible! Where’s the parity? The difference is sample size. The best baseball teams aren’t that much better than the worst teams: they only win an average of 27% more of their games. But by spreading that small advantage over 162 games, it appears that they are much better. For example, in 2010 the Yankees finished 38 games ahead of the Pirates. Sounds like a lot, right? But if we applied their percentages to a 16 game NFL schedule, the Yankees would have finished the year 9-7, while the Pirates would have finished about 6-10. Meanwhile, if the 2010 NFL season had lasted 162 games, the best team (New England Patriots) would have finished 142-20, while the worst team (Carolina Panthers) would have finished 122 games back with a record of 20-142!
You may think, "So what if the NFL has the widest discrepancy between the best and worst team? It's different teams at the top and bottom of the NFL standings each year. Baseball may have a smaller difference between the best and worst teams, but the Yankees are always the best team and the Pirates are always the worst. Nothing like that happens in football!"
So, I can’t look at one statistic and be confident in saying one league has more parity than another. Next, I’ll delve deeper to look at how each franchise has performed over the last 15 seasons.
Photograph by RonAlmog. Images licensed under Creative Commons Attribution ShareAlike 2.0.