In The Success Equation, Michael J. Mauboussin makes an interesting point about the role of luck in sports. He frames it in a way I hadn’t thought of before. In sports, any win is part skill and part luck. Different sports depend on different ratios of skill to luck.
Any sport that depended entirely on luck would have all teams normally distributed around a .500 winning percentage (for instance, if competitive coin flipping was a sport). If a win were determined entirely by skill, there would be an even distribution across the spectrum of winning percentages–the best player or team would be undefeated, the worst winless, and everyone else distributed throughout based on skill level. Where a sport lies between these two extremes is a measure of how much luck is in play.
Baseball, on this scale, is about one-third luck. Soccer and football are similar, hockey a little more dependent on luck (53%), and basketball far more dependent on skill than luck (12% luck).
The role of luck is dependent in large part on parity. If two players are exactly equal in skill, then by definition luck would determine the outcome. Basketball teams, because they are restricted mostly to very tall people (and the average height has been rising over time), have a much smaller population of potential players, which leads to a higher variance of skill (the larger the pool of potential players, the easier it is to stock teams with players near the upper limit of skill).
Isn’t that interesting?
How can we determine the role of luck in the success of instruction?