Agreed, although the variance of the data is probably even more important to consider. In other words, is KenPom consistently bad or inconsistently less bad at predicting Michigan games? Looking at the data, and given Michigan's play this year, I imagine it's the latter.
Very nice post, but I'm not sure that dividing the number of medals won by the number of athletes should be represented by a percentage. A percentage is a dimensionless proportion and, as such, the units of the numerator and denominator have to cancel. In your case, the numerator units (medals) and denominator units (athletes) are different. (This metric is fine to calculate and talk about but it should be properly classified as a rate.)
However, as the responses here indicate, rates can be tricky to interpret. For example, if each of the Korean athletes competed in two events, while the Dutch athletes each only competed in one, how does that affect the interpretation of medals/athlete? Is it really accurate to say that the Koreans won more medals per athlete (as a measure of overall performance) if they had more opportunities per athlete?
Instead, we should probably calculate the number of medals won by the total possible medals that could have been won. Of course, that would take considerably more time.
Recent Comments