Clustering in the denominator: non-independence of starts in racing fatality studies

I thought about this in light of Efron’s paper on the Stein paradox which I love because baseball was really what got me interested in statistics. Batting averages and ERA are what we had growing up… its a much richer milieu of statistics now.

stein_whitepaper.pdf (869.9 KB)

steinsparadox.pdf (965.8 KB)