[Papa John]

You would have to compare it to the entire universe of coaches, not just the ones who are household names (in part because they've been especially successful).  For what it's worth Duke has outperformed expectations; simulations would show a median number of wins of 72, while Mike Krzyzewski has actually won 86 tournament games (this is almost statistically significant; the probability of seeing 86 or more wins is about 5.5%). 

BC, please forgive me for asking, but how can you claim almost statistical significance in this case? You are not taking random samples from a population, you actually know the entire population of data. So aren't we dealing with descriptive rather than inferential probabilities?

And before you laugh at me, let me warn you that I have a doctorate...in Music (OK, go ahead and laugh).

[Papa John]

You would have to compare it to the entire universe of coaches, not just the ones who are household names (in part because they've been especially successful).  For what it's worth Duke has outperformed expectations; simulations would show a median number of wins of 72, while Mike Krzyzewski has actually won 86 tournament games (this is almost statistically significant; the probability of seeing 86 or more wins is about 5.5%). 

BC, please forgive me for asking, but how can you claim almost statistical significance in this case? You are not taking random samples from a population, you actually know the entire population of data. So aren't we dealing with descriptive rather than inferential probabilities?

And before you laugh at me, let me warn you that I have a doctorate...in Music (OK, go ahead and laugh).