The Victors

I have simple threshold criteria: the "Casey Stengel" (maybe Dan Campbell"?) and "Bear Bryant" effects. Is a seemingly unremarkable coach sufficiently competent to not fuck up quality players? On the other end, Bryant's aphorism (I think it was Bryant): can a coach "beat you with his players, and beat you with yours?"

A large majority of coaches at higher levels can probably pass the low Stengel bar: The ones who dont usually lose their jobs and even careers eventually. (See Ellerbe, Brian).

Bryant criterion: At high competitive levels, only a few coaches can pass that. At high levels most coaches have already proven themselves as quite good just moving up in the profession. Because the difference between guys at top levels of the profession - say, 1.5 standard deviations vs 3 standard deviations above average -- is small (93 vs 99th percentiles), important edges purely from an edge over the opponent's coaching will usually be minimal or inconsistent.

But there must be some who are more demonstrably better than peers even at the top levels. The problem is it's hard to know who those are, and "because his team wins more" isnt really an explanation, just getting us back to a Brady-Belichick debate. The test of greatness isnt just winning, since Casey Stengel could do that as long as he had DiMaggio/Mantle etc. It is "What more specifically about the unicorn great coach sets him apart from more baseline competent peers?"

I dont think Harbaugh, say, passes that test, while Dusty May might - because even I can identify specific skills for May, but not Harbaugh, that separate the great from other more commonplace good coaches. For instance, May believing multiple centers/big men can play together rather than just clogging the court -- and then knowing how to actually do it, no less in back to back years with totally different players? (On the other hand, it's still a small number of trials...)