I’m willing to take it at face value that coaching matters, even though hard evidence is sometimes less than compelling. The average man on the street probably would not do as good a job as the average college basketball coach. But we know less about coaching than any other area of sports. We can look at game theory, we can look at schemes, we can look at recruiting rankings: all true. But we can’t capture much that matters: the camaraderie, the psychology, and all the other soft factors. We can’t quantify these things, which leads many to think stats people believe they don’t exist. I’m sure they do exist; I just have no idea how they affect the game.
Because of that knowledge gap, it is difficult to evaluate coaches. Good coaching decisions sometimes don’t work, and we can’t go back and find out what would have happened otherwise. There’s a grating tendency among fans and media that assumes if one plan did not work, another would have. Often, coaches have many options available to them, but all of them will fail. The least bad option is also the best.
Then again, sometimes the bad options work. Think of blackjack. Perhaps you get dealt an 18 and the dealer is showing a nine. Mathematically, holding is the correct play; if you got this hand a million times, you’d almost certainly make more money by holding than by hitting. But lots of times, you hold and lose. When a basketball coach holds and loses, is that a bad coaching decision? And how do we know whether it is?
Even worse, some people will hit on 18 and win. Hitting on 18 is a disaster over the long run, but every once and a while you’ll draw a three. Does a coach that plays a hunch and hits on 18 make a good coaching decision? Can we know? And when can we trust that an innovative “hit on 18” coach is on to something?
Let me step back from the existential ledge for a moment. We have a large enough sample size with Tom Izzo to know that he is a very, very, very good coach. A great coach, a Hall of Fame coach, a legend. I could not have more respect for Tom Izzo, and I trust that he would never do the college basketball coaching equivalent of hitting on 18.
But watching Michigan State the past few years reminded me of a slightly different scenario. Let’s stick with blackjack. You draw a sixteen, and the dealer shows a face card. This is awful; you should hit, but you are likely in trouble either way. The Spartans have been drawing 16s the past few years—injuries at the worst possible moments, elite recruiting by close rivals, some well-regarded recruits not panning out, etc. Before this year, Tom Izzo had hit on 16 and drawn fives. Izzo deserves credit for knowing to hit rather than stand.
Still, you can’t keep drawing fives forever. At some point, the law of averages catches up to you. Dealt enough bad hands, you will bust. That was the 2011 Michigan State season; you draw another hand in the mid-teens, you hit again, and you lose. It has happened to every great head coach in every sport; that it happened to Izzo is no surprise. Yet it certainly felt like a surprise; how many people did you know that had MSU going to the Sweet Sixteen or even the Final Four again this season? Draw enough fives on 16 and you begin to feel like the player can has the power to draw fives—that drawing fives is a skill that can be learned.
How do we know that Izzo made the right moves? Unfortunately, unlike blackjack, we can’t know. The game is just too complicated, and there’s too much we can’t capture. And “hitting” isn’t just a binary decision, but a thousand little choices during the course of a month, or week, or game. At some point, you simply have to trust; coaches don’t make six Final Fours in a dozen years without making the right decisions.
