Game coaching is only a small part of the overall job of a coach. However, coaches are measured by results as most people only watch the games, not practices. The effort between games goes unnoticed because its is hidden from view.
One popular measure for a coach’s effectiveness during games is performance during close games, usually games decided by one possession (<3 points) or maybe games decided by five points or less.
Is this a fair measure of a coach? Would Butler’s Brad Stevens be a better coach if Gordon Hayward’s half-court heave found the bottom of the net in the NCAA National Championship Game?
In my research design class, we studied observable score and true score. Essentially, your observable score equals your true score plus the error score. From a coaching standpoint, the error score accounts for lucky shots, unlucky shots, bad calls, off-nights, etc.
In the Butler example, the observable score was a two-point loss. If a coach is measured by performance in one-possession games, Stevens would be 0-1. Is that fair? If Hayward’s shot would have fallen, would it be fair for Coach K to fall 0-1 based on a half-court heave?
It is hard to know how much error there is in a given game. Is it the coach’s fault when a 90% free throw shooter misses two free throws at the end of a loss? What about a missed call or a banked in three-pointer or a shot that rolls in and out? It’s easy to say that it’s a part of the game, and it is, but should a coach’s worth be determined, at least in part, by these situations when the margin of victory and defeat is so small?
When I coached a professional team, we lost several close games on the road with an under-manned team. Ultimately, I was fired. However, I felt being competitive on the road against more talented teams was a sign of good coaching, not bad coaching. We were in the game and one or two plays or one break away from stealing a win.
What if we considered the error score when evaluating a coach’s performance? Rather than decide a coach’s value based on games decided by three points or less, what if we set the error score at five and measured a coach’s record in games decided by more than five points?
Would a coach’s record in games decide by five points or more more accurately reflect a coach’s value than a coach’s record in close games?
By Brian McCormick
Author, Cross Over: The New Model of Youth Basketball Development
Director of Coaching, Playmakers Basketball Development League