Friday, August 10, 2007

Back in the saddle

Finally made it back to the gym yesterday afternoon to play some basketball for the first time since I took the bar exam two weeks ago. You'd think that with the bar exam out of the way I'd be playing ball even more often than I was before. No such luck. Between various trips out of town since the bar exam and the departure from Boston of the people that I'd been playing basketball with this summer, I just hadn't found a good time to go. Even today, I hadn't really expected to find a game. No one was on the court when I arrived, so I did four miles on the treadmill (2.5 running, 1.5 warm-up and cool-down) and then went back to the court to shoot around for a little bit afterwards. Lo and behold, that's when people started showing up.

Why am mentioning my pick-up game from yesterday to you? As I try and figure out which of the various available statistical measures I find most compelling in evaluating players, I've been spending some time thinking through how I would analyze player performance if I could build an evaluation system from scratch. Now, when I'm playing at the gym, I find that I notice when a particular statistic plays a really important role in the outcome of my pick-up games. Today, my team got absolutely hammered on the boards. The other team had a guy who was 6'4", and I was the tallest person on my team at 5'10". Plus, the other four players on the other team were stronger than our players for the most part. Besides the rebounding, I'd say both teams were relatively equal (which, in this case, means equally bad). But with the huge disparity in rebounding, we absolutely got crushed. My initial reaction to the weight given to rebounds (i.e. as much as points) in some statistical measures such as Win Score had been "that's way too high," even though I've always been more focused on defense and rebounding as a player. My thinking has been definitely been changing on that front, and today's games were simply yet another example driving home the importance of rebounding.

What statistics do you think are really important, and which ones do you think are highly overrated?

1 comment:

Louis said...

I think that the assist-to-turnover ration is not a useful statistic. Turnovers and assists are important on their own but they do not relate to each other. I would rather know a player’s steals-to-turnover ratio because stealing the ball is the opposite of giving it away.