blog | pomeroy ratings | stats

Friday, February 02, 2007

Hurricane Dicta

On this eve of Super Bowl Weekend involving Da Bears, it’s appropriate to acknowledge that we all have a little Dicta inside of ourselves. Big Ten Wonk reminded me of one my own dicta a while back, and I decided to create a complete list of the the others right here. If you’re ever wondering why things are so darn messed up around here, let this be your reference guide. These are my core hoops beliefs and philosophies.

Don’t confuse achievement with performance. If I somehow happened to become a coach, I would totally write this on the chalkboard after a close win over a lousy team. Future achievement can be better predicted by past performance than past achievement. How to measure performance is the tricky part, but I know it goes deeper than recording W’s and L’s, or even looking at who a team beat and lost to. If you’re wondering why I am often unconcerned about past achievements, it’s because so many other people have that covered. You can look at the conference standings and see achievement. Trying to figure out a team’s level of performance is more challenging, but also more insightful. Naturally, there comes a time in the season where achievement trumps performance, and that time is called March Madness.

The RPI is nearly useless. Following from above, the RPI measures achievement for the most part. I don’t have much use for it these days. (Although, a hearty thanks to Mr. David Rueth who continues to ensure that my site designations continue to match what the NCAA has in their database, thus ensuring a totally accurate RPI on this site for those of you that need that sort of info.) I know I once defended the thing, and in its original form, the combination of simplicity and accuracy was admirable. But the RPI was invented in 1981! In 1981, it was a state of the art formula. In 1981, Atari was a state of the art video game system. Kids aren’t playing games on an Atari, yet the RPI is still the gold standard for the NCAA.

One game matters little in evaluating performance. Honestly, I don’t think my mind is ever changed by one game. If you want to bore me, try to convince me that team A is better than team B because A beat B. I need a little more than that. It gets even better when you tell me that team A beat team C who beat team B. I need to see longer-term trends to base an opinion on it.

Stats don’t lie. Seriously. If Joe Blow scores 20 points in a game, he scores 20 points. If he does it on 10 for 36 shooting with 12 turnovers, and Johnny Analyst praises Joe Blow’s brilliant offensive performance, then it’s Johnny Analyst doing the lying, not the stats.

Offensive and defensive rebounding are different things. I mean, it was 14 years ago that the NCAA began distinguishing between offensive and defensive rebounds. It was a major advancement in stats keeping. Yet, with the exception of Fran Fraschilla, analysts will simply refer to a team’s “rebounding” either qualitatively or quantitatively, which often is misleading. These are two distinct, albeit related, skills. Offensive rebounding requires individual effort. Defensive rebounding is more team oriented. Florida is a great defensive rebounding team. They are not a great rebounding team - they are less than great on the offensive end.

PredictionsAbsolute statements are stupid. This is a slight adjustment from the old ‘predictions are stupid’ mantra. Predictions aren’t stupid, but applying total certainty to them is. I fall into the trap myself, and I usually regret it.

It’s more important what a coach does than what he says. I’m not saying to ignore what a coach says, but see if his quotes jive with the statistical record before you take his preaching as gospel.

The single most important factor in determining the outcome of close games is chaos. I used to call this luck, but that hurts people’s feelings, so the new PC term is “chaos.” Anyway, there’s no silver bullet to winning close games. It helps to be better than the opponent, but aside from that, there are dozens of factors that influence the outcome of the one-possession game, some of which a team has no control over.

Statistical and visual analysis go together. You’re foolish using one without the other, but so many people do, and they’re worse off for it. We all have biases, and of course stats do also, some much worse than others. By looking at the proper stats you can check your biases at the door and get another “opinion” on what is happening.

The ultimate goal should be to find the truth. And if somebody says they know it, that’s when you should get suspicious.

Posted on 02/02 at 05:00 AM
Permalink • E-mail me

Powered by ExpressionEngine