KenPom Explanation?

[ChillinDuke]

Hey all,

Does anyone know what KenPom's statistical columns stand for? I'm sure the model is proprietary, but does he (or did he ever) explain somewhere what AdjO and AdjD and AdjT, etc mean? (I assume Adjusted Offense, Defense, T?)

Just wanted to know if there is some background in how to read some of his inputs. For example, what can I learn from us being ranked #93 in SOS Pyth?

Feel free to PM me if that is a better avenue to discuss this. If this is all proprietary, then I apologize and please disregard.

Go Duke!

- Chillin

[Kedsy]

Hey all,

Does anyone know what KenPom's statistical columns stand for? I'm sure the model is proprietary, but does he (or did he ever) explain somewhere what AdjO and AdjD and AdjT, etc mean? (I assume Adjusted Offense, Defense, T?)

Just wanted to know if there is some background in how to read some of his inputs. For example, what can I learn from us being ranked #93 in SOS Pyth?

Feel free to PM me if that is a better avenue to discuss this. If this is all proprietary, then I apologize and please disregard.

Go Duke!

- Chillin

Check these out. They might help.

http://kenpom.com/blog/index.php/weblog/stats_explained/

http://kenpom.com/blog/index.php/weblog/ratings_explanation/

http://kenpom.com/blog/index.php/weblog/national_efficiency/

[Wander]

Everything is a rating adjusted for number of possessions per game. Adjusted offense is the expected number of points the team would score if it had 100 possessions against an average defense. Adjusted defense is the expected number of a points a team would give up in 100 possessions against an average offense. Adjusted tempo is the average number of possessions that team would have in a 40 minute game (ie, a measure of how fast that team plays).

You probably already know this, but the idea is to keep a team like UNC from having their offense look really good (or defense really bad) just because they play at a fast pace, and therefore more points are scored in an average UNC game.

[CDu]

Hey all,

Does anyone know what KenPom's statistical columns stand for? I'm sure the model is proprietary, but does he (or did he ever) explain somewhere what AdjO and AdjD and AdjT, etc mean? (I assume Adjusted Offense, Defense, T?)

Just wanted to know if there is some background in how to read some of his inputs. For example, what can I learn from us being ranked #93 in SOS Pyth?

Feel free to PM me if that is a better avenue to discuss this. If this is all proprietary, then I apologize and please disregard.

Go Duke!

- Chillin

AdjO is adjusted offensive efficiency, in which he estimates the expected number of points per 100 possessions after adjusting for quality of opponents. AdjD is adjusted defensive efficiency, which is the same concept as AdjO but points allowed per 100 possessions.

AdjT is adjusted tempo, which estimates the number of possessions per game. It's just a measure of the pace at which a team plays.

Pyth is the Pythagorean estimated win percentage. It is based on the adjusted offensive and defensive efficiency stats. It's ultimately what Pomeroy uses to rank teams. There's also a strength of schedule Pyth (which measures the quality of your opponents) and the non-conference SOS Pyth (which measures the quality of your non-conference opponents).

Being ranked #93 in SOS Pyth states that, at this point, we've played the 93rd toughest schedule according to Pomeroy's metrics. It's largely meaningless this early in the season though. In general, I'd say the SOS is an overrated measure anyway. I'd focus more about our own Pyth and our AdjO and AdjD stats.

[Billy Dat]

Interesting that Wisconsin, and their #1 ranked AdjD, are currently the #1 ranked KenPom squad. We, and our 28th ranked AdjD, are ranked 9th. Our strength of schedule rankings are insane - along with Kansas and Long Beach State - all caveats included. Need to improve that D.