Luke Winn article on team efficiencies

[jzp5079]

http://sportsillustrated.cnn.com/2008/writers/luke_winn/01/02/early.warnings/index.html


delete my thread if someones posted already, I glanced over the threads but may have missed it. (I'm tired, been up all night looking for a new car online)

However, if no one has posted this yet, read through and on page 2, I think this will have some people feeling good - its at least interesting to say the least.

[gotham devil]

http://sportsillustrated.cnn.com/2008/writers/luke_winn/01/02/early.warnings/index.html


delete my thread if someones posted already, I glanced over the threads but may have missed it. (I'm tired, been up all night looking for a new car online)

However, if no one has posted this yet, read through and on page 2, I think this will have some people feeling good - its at least interesting to say the least.

Interesting article
Thanks for posting it.

[Troublemaker]

I suspect Duke's efficiency is a bit inflated right now because our style of play lends itself to beating bad teams worse than other contenders can beat bad teams. Specifically, we like to shoot threes and we hit a good percentage of them and we like to pressure and force turnovers, which can lead to efficiency-inflating blowouts. There's some "duh" factor here, of course, since all contenders play better against bad teams than against good teams, but I do think the effect is more pronounced with Duke. I really like this Duke team and I think we were underrated in the preseason, BUT right now, all the computer / nerd rankings are overrating us.

[mepanchin]

I suspect Duke's efficiency is a bit inflated right now because our style of play lends itself to beating bad teams worse than other contenders can beat bad teams. Specifically, we like to shoot threes and we hit a good percentage of them and we like to pressure and force turnovers, which can lead to efficiency-inflating blowouts. There's some "duh" factor here, of course, since all contenders play better against bad teams than against good teams, but I do think the effect is more pronounced with Duke. I really like this Duke team and I think we were underrated in the preseason, BUT right now, all the computer / nerd rankings are overrating us.

Kansas and UNC live off turning over inferior teams too, what's your point?

Duke also shoots as well inside the arc as UNC does, and has protected the ball. There's no "overrating" involved, it's just a statement of who has played the best on both sides of the ball so far. Duke is appropriately placed there. How we will look in 2 months is anyone's guess, though.

[Troublemaker]

Kansas and UNC live off turning over inferior teams too, what's your point?

Duke also shoots as well inside the arc as UNC does, and has protected the ball. There's no "overrating" involved, it's just a statement of who has played the best on both sides of the ball so far. Duke is appropriately placed there. How we will look in 2 months is anyone's guess, though.

We don't disagree (too much). Yes, the stats are reflective of how well Duke has played so far. However, they hold little predictive value for Duke's future games, or more precisely, they hold less predictive value for Duke than for other contenders. I suspect KU is overrated as well (from a predictive value standpoint) so I'm not saying Duke is alone.

Take a look at KU's schedule, for example. The only two top-50 teams they've played have given them very close games. They are ranked #1 in efficiency on the strength of blowing out bad teams worse than other contenders. I'm not saying that's a bad skill to have. I am saying it's a misleading skill since stats for games against good teams and for games against bad teams are not separated, but what ultimately decides how successful a team is, are only the games against good teams.

You're also conflating (or perhaps it's de-conflating) the issue a bit. It's not just "shooting a good percentage from three." It's shooting them well AND shooting a lot of them. AND forcing a lot of turnovers. It's the entire package I talked about. UNC, for example, is one of the least 3-reliant teams in the country, so it's cheating to count them.