Advanced metrics and UD

[Medford]

took a brief look at kenpom today..

http://kenpom.com/

I'll admit I'm not the biggest advanced metrics guy and I don't subscribe to kenpom so I don't have any of the "inside info" However I've seen kenpom referenced many times in terms of the NCAA selection process and how teams are looked at, from Jay Bilas to Hoops HD, Lunardi, etc... its always felt like the kenpom ratings were held to a bit higher standard than other rating systems like the KPI, BPI, Sags, etc...

I believe I've also seen it referenced that the selection committee has things like kenpom available to them to look at rankings beyond just the rpi (Chris can you confirm from your mock selection process?)

anyhoo, a couple of things stood out to me. Its been mentioned before, but UD is really, really good at D, currently ranking 11th in adjusted defensive efficiency. I know some announcer in a game a while back mentioned that they were 8th, so they've been pretty solid all season long. On the flip side, they are rather mediocre in offensive efficiency, 86th. I don't know where that ranking has stood all season long, if its been fairly steady, improving or declining. Right now, UD is ranked 31st in the kenpom; from past history, it feels like if UD remains in the top 40 their at large chances are very solid. If you break it down just a bit, 80% of the top 20 defensive teams are in the current top 40. On the flip side, only 1 of the top 20 offensive teams reside outside the top 40 (#20 UNC-Wilimington sitting at 54th in the kenpom ratings).

With Cunningham set to return soon, it would seem that their defense may improve slightly, however there is only so much room to improve from 11th overall, specifically if he's healthy and at his full self, I'd expect the rebounding numbers to improve (don't know what they are) based upon his rep and 2 game UD sample size. On the flip side, based upon the very limited 2 game sample set, it appears that a healthy and confident Josh Cunningham could/may help them offensively as well. If UD is to improve their kenpom rating, they'll need to improve their offensive efficiency. Of the top 40, they currently have the 3rd worst offense (according to kenpom anyways).

So my question is, how much can they therotically improve down the stretch? Could they potentially push their offensive rating up into the 50s and similarly their overall kenpom ratings into the 20s, or is it just too late in the season for an improvement like that to be made? If they can improve, what areas of improvement would make the biggest difference?

Their PPG, 2pt%, 3pt% and a/to ratings are all near the top of the A10, but a good pace behind the best teams in the B12, B10, ACC and other places that I've looked. Is the only way to improve their kenpom to go on a tear and get their points and shooting % up closer to a team like Kansas or W Virginia? anyone that can shed some light on how those ratings work, what is used to determine them, etc.. would be greatly appreciated.

[priceg75]

Our offensive efficiency was in the 100-140 range much of the year, got up into double digits, then dropped back after the VCU game.

85 is one of our best figures of the year. It will improve as we play the weaker conference teams.

[springborofan]

Just being healthy should help. This team has been without Pollard, Cooke, and KD (along with Josh C.). That does a number of things. Obviously, some of your better players aren't on the court. It also puts different rotations in play which impacts offensive efficiency.

It's on the offensive end that I think this team will surprise people in the tournament. I'll keep banging on this drum but we need DD to heat up over the next 6 weeks. If he can hit 45% from the 3 pt. range down the stretch, this team becomes scary strong.

[Medford]

Never really got an answer to what drives the offensive efficiency numbers exactly, but none the less, UD remains overall about where they were 32 overall in kenpom, their defense has fallen down to 25 as of this morning, while their offense has improved to 59 in the country. So it would appear that with a strong finish, they could push their offense up into the top 50, maybe get their defense back down into the top 20. I'm not sure how much it matters, if at all for actual NCAA tournament selection or seeding, but the improvement on offense shows that even this late in the season there is room for some reasonable adjustment in his stats.

[flyer016]

Originally Posted by Medford Never really got an answer to what drives the offensive efficiency numbers exactly, but none the less, UD remains overall about where they were 32 overall in kenpom, their defense has fallen down to 25 as of this morning, while their offense has improved to 59 in the country. So it would appear that with a strong finish, they could push their offense up into the top 50, maybe get their defense back down into the top 20. I'm not sure how much it matters, if at all for actual NCAA tournament selection or seeding, but the improvement on offense shows that even this late in the season there is room for some reasonable adjustment in his stats.

Kenpom ratings are based on efficiency and pace.

A simplified understand: On offense, it's the number of points scored per possession, then calibrated to 100 possessions for the sake of comparison. Defense is calculated by the inverse, number of points allowed per possession, then calibrated to 100 possessions.

It's a really simple system, but fairly accurate (to my mind) and does a good job accounting for pace. To answer your question, improvement would come by scoring more efficiently (no matter the pace of the game) or defending more efficiently.

[Medford]

thanks, What constitutes a "possession"? Lets say Scoochie fires up a 3 late in the shot clock, and it bounces off the rim right to Pollard, who tries to get a quick put back, but its blocked to the perimeter where Kyle grabs it, they run some offense which leads to Charles attempting a driving layup only to have it roll off the williams who grabs and dunks it for 2 points.

Is that 1 possession, 3 (accounting for 2 offensive rebounds) or 4 (accounting for the blocked shot as well)? If the answer is 1, it seems the best way for UD to improve their offensive efficiency rating would be to get better at grabbing offensive rebounds and give themselves an additional chances at points.

[Iguomaniac]

Originally Posted by Medford thanks, What constitutes a "possession"? Lets say Scoochie fires up a 3 late in the shot clock, and it bounces off the rim right to Pollard, who tries to get a quick put back, but its blocked to the perimeter where Kyle grabs it, they run some offense which leads to Charles attempting a driving layup only to have it roll off the williams who grabs and dunks it for 2 points.

Possessions are not an official stat, so they have to be estimated. The formula KenPom uses is

# of Possessions = [Field Goals Attempted]-[Offensive Rebounds]+[Turnovers]+0.475*[Free Throws Attempted]

So in your example, there would be 1 possession with 4 FGA and 3 OR.

[Medford]

thanks, I remember seeing that formula once upon a time now that I see it again.