MH20

February 28th, 2018 at 3:54 PM ^

Evidently none of them have looked at their RPI metrics.  Even ignoring the current quadrant system, Nebraska is only 3-8 against the RPI top-100 which is why their own RPI figure is 59th.  Despite having a much better conference record, the Huskers kind of remind me of Michigan's 2016 team that also only had three top-100 RPI victories and had to beat IU just to get into the First Four.

Joined: 10/30/2008

MGoPoints: 40408

tenerson

February 28th, 2018 at 3:55 PM ^

I think Miles is just having some fun here at Kenpoms expense. This doesn't seem as much mean spirited as just a little fun with the model that was cleary a ways off. 

Joined: 10/04/2009

MGoPoints: 2371

Boner Stabone

February 28th, 2018 at 3:56 PM ^

They have a lousy KenPom ranking and a lousy RPI ranking.  It is the direct effect of feeding off the bottom teams in the conference all season and inflating their win total.

Joined: 01/06/2015

MGoPoints: 23995

Blukon Cornelius

February 28th, 2018 at 4:49 PM ^

If the first 3 (MSU, OSU and Purdue) each had, as of the taking of this snapshot, 100% chance of finishing in the Top 4, and Nebraska had a 93.3% chance of finishing in the Top 4, how in the hell did Michigan have a 23.0% chance of finishing there?  The first 3 take 3/4 of the Top 4 away, meaning the remaining teams are playing for only 1 remaining Top 4 position.  Undere thos circumstances, shouldn't remaining possibilities add up to 100% (multiple teams fighting for 1 remaining spot, rather that multiple teams fighting for multiple remaining spots)?  Statistics am hard. 

Joined: 07/14/2015

MGoPoints: 259

MGoPoe

February 28th, 2018 at 4:24 PM ^

I have a wholesome South Dakota boy exterior, I had no idea he was from my home state.  I.....I don't know how to feel about this right now.

 

Edit:  Nm he's from east river, I get it now.

Joined: 10/06/2013

MGoPoints: 1446

MGoPoe

February 28th, 2018 at 6:31 PM ^

The Missouri River splits the state roughly in half, east river is a lot like Iowa/southern Minnesota/Illinois:  corn, soy beans, hogs, it's flat, lot more water.

West river is entirely rolling hills and plains, can't grow much, not a lot of water, lot of cattle ranches, but dammit we have the Black Hills and I'll take that over a 2018 NCAA hockey regional site any day.

Joined: 10/06/2013

MGoPoints: 1446

Kamtheram

February 28th, 2018 at 6:42 PM ^

Anyone see the Lunardi predictions for SEC. Bama a 9?. Tennessee a 3 seed? Kentucky a 5??? Nebraska is better than the 8th best sec team, for sure, and has about 7 more wins

Joined: 02/11/2013

MGoPoints: 108

jbrandimore

February 28th, 2018 at 7:45 PM ^

If anyone knows on “tempo.” UVa is amazingly efficient, but the way they play it would seem leads to fewer possessions for both teams in a 40 minute game. To me, fewer possessions would automatically increase the luck/randomness factor. Is this correct? Also, if we are going to place the highest value on an equation like (net efficiency=offensive PPP-defensive PPP) and (expected margin of victory=net efficiency*total possessions) wouldn’t there be a case to be made that a team with a lower net efficiency, but one who consistently generated many more possessions per 40 minutes might be the better team? To me, an example of this us old timers remember is Loyola Marymount. They got a Michigan team that loved to run and shoot 3’s into a 3 point shooting contest. Michigan scored well over 100 points, but was never even competitive in that game. Does KenPom address the abilities of a team to impose it’s preferred pace and ability to force additional possessions?

Joined: 01/06/2015

MGoPoints: 13706

J.

February 28th, 2018 at 8:28 PM ^

1 - Yes.  Fewer possessions absolutely increases the luck factor; all other things being equal, a team with a faster tempo will win more frequently than a team with slower tempo.  (All other things are rarely equal, of course. :)

2 - Not really.  I see what you're saying, but each game only has one tempo, so the more efficient team has an advantage in any given game.  Now, the higher-tempo, less-efficient team could win more games due to luck, but KenPom is a predictive system -- when you're trying to predict the winner, you want the more efficeint team.

3 - That game was an outlier by any margin, though.  I get about 108 possessions, meaning Loyola-Marymount scored close to 1.4 points per possession.  Loyola also made 21/40 threes and turned the ball over half as often as Michigan.  If anything, that game probably exemplifies the advantage that a highly-efficient offensive team can have.  You can be slow & efficient or fast & efficient; Loyola was the latter.

4 - That's a good question.  KenPom projects a pace number for each game; it doesn't seem to be exactly equal to the average of the two, so maybe he is taking some kind of measurement to determine whose pace is more likely to prevail.

Joined: 03/11/2012

MGoPoints: 13057

HailHail47

March 1st, 2018 at 1:20 AM ^

I don’t think KenPom was “inaccurate” as much as Nebraska got better. No model can account for the coaching and improvement factor. Some people work harder than others to get better. Further, any early season modeling is going to be inaccurate due to limited data. Now the model has them as a mediocre team and that seems right.

Joined: 10/07/2017

MGoPoints: 5305