Why is our KenPom dropping?

Submitted by JamesBondHerpesMeds on January 11th, 2019 at 12:23 PM

I've not had a chance to wonk away at KenPom, but I find it intriguing/concerning that we've dropped to 7th in the KenPom rankings. Is this because:

- Other teams are simply refining their offense/defense and winning more emphatically than us?

- We're not winning games as emphatically as we used to?

- Something else?

 

kjaskolski

January 11th, 2019 at 12:45 PM ^

Michigan has not really blown out any teams lately.  Solid 10 point (or thereabouts)wins, but nothing where they put the hammer down.

Take for example MSU, they would have won the game last night by about 20, instead of 10.  Their Purdue game is a great example, it was about a 7 point game with about 5 minutes left, then they went out and buried them to win by almost 20, that is a big difference in KenPom when compared to Michigan winning against Illinois just by 10, Penn State by 11, etc.

Another team is Virginia who routinely are blowing teams out by 20 + points.  Michigan just does not do that as often (Nova game notwithstanding) as other squads.  But, at the end of the day, Michigan is 16-0, and most games (except for the NW game) have been pretty comfortable, especially down the stretch

Joined: 02/03/2011

MGoPoints: 432

Lumpy_wolverine

January 11th, 2019 at 12:54 PM ^

Yes, other teams are jumping us in the Kenpom adjusted efficiency margin.

We win in Illinois by 10 points (expected Kenpom margin, our adjusted efficiency doesn't change).

Tennessee wins in Missouri (similar team to Illinois, in Kenpom) by 24.  This is after beating Georgia by 46.  (I'm in Knoxville, so the Vols are the other team that I follow closely).

No surprise that Tennessee passes us up in Kenpom.

Daily Kenpom rankings are like daily stock market valuation.  If one is invested in these rankings, then it's best only to look at them once a month.

Joined: 02/01/2013

MGoPoints: 1554

MH20

January 11th, 2019 at 3:15 PM ^

Daily Kenpom rankings are like daily stock market valuation.  If one is invested in these rankings, then it's best only to look at them once a month.

I have not yet figured this out. Last night the nerd in me was really hoping that Trent Frazier missed the three he jacked up that cut it to a 10-point game with :18 left because Michigan would have most likely rebounded and then dribbled out the clock to a 13-point win (and a 3-point clear of the KP projection). But no, he had to make that three so then Brad Underwood decided that he could still win the game and waste everyone's time.

Joined: 10/30/2008

MGoPoints: 40643

MH20

January 11th, 2019 at 1:01 PM ^

Michigan moved up to #4 on December 6th, its highest overnight ranking (there were a few times where they temporarily went up to #3 before the remainder of the night's results came in). For the next several weeks, they defeated a bunch of bad teams at home by less than the expected margin of victory according to the KenPom projection. While this was happening, the teams around Michigan (most notably MSU, Tenn, and Gonzaga) were also winning but by the same margin or larger as compared to the KP projection. Only recently (starting with PSU) has Michigan resumed meeting or exceeding the KP number. 

Michigan's high water efficiency margin was 26.65 on the morning of December 4th (good for #5 overall). Today it's 26.28, good for #7 overall.

In conclusion: 

¯\_(ツ)_/¯

Joined: 10/30/2008

MGoPoints: 40643

northernmich

January 11th, 2019 at 1:05 PM ^

There is reason for concern for this team. 16-0 is great but I don’t think we are really getting better every game. Can we? We sure as hell can, but only beating a super shitty Illinois team by 10, when other highly ranked teams like Michigan are beating their Illinois equivalent teams by 25+, we will get passed. Idk how many times Michigan could have blown that game open, but there was just so many dumb ass mistakes and lapses in defense. Did we peak too soon? I hope not, but I wouldn’t rule it out of the realm of possibility. Look at Tennessee and Duke, even Virgina now, every single player is a threat to take it to the hoop or hit a 3. Charles bogs the offense down, and Z isn’t a realible shooter. Poole was not good defensively last night either, Iggy wasn’t great either. Good teams will expose them, and we need to have realible scorers. We have scorers, but are they realible? We will figure these questions out soon I hope.

Joined: 10/12/2018

MGoPoints: 3032

MaizeNBlue_Kzoo

January 11th, 2019 at 1:14 PM ^

Until the team actually loses a game (and even after), there is really no reason to be so concerned about this team.  Does it have vulnerabilities?  Sure.  So does every other team.  This team is SO FUN TO WATCH if you stop worrying about exceeding expectations and just enjoy them.  They play defense.  They hustle.  They play together as a team.  They enjoy the journey.  We should too. 

Joined: 01/09/2011

MGoPoints: 420

Starko

January 11th, 2019 at 2:03 PM ^

On a per-possession basis, Michigan has been giving up slightly more points than it had earlier in the season and scoring slightly fewer. 

Joined: 10/08/2008

MGoPoints: 782

cornman

January 11th, 2019 at 3:10 PM ^

Have you not watched the games?  We've struggled with basically everybody we've played in the last month.  If you want to be a top ten team in today's NCAA, you need to blow out bad teams.  We haven't been able to do that.

Joined: 12/17/2018

MGoPoints: 639

TrueBlue2003

January 11th, 2019 at 3:24 PM ^

It is 100% other teams passing Michigan by destroying teams by a lot more than expected while Michigan has only slightly exceeded expectations in the past couple weeks. (Prior to that - the WMU/AirForce stretch was different).

Kenpom ranks teams based on his Adjusted Efficiency Margin metric which is basically an estimate of how much you'd be expected to beat an average D1 team by in a game in a game with an average number of possessions.

Michigan has actually met or exceeded KenPom expectations in the last three games so their AdjEM has GONE UP.

But other teams have exceeded their expectations even more and have increased their AdjEM. UVA, MSU, UNC, Tennessee and Gonzaga are all covering their kenpom spreads by like 10-20 points so the algorithm is bumping their AdjEM by more than M's which has been covering by only a couple points.

NOTE: I know I'm pretty late to the game with that and it's been mentioned elsewhere but I had this mostly written up when work distracted me from posting, wtf!  Just listen to J. That guy knows what he's talking about.

Joined: 01/05/2010

MGoPoints: 38075

notYOURmom

January 11th, 2019 at 3:29 PM ^

It's a RANK, so a possibility is that other team's stats are improving, even if ours are not in decline.  

In the first exam of the semester, you do well and many other people bomb.  In the second, the bomb-ers have figured it out and do well and so do you.   Your rank may well fall, just because everyone else caught up.

Joined: 03/13/2009

MGoPoints: 1647

SysMark

January 11th, 2019 at 5:03 PM ^

I'll take the wins over kenpom at this stage.  We're so far ahead of where were supposed to be at the start of the season that it's really hard to complain.

Joined: 11/15/2008

MGoPoints: 21752