MAC OOC Results 2019

[DocCas86]

Here are some results of top MAC teams (based on rankings as of 12/31 - median of barttorvik, kenpom and massey).

Initial observations:

Perhaps a UB bias - if we win Dartmouth / Army games - we would likely be top ranked MAC team heading into the regular season.  Water under the bridge, I know.  Hopefully, an indication that the potential is there for this team in MAC play.

Bowling Green - Their ranking has been hurt by losses to Norfolk St and Quinnipiac playing without Justin Turner.  He returned with 22 points last night against Hartford. 

Ball State - At first glance their ranking seems high - 2-5 against teams with ranking 200 or better.

Other MAC teams - I did not review their rankings to see if they warrant consideration.

Top 100 results

Overall record for top six MAC teams is 3-10.  Buffalo is the only team with a winning record against top 100 teams (DePaul / Harvard).  Bowling Green is the only other team with win against top 100 (Cincinnati 91-84 at a neutral site).  Akron played Louisville close (82-76 loss @Louisville), and Toledo lost by 2 to Notre Dame 64-62 loss @ND).

Additional Analysis on Rankings

Here are couple of charts that compare current rankings to results in OOC games.  A little background on the charts:

Team rank - is the median rank from barttovik.com, kenpom.com and masseyratings.com

x-axis - Team rank - opponent rank

y-axis - PF - PA normalized to neutral court.  If away game, add 3.5 points, if home game subtract 3.5 points

The trend line presented is my estimate of the expected point differential given rank differential.   The estimate is about 1 point differential for every 10 difference in ranking.  If the data point is above the line, the team performed better than expected, and below the line, worse than expected.

Akron - Game results appear to be in line with their ranking.  Games helping their ranking is previously mentioned 6 point loss at Louisville.  They did underperform a bit in 80-67 loss to Liberty (#54) at a neutral site.

Kent St - Again results appear in line with their ranking.  The game that appears to influence their ranking is a 96 - 68 loss at Mississippi St.

Buffalo - The ranking seems fair based on the results however, you can see the variance in the performance thus far - wins against DePaul and Harvard are much better than current ranking would suggest, while losses against Dartmouth and Army are much worse than expected.

Ball State / Toledo / Bowling Green  - Grouping together as they all appear to have been inconsistent over the course of the year - Bowling Green in particular has performed above their ranking for teams that are ranked above them and below expectations for teams ranked below them, with several of these games coming when Justin Turner was out.  Given that, I would say Bowling Green ranking a bit understated.

 

 

 

[John]

5 hours ago, DocCas86 said:

Here are some results of top MAC teams (based on rankings as of 12/31 - median of barttorvik, kenpom and massey).

Initial observations:

Perhaps a UB bias - if we win Dartmouth / Army games - we would likely be top ranked MAC team heading into the regular season.  Water under the bridge, I know.  Hopefully, an indication that the potential is there for this team in MAC play.

Bowling Green - Their ranking has been hurt by losses to Norfolk St and Quinnipiac playing without Justin Turner.  He returned with 22 points last night against Hartford. 

Ball State - At first glance their ranking seems high - 2-5 against teams with ranking 200 or better.

Other MAC teams - I did not review their rankings to see if they warrant consideration.

Top 100 results

Overall record for top six MAC teams is 3-10.  Buffalo is the only team with a winning record against top 100 teams (DePaul / Harvard).  Bowling Green is the only other team with win against top 100 (Cincinnati 91-84 at a neutral site).  Akron played Louisville close (82-76 loss @Louisville), and Toledo lost by 2 to Notre Dame 64-62 loss @ND).

Additional Analysis on Rankings

Here are couple of charts that compare current rankings to results in OOC games.  A little background on the charts:

Team rank - is the median rank from barttovik.com, kenpom.com and masseyratings.com

x-axis - Team rank - opponent rank

y-axis - PF - PA normalized to neutral court.  If away game, add 3.5 points, if home game subtract 3.5 points

The trend line presented is my estimate of the expected point differential given rank differential.   The estimate is about 1 point differential for every 10 difference in ranking.  If the data point is above the line, the team performed better than expected, and below the line, worse than expected.

Akron - Game results appear to be in line with their ranking.  Games helping their ranking is previously mentioned 6 point loss at Louisville.  They did underperform a bit in 80-67 loss to Liberty (#54) at a neutral site.

Kent St - Again results appear in line with their ranking.  The game that appears to influence their ranking is a 96 - 68 loss at Mississippi St.

Buffalo - The ranking seems fair based on the results however, you can see the variance in the performance thus far - wins against DePaul and Harvard are much better than current ranking would suggest, while losses against Dartmouth and Army are much worse than expected.

Ball State / Toledo / Bowling Green  - Grouping together as they all appear to have been inconsistent over the course of the year - Bowling Green in particular has performed above their ranking for teams that are ranked above them and below expectations for teams ranked below them, with several of these games coming when Justin Turner was out.  Given that, I would say Bowling Green ranking a bit understated.

 

 

 

Thanks for some additional insight on the metrics. From a common sense stand point it hard to grasp that both Akron and Kent State are ahead of UB when all their wins are against teams ranked in the 100’s to over 200’s. The bottom line is UB has to win their MAC games. Go Bulls!!!!

[DocCas86]

5 minutes ago, John said:

Thanks for some additional insight on the metrics. From a common sense stand point it hard to grasp that both Akron and Kent State are ahead of UB when all their wins are against teams ranked in the 100’s to over 200’s. The bottom line is UB has to win their MAC games. Go Bulls!!!!

Akron's results by rank are all fairly close to the predicted result - in my opinion their ranking is reasonable.  Kent St has performed well but most games against weak teams.  They have under-performed against top teams they have played.  This may be in line with your observation that their ranking is too good.  As for UB, I previously mentioned that we would be ranked at least comparable to Akron, if you didn't have the Army / Dartmouth games.   I want to reiterate that Bowling Green is probably better than their rank

I would say Akron, Bowling Green and Buffalo are the teams that I believe should be favorites to win MAC.  I was hoping to spur discussion from others as to how they see the MAC - common sense stand points welcome, as it helps me take a critical eye at the metrics.

[Ubbulls84]

I don’t know if I believe in all the metrics, sometimes can overrate or under rate teams. Bottom line, it looks like Bowling Green and Akron will be the toughest teams we will be facing in the MAC. Kent could be good, but they played a really weak schedule and did not look good against higher level competition so jury still out on them. 
 

The conference is clearly down this year as far as rankings go, mainly due to us not being as highly ranked as last year. But also, no other teams really having many high level wins. Our game against DePaul was the best win for the conference. 
 

Dartmouth/Army losses were brutal to our rankings as those are low level D1 teams and that just can’t happen on your home court. 

Hopefully we can go 13-5 in conference which should set us up nicely for the conference tourney. 

[DocCas86]

14 minutes ago, Ubbulls84 said:

I don’t know if I believe in all the metrics, sometimes can overrate or under rate teams. Bottom line, it looks like Bowling Green and Akron will be the toughest teams we will be facing in the MAC. Kent could be good, but they played a really weak schedule and did not look good against higher level competition so jury still out on them. 
 

The conference is clearly down this year as far as rankings go, mainly due to us not being as highly ranked as last year. But also, no other teams really having many high level wins. Our game against DePaul was the best win for the conference. 
 

Dartmouth/Army losses were brutal to our rankings as those are low level D1 teams and that just can’t happen on your home court. 

Hopefully we can go 13-5 in conference which should set us up nicely for the conference tourney. 

I agree the metrics are not infallible and can be misleading as only based on a handful of observations, so range of possible ranks is large.   In addition they do not consider things such as injuries or other circumstances they may impact individual game results.

[121Merrimac]

I find the 200+ bucket to be quite large.  Here are the quantity of wins versus teams with a 300+ massey rating, or a non-D1 team.

What value can EMU gain from playing 4 non-D1 teams?  I suddenly realize Buffalo's OOC schedules over the years could have been much worse.

[skrabukes]

10 minutes ago, 121Merrimac said:

I find the 200+ bucket to be quite large.  Here are the quantity of wins versus teams with a 300+ massey rating, or a non-D1 team.

What value can EMU gain from playing 4 non-D1 teams?  I suddenly realize Buffalo's OOC schedules over the years could have been much worse.

Based on this, I think some people who constantly gripe about our "weak" opponents at home in non-league need to reconsider their position. I know that it's a challenge for UB to get "big" teams to play at AA, but clearly some of our MAC peers sign on to play just about anyone, whereas (at least in recent years) UB has some reasonable standards.

[DocCas86]

17 minutes ago, 121Merrimac said:

I find the 200+ bucket to be quite large.  Here are the quantity of wins versus teams with a 300+ massey rating, or a non-D1 team.

What value can EMU gain from playing 4 non-D1 teams?  I suddenly realize Buffalo's OOC schedules over the years could have been much worse.

Thank you for the additional cut.  This also points out that care is needed in dealing with rankings - massey.com has Eastern Michigan ranked at 136 - this is way to high for Eastern, kenpom.com has ranking of 190 - which upon review seems like a reasonable ranking for them.

[rma]

3 hours ago, 121Merrimac said:

I find the 200+ bucket to be quite large.  Here are the quantity of wins versus teams with a 300+ massey rating, or a non-D1 team.

What value can EMU gain from playing 4 non-D1 teams?  I suddenly realize Buffalo's OOC schedules over the years could have been much worse.

Amazingly CMU has a better NCSOS on Kenpom than us, which is quite the break from the past 7 or so years when they would consistently be in the lowest decile of D1.

My hypothesis is that at some point the Mt. Pleasant CYO and junior varsity high school basketball teams said enough is enough. 

[121Merrimac]

23 hours ago, DocCas86 said:

... This also points out that care is needed in dealing with rankings ...

I definitely agree with your concept of averaging a few ranking publications.  I was just lazy, and selected Massey’s site to quickly review every team’s opponents.  Originally I was curious how many non-D1 teams everyone played.  Then I got side tracked by how many 300+ teams there were, so I lumped them in too.  

After checking the NET, or Kenpom, or another site, I also like to review the following compilation.  More publications are added to the table as the season progresses- it’s up to 50!   Even though I’m not familiar with most of them, the big hitters are all there, and for me the average is a good gut check.  
https://www.masseyratings.com/cb/compare.htm

FWIW, I think the quantity of non-D1 schools is an example of why the rankings are so different from one site to another... sometimes those scores are used, and sometimes they aren’t. 

20 hours ago, rma said:

Amazingly CMU has a better NCSOS on Kenpom than us, which is quite the break from the past 7 or so years ..,

I remember you saying that some team (I couldn’t recall which) always scheduled 4 or so non-D1s.  It’s what prompted my post.  

[DocCas86]

2 hours ago, 121Merrimac said:

I definitely agree with your concept of averaging a few ranking publications.  I was just lazy, and selected Massey’s site to quickly review every team’s opponents.  Originally I was curious how many non-D1 teams everyone played.  Then I got side tracked by how many 300+ teams there were, so I lumped them in too.  

After checking the NET, or Kenpom, or another site, I also like to review the following compilation.  More publications are added to the table as the season progresses- it’s up to 50!   Even though I’m not familiar with most of them, the big hitters are all there, and for me the average is a good gut check.  
https://www.masseyratings.com/cb/compare.htm

FWIW, I think the quantity of non-D1 schools is an example of why the rankings are so different from one site to another... sometimes those scores are used, and sometimes they aren’t. 

I remember you saying that some team (I couldn’t recall which) always scheduled 4 or so non-D1s.  It’s what prompted my post.  

I am not sure who mentioned non-D1 opponents, but I was not that thorough.  I did look at a couple of  kenpom SOS and at first glance did not make sense... If I have a chance I may take a second look.

Trying to rank all D1 teams, there is bound to be more to it then the numbers, and things missed. From watching the Bulls, I think we agree they have potential to be very good and  I do like our chances in the MAC although I tend to be more conservative then some thinking 11/12 wins.

Looking forward to getting first glance of how good they will be on Saturday. 

Oh and thank you for the link.

Edited January 3, 2020 by DocCas86

[121Merrimac]

35 minutes ago, DocCas86 said:

I am not sure who mentioned non-D1 opponents,

Maybe I'm wrong, but I though rma made a comment about that one of the past couple years.  Aanyway, I realize I didn't actually share this year's non-D1 only numbers.  If I remember correctly, Toledo and Ball St are the only others with only 1.  A majority had 2, one or two teams had 3, and EMU had the most with 4.

[DocCas86]

8 minutes ago, 121Merrimac said:

Maybe I'm wrong, but I though rma made a comment about that one of the past couple years.  Aanyway, I realize I didn't actually share this year's non-D1 only numbers.  If I remember correctly, Toledo and Ball St are the only others with only 1.  A majority had 2, one or two teams had 3, and EMU had the most with 4.

Yes that was my mistake I thought you were saying I posted... Wasn't until I reread that I realized you were talking to RMA at that point. Apologies.

[rma]

2 hours ago, DocCas86 said:

I am not sure who mentioned non-D1 opponents, but I was not that thorough.  I did look at a couple of  kenpom SOS and at first glance did not make sense... If I have a chance I may take a second look.

CMU's OOC typically consists of CYO Basketball squads, local community colleges, Mt. Pleasant High School, and if they're really feeling ballsy maybe a D2 or NAIA squad.  This year they really stepped up their scheduling.