Jump to content

NET rankings


Recommended Posts

I can see why you guys don't like NET. I use this site to check all the different rankings - https://www.masseyratings.com/cb/compare.htm

The ranking systems are at the top, so each column is a different system. The blue and red numbers show a team's lowest and highest rankings. So a ranking system with a bunch of red/blue numbers would be suspect as it would show a large number of deviations. And the NET actually looks pretty good there, only 1 blue number. Which is of course Indiana. And it really sticks out because it's way off from where all the other systems have you guys. Really strange....

Link to comment
Share on other sites

  • Replies 279
  • Created
  • Last Reply
4 minutes ago, BruceDouglas said:

I can see why you guys don't like NET. I use this site to check all the different rankings - https://www.masseyratings.com/cb/compare.htm

The ranking systems are at the top, so each column is a different system. The blue and red numbers show a team's lowest and highest rankings. So a ranking system with a bunch of red/blue numbers would be suspect as it would show a large number of deviations. And the NET actually looks pretty good there, only 1 blue number. Which is of course Indiana. And it really sticks out because it's way off from where all the other systems have you guys. Really strange....

Indiana around 32 seems about right.  

Link to comment
Share on other sites

On 2/21/2020 at 10:56 AM, PritchardPutBack said:

Image

 

The NCAA tweeted out "most" of the formula with this picture once. It says that the weight of each numbered section is in descending order but I wouldn't be surprised if the lack of road wins are hurting and the scoring margin in conference games, even with the differential in a game being capped at 10

So if the weight of each numbered section is in descending order then I could see an argument that location is actually factored in twice and therefore neutral and away games are factored essentially twice.  

Box one- team value= opponent, LOCATION, winner.  

Box 4 doubles down and gives extra points for road wins.  And something we haven't talked about.  Losing at home hurts you just as bad as not winning on the road.  Maryland, Purdue, Arkansas.  

Link to comment
Share on other sites

24 minutes ago, NotIThatLives said:

So if the weight of each numbered section is in descending order then I could see an argument that location is actually factored in twice and therefore neutral and away games are factored essentially twice.  

Box one- team value= opponent, LOCATION, winner.  

Box 4 doubles down and gives extra points for road wins.  And something we haven't talked about.  Losing at home hurts you just as bad as not winning at home.  Maryland, Purdue, Arkansas.  

You may be on to something.  But even if what you pointed out is not true, something is obviously broke. 

Also, #3 and #5 blatantly ignore the quality of the opponent.  Beating 1-25 Chicago State at home by 11 is better than beating Kansas at home by 1.  That's stupid and the NCAA is stupid for implementing it.

Link to comment
Share on other sites

3 hours ago, NotIThatLives said:

So if the weight of each numbered section is in descending order then I could see an argument that location is actually factored in twice and therefore neutral and away games are factored essentially twice.  

Box one- team value= opponent, LOCATION, winner.  

Box 4 doubles down and gives extra points for road wins.  And something we haven't talked about.  Losing at home hurts you just as bad as not winning on the road.  Maryland, Purdue, Arkansas.  

Good points. I know we've ran into teams on the road  either ranked or while they are getting hot (Penn State/Maryland for ranked, Rutgers/Michigan/Ohio State for getting hot) but if we could have squeezed some road wins in there I'm really curious what difference it would have made. Big Ten teams also with around 3 home losses include Michigan State/Illinois at 10-3 and MIchigan/Minnesota/Purdue at 10-4 when home (we're 14-3). When you look at their NET and road records they are

13 Michigan State 5-4 (plus 5 neutral court wins)

22 Michigan 4-5 (plus 4 neural wins)

35 Illinois 5-5

36 Purdue 3-8 (plus 1 neutral win)

42 Minnesota 2-8 (plus 1 neutral win)

52 Indiana 2-8 (plus 2 neural wins)

 

This does ignore if there are any efficiency issues since that is ranked number 2 in importance, but just something I've noticed that seems to go together

Link to comment
Share on other sites

1 hour ago, milehiiu said:

Two ranked teams that IU has beaten are doing their best in terms of keeping IU's NCAAT hopes alive.....  won last night (Feb.25 ). Michigan State and Florida State.

And yet IU drops from 52 to 53. FL ST, after beating #11 LVille, dropped from 8 to 9. LVille after losing by 15 to FSU, remained #11. Meanwhile Duke, which lost to #109 Wake Forest by 12, remained #6.

NET is completely broken

Link to comment
Share on other sites

On 2/25/2020 at 12:10 PM, BruceDouglas said:

I can see why you guys don't like NET. I use this site to check all the different rankings - https://www.masseyratings.com/cb/compare.htm

The ranking systems are at the top, so each column is a different system. The blue and red numbers show a team's lowest and highest rankings. So a ranking system with a bunch of red/blue numbers would be suspect as it would show a large number of deviations. And the NET actually looks pretty good there, only 1 blue number. Which is of course Indiana. And it really sticks out because it's way off from where all the other systems have you guys. Really strange....

Thanks,

That helps illustrate why the NET rankings seems so off to us IU fans. The NET rankings in your attachment only appear  “off” (ie shaded blue on your spreadsheet) for 2 schools out of 353.  IU and Tennessee-Martin. I have to think that will be pointed out to the committee.

In any case, win 2 of the next 4 and we will be 10-10 in the conference and a shoo-in for the tourney.

Link to comment
Share on other sites

29 minutes ago, IU878176 said:

Thanks,

That helps illustrate why the NET rankings seems so off to us IU fans. The NET rankings in your attachment only appear  “off” (ie shaded blue on your spreadsheet) for 2 schools out of 353.  IU and Tennessee-Martin. I have to think that will be pointed out to the committee.

In any case, win 2 of the next 4 and we will be 10-10 in the conference and a shoo-in for the tourney.

Must be someone at TN-Martin with an IU connection

Link to comment
Share on other sites

Story from Matt Norlander on the Big Ten.

https://www.cbssports.com/college-basketball/news/court-report-how-the-big-ten-changed-its-schedule-embraced-the-net-and-became-the-nations-top-conference/

He has this blurb regarding Indiana and the NET:

I noticed this recently for Indiana:

  • Strength of Record: 22nd
  • Sagarin: 25th
  • KPI: 36th
  • KenPom: 37th
  • BPI: 38th
  • NET: 53rd (!)

That gap though. How there can be a 30-spot space between SOR and NET with 27 games worth of data is vexing, particularly when you consider that NET and SOR take nonconference strength of schedule into account and are results-based, not predictive. The Hoosiers are angling toward the NCAAs, but in terms of how they get seeded, it wouldn't surprise me if we get relatively volatile disagreement amongst bracketologists on the morning of Selection Sunday. 

 

 

Link to comment
Share on other sites

8 minutes ago, 5fouls said:

Story from Matt Norlander on the Big Ten.

https://www.cbssports.com/college-basketball/news/court-report-how-the-big-ten-changed-its-schedule-embraced-the-net-and-became-the-nations-top-conference/

He has this blurb regarding Indiana and the NET:

I noticed this recently for Indiana:

  • Strength of Record: 22nd
  • Sagarin: 25th
  • KPI: 36th
  • KenPom: 37th
  • BPI: 38th
  • NET: 53rd (!)

That gap though. How there can be a 30-spot space between SOR and NET with 27 games worth of data is vexing, particularly when you consider that NET and SOR take nonconference strength of schedule into account and are results-based, not predictive. The Hoosiers are angling toward the NCAAs, but in terms of how they get seeded, it wouldn't surprise me if we get relatively volatile disagreement amongst bracketologists on the morning of Selection Sunday. 

 

 

I truly don’t get it. We’re going to get punished for losing by more than we should have to good teams on the road. Meanwhile our wins stack up with anyone in the country but don’t seem to matter at all.

Link to comment
Share on other sites

8 minutes ago, 5fouls said:

Story from Matt Norlander on the Big Ten.

https://www.cbssports.com/college-basketball/news/court-report-how-the-big-ten-changed-its-schedule-embraced-the-net-and-became-the-nations-top-conference/

He has this blurb regarding Indiana and the NET:

I noticed this recently for Indiana:

  • Strength of Record: 22nd
  • Sagarin: 25th
  • KPI: 36th
  • KenPom: 37th
  • BPI: 38th
  • NET: 53rd (!)

That gap though. How there can be a 30-spot space between SOR and NET with 27 games worth of data is vexing, particularly when you consider that NET and SOR take nonconference strength of schedule into account and are results-based, not predictive. The Hoosiers are angling toward the NCAAs, but in terms of how they get seeded, it wouldn't surprise me if we get relatively volatile disagreement amongst bracketologists on the morning of Selection Sunday. 

 

 

Good synopsis.  I'm nearly certain the discrepancy in the NET is due to our blowout road losses.  Indiana is a bit of a study case here.  It is fairly unusual for a team to beat several highly rated teams and also to get crushed in every road game but 2.  There is enough data that suggests NET probably overvalues road results. 

Link to comment
Share on other sites

What was wrong with the old system when the committee looked at record and strength of schedule?  There were disagreements—which there would always be short of Jesus raising his hand and decreeing a bracket—but nothing to the extent of what this ridiculous system has produced.  

Rule of thumb, if the NCAA came up with it, it will be a mess, inconsistent, arbitrary, incongruous, etc.  The NET came from the NCAA.   It looks like they did a poor job running the numbers as they overweight road wins, even against lower comp, margin of victory, etc.  It is a mess.  I have worked on similar things.  We put together a calculation on bonus programs.  We ran a bunch of hypotheticals and it helped us iron out anomalies and gave us insight into tweaks.  I’m a science guy so I understand rigorous protocol, math, etc.   It’s pretty obvious that the NCAA wildly failed in these aspects.  

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.




×
×
  • Create New...