Official complain about net rankings thread

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
The NET rankings are a horse crap metric


feel free to agree with me in this thread
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
Penn state currently higher in the NET than Minnesota.

I guess they Are more efficient or something


All you need to know is that St Louis is 12 and Minnesota is in the 50s.



We should have a metric where winning games is more important than how close you lose and how much you win by.
 
  • Like
Reactions: ZPI

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
Don’t they only update once a week?
First one was released today.

Somehow Minnesota is one of 3 teams in the country with 10’wins
One of 8 teams in the country with 3+ quad 1 wins

But is only 43rd
 

Otis

Well-known member
Joined
Aug 13, 2014
Messages
3,974
Reaction score
767
Points
113
Lol!!!

NET rankings don't pass the eye test!

No doubt MSU and OSU are ranked higher than the Gophers as well.
 

Go Gophs

Active member
Joined
Jan 16, 2009
Messages
801
Reaction score
163
Points
43
Doesn't matter. If the season ended today, Gophers are probably a 4 or 5 seed and that's a fact.
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
Lol!!!

NET rankings don't pass the eye test!

No doubt MSU and OSU are ranked higher than the Gophers as well.
Ohio state with more losses, fewer wins, and a 1-3 record vs quad 1
Rank: 28

Michigan state 119 they aren’t even a quad 2 win for the gophers
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
Doesn't matter. If the season ended today, Gophers are probably a 4 or 5 seed and that's a fact.
That is maybe true. It doesn’t mean this is a good metric. The fact that it is used at all is a joke.

right now, beating liberty on the road is a better win than beating Michigan state at home according to NET
 

matt

Let's get weird
Joined
Apr 21, 2010
Messages
3,376
Reaction score
495
Points
83
I find it interesting how far apart the NET rankings (#43) and KenPom (#23) are, especially since they tweaked the NET rankings so that offensive and defensive efficiency numbers are more heavily weighted.
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
KenPom and Net do some things better than RPI


RPi is a better measure of achievement.
Once you have a large enough sample of games (I’d argue you never have a large enough sample when you only play 30 games and there are 300+ teams) at some point efficiency based metrics become better predictors than RPI.


I think we should measure teams based on achievement rather than based on how we think they might hypothetically achieve moving forward
 

cjbfbp

Well-known member
Joined
Feb 23, 2014
Messages
7,384
Reaction score
1,984
Points
113
Penn state currently higher in the NET than Minnesota.

I guess they Are more efficient or something


All you need to know is that St Louis is 12 and Minnesota is in the 50s.



We should have a metric where winning games is more important than how close you lose and how much you win by.
We do. It's called RPI and it was the preferred method of picking team strength for many years.

I really don't like NET (regardless of where the Gophers are ranked). We could have emptied the bench with at least 5 minutes left against Michigan State but can't do that because of NET concerns. We could have thrown in the towel and let the subs play for the last three minutes against Wisconsin but, again, had to consider NET.
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
We do. It's called RPI and it was the preferred method of picking team strength for many years.

I really don't like NET (regardless of where the Gophers are ranked). We could have emptied the bench with at least 5 minutes left against Michigan State but can't do that because of NET concerns. We could have thrown in the towel and let the subs play for the last three minutes against Wisconsin but, again, had to consider NET.
Yup.
I just said in another thread RPI is a better measure of achievement.

Flukey wins should still count as wins and flukey losses should still count as losses.

efficiency based metrics are better for predicting in many cases. But the tools used to evaluate shouldn’t be predictive metrics. They should be metrics to measure what actually happened.

3 1 point losses is not better than 2 1 point wins and 1 25 point loss.....but if you use efficiency metrics...it is.
 

howeda7

Well-known member
Joined
Nov 22, 2008
Messages
49,918
Reaction score
7,683
Points
113
Lol!!!

NET rankings don't pass the eye test!

No doubt MSU and OSU are ranked higher than the Gophers as well.
I demand a full and complete audit! Our RPI is #10.
 

cjbfbp

Well-known member
Joined
Feb 23, 2014
Messages
7,384
Reaction score
1,984
Points
113
I will give NET credit for something: at least that metric doesn't have Duke in the top 25 like the polls still do. Duke is currently #115 in the NET rankings.

The fact that we are behind Penn State does seem kind of crazy but that probably is a feature of environmental factors this season and our lopsided 2 losses. Because of cancellations, Penn State has only 7 games overall with zero games against Quad 3&4 teams. We have 12 games with 7 against Quad 3&4 teams which, as someone else mentioned, apparently includes Michigan State at this point of the season.
 

howeda7

Well-known member
Joined
Nov 22, 2008
Messages
49,918
Reaction score
7,683
Points
113
I will give NET credit for something: at least that metric doesn't have Duke in the top 25 like the polls still do. Duke is currently #115 in the NET rankings.

The fact that we are behind Penn State does seem kind of crazy but that probably is a feature of environmental factors this season and our lopsided 2 losses. Because of cancellations, Penn State has only 7 games overall with zero games against Quad 3&4 teams. We have 12 games with 7 against Quad 3&4 teams which, as someone else mentioned, apparently includes Michigan State at this point of the season.
This says MSU is #7?

Week 7
Poll Date : January 03, 2021
NET Rankings






TeamW/LConfSoS RankNET 1-5051-100101-150151-200200+
.
Dixie State4-2Western Athletic2090-10-00-00-12-0
.
Bellarmine3-3Atlantic Sun2350-10-10-00-11-0
.
UC San Diego2-0Big West0-00-00-00-00-0
.
Tarleton State2-2Western Athletic250-00-00-10-10-0
1.
Gonzaga10-0West Coast635-01-00-00-03-0
2.
Kansas8-2Big 12404-10-11-00-02-0
3.
Dayton5-2Atlantic 10541-01-11-01-11-0
4.
San Diego State6-2Mountain West161-12-02-10-00-0
5.
Baylor9-0Big 122281-04-00-00-04-0
6.
Duke3-2Atlantic Coast2290-21-00-00-01-0
7.
Michigan State7-3Big Ten2311-21-00-02-13-0


 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
This says MSU is #7?

Week 7
Poll Date : January 03, 2021
NET Rankings






TeamW/LConfSoS RankNET 1-5051-100101-150151-200200+
.
Dixie State4-2Western Athletic2090-10-00-00-12-0
.
Bellarmine3-3Atlantic Sun2350-10-10-00-11-0
.
UC San Diego2-0Big West0-00-00-00-00-0
.
Tarleton State2-2Western Athletic250-00-00-10-10-0
1.
Gonzaga10-0West Coast635-01-00-00-03-0
2.
Kansas8-2Big 12404-10-11-00-02-0
3.
Dayton5-2Atlantic 10541-01-11-01-11-0
4.
San Diego State6-2Mountain West161-12-02-10-00-0
5.
Baylor9-0Big 122281-04-00-00-04-0
6.
Duke3-2Atlantic Coast2290-21-00-00-01-0
7.
Michigan State7-3Big Ten2311-21-00-02-13-0


Not sure who is doing data entry at yahoo. But the ncaa official net on their website has MSU at 119

duke at 115
 

cjbfbp

Well-known member
Joined
Feb 23, 2014
Messages
7,384
Reaction score
1,984
Points
113
This says MSU is #7?

Week 7
Poll Date : January 03, 2021
NET Rankings






TeamW/LConfSoS RankNET 1-5051-100101-150151-200200+
.
Dixie State4-2Western Athletic2090-10-00-00-12-0
.
Bellarmine3-3Atlantic Sun2350-10-10-00-11-0
.
UC San Diego2-0Big West0-00-00-00-00-0
.
Tarleton State2-2Western Athletic250-00-00-10-10-0
1.
Gonzaga10-0West Coast635-01-00-00-03-0
2.
Kansas8-2Big 12404-10-11-00-02-0
3.
Dayton5-2Atlantic 10541-01-11-01-11-0
4.
San Diego State6-2Mountain West161-12-02-10-00-0
5.
Baylor9-0Big 122281-04-00-00-04-0
6.
Duke3-2Atlantic Coast2290-21-00-00-01-0
7.
Michigan State7-3Big Ten2311-21-00-02-13-0


That's not what I'm seeing.

 

tjgopher

Well-known member
Joined
Nov 21, 2008
Messages
4,603
Reaction score
277
Points
83
The complaint about the NET rankings is they measure which team SHOULD win, not which team actually wins. So, if you play efficiently and lose you don't get docked for the loss. But, if a highly-talented inefficient team wins, they don't get credit for the wins. It is great for gamblers to figure out which teams to bet on, but should only be one small factor in determining the Field of 68.
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
The complaint about the NET rankings is they measure which team SHOULD win, not which team actually wins. So, if you play efficiently and lose you don't get docked for the loss. But, if a highly-talented inefficient team wins, they don't get credit for the wins. It is great for gamblers to figure out which teams to bet on, but should only be one small factor in determining the Field of 68.
Spot on


and maybe it is a small factor...which is then fine....but the ncaa advertises it like it’s the biggest of the metric (even if the metric only accounts for 15-20% of their decision making)
 

UpAndUnder43

Well-known member
Joined
Jan 3, 2017
Messages
6,102
Reaction score
1,809
Points
113
The first NET rankings to me are unimportant at best. Like preseason polls. If you're good you'll be fine. If you're bad you won't be. Net rankings maybe factors into 2-3 at large bids but only after a lot of other things are considered. In my opinion at least.
 

alchemy2u

Well-known member
Joined
Nov 13, 2009
Messages
4,495
Reaction score
715
Points
113
KenPom and Net do some things better than RPI


RPi is a better measure of achievement.
Once you have a large enough sample of games (I’d argue you never have a large enough sample when you only play 30 games and there are 300+ teams) at some point efficiency based metrics become better predictors than RPI.


I think we should measure teams based on achievement rather than based on how we think they might hypothetically achieve moving forward
It would be interesting to measure how accurate the NET ranking is in predicting wins. Isn't that they purpose? To determine which team would do better in a game? I would bet that the NET is awful in predicting the winner in those obvious game where it look wrong.
 

Some guy

Well-known member
Joined
Aug 1, 2014
Messages
8,368
Reaction score
1,622
Points
113
It would be interesting to measure how accurate the NET ranking is in predicting wins. Isn't that they purpose? To determine which team would do better in a game? I would bet that the NET is awful in predicting the winner in those obvious game where it look wrong.
I think Kenpom and other efficiency rankings are meant to be predictive.

net is heavily weighted towards efficiency but I’m not even sure they release their exact formula
 

UpAndUnder43

Well-known member
Joined
Jan 3, 2017
Messages
6,102
Reaction score
1,809
Points
113
No, I said they were #119 and posted that source. You must have me confused with someone else.
Oh, it said "thats not what I'm seeing" with a big bold #7 but you had quoted them. My bad.
 

cjbfbp

Well-known member
Joined
Feb 23, 2014
Messages
7,384
Reaction score
1,984
Points
113
The complaint about the NET rankings is they measure which team SHOULD win, not which team actually wins. So, if you play efficiently and lose you don't get docked for the loss. But, if a highly-talented inefficient team wins, they don't get credit for the wins. It is great for gamblers to figure out which teams to bet on, but should only be one small factor in determining the Field of 68.
Great point. I agree that NCAA tournament selection should be because of achievement based on the team's record, not on point spreads. Selection is a reward for a good season based on wins and strength of schedule. A team with an 18-14 record shouldn't get in because they beat the hell out of multiple weaker teams and lost respectably to many better ones. Let them go to the NIT.
 

howeda7

Well-known member
Joined
Nov 22, 2008
Messages
49,918
Reaction score
7,683
Points
113
Not sure who is doing data entry at yahoo. But the ncaa official net on their website has MSU at 119

duke at 115
I see that. Yahoo is way off but right for some teams...
 

howeda7

Well-known member
Joined
Nov 22, 2008
Messages
49,918
Reaction score
7,683
Points
113
Ironically, the Gophers may have had a higher Net ranking one year ago, with a record of ~7-5 than they do now....
 

WhoFellDownTheGopherHole?

Well-known member
Joined
Feb 11, 2011
Messages
3,416
Reaction score
2,309
Points
113
It's been a while since we've had the luxury of thinking about it, but isn't this what we want - don't even the "good" Gopher teams always play the disrespected underdog card/role?

Awwww...it's just like the good ol' days. :love:
 
Top Bottom