Offical Net ranking thread

This article does a good job explaining it, but the opponent ranking needed for a win or loss to count as Quad 1 vs Quad 2 changes depending on where the game was played. So you shouldn't have to apply any sort of correction and in theory, all Quad 1 wins are better than all Quad 2 wins, regardless of location.

Quadrant 1: Home 1-30, Neutral 1-50, Away 1-75.
Quadrant 2: Home 31-75, Neutral 51-100, Away 76-135.
Quadrant 3: Home 76-160, Neutral 101-200, Away 135-240.
Quadrant 4: Home 161-353, Neutral 201-353, Away 241-353.

Explanation article: https://www.ncaa.com/news/basketbal...-what-know-about-college-basketballs-new-tool
 

By any metric we have two horrible losses to BC and Illinois as it stands now. Those teams play poorly, lose alot and manhandled us.

Illinois has talent. They beat Maryland as well. Boston College does as well, and they have beaten a couple of good teams. They are still teams with talent that can put a good game together. Wisconsin has lost to us, goes entire halves of games scoring less than 15 points, and somehow they are efficient - they just play slow.


Sent from my iPhone using Tapatalk
 


NC State (16-4) is probably the other biggest high major outlier. At least they have a better record than Gators and a win over Auburn (although that is growing less impressive).

NC STATE:

currently: #107 in RPI with a #277 SOS. Wouldn't even sniff an NIT berth in old system.

currently: #29 in NET whipping up on inferior teams and playing with game-long "efficiency"
currently: #32 in KenPom

More examples of the deeply flawed NET rankings.


Sent from my iPhone using Tapatalk
 

So lets take a look at Iowa st vs Minnesota resume

Both are 15-5 (4-3 vs q1, 1-2 vs q2, 4-0 vs q3, and 6-0 vs q4)
Average net win for iowa st is 151 for the gophers 133
Average net loss for iowa st 38 for the gophers 59
strength of schedule iowa st 37 for the gophers 59
Iowa st is 3-1 in nuetral games (0-1 vs quad 1/2) gophers are 4-0 (2-0 vs quad 1/2)
Iowa st is 3-3 on road(all vs quad 1) gophers are 1-4 away (1-2 vs quad 1)

oh btw iowa st has a net rating of 15 while minnesota is 51

so quality road wins, strength of schedule (which really their isnt that big of a difference there)

Goes to show how much importance they really are putting in efficency numbers
 


So lets take a look at Iowa st vs Minnesota resume

Both are 15-5 (4-3 vs q1, 1-2 vs q2, 4-0 vs q3, and 6-0 vs q4)
Average net win for iowa st is 151 for the gophers 133
Average net loss for iowa st 38 for the gophers 59
strength of schedule iowa st 37 for the gophers 59
Iowa st is 3-1 in nuetral games (0-1 vs quad 1/2) gophers are 4-0 (2-0 vs quad 1/2)
Iowa st is 3-3 on road(all vs quad 1) gophers are 1-4 away (1-2 vs quad 1)

oh btw iowa st has a net rating of 15 while minnesota is 51

so quality road wins, strength of schedule (which really their isnt that big of a difference there)

Goes to show how much importance they really are putting in efficency numbers

Lets look closer at the wins and the losses. Gophers do not have a win that even comes close to Kansas or Texas Tech and they have 2 losses far worse than anything Iowa State has. They have those wins over highly efficient teams and avoided losses to teams that are not efficient like BC and Illinois. Most people that are not fans of either would say that right now that Iowa State is the better team.
 

Lets look closer at the wins and the losses. Gophers do not have a win that even comes close to Kansas or Texas Tech and they have 2 losses far worse than anything Iowa State has. They have those wins over highly efficient teams and avoided losses to teams that are not efficient like BC and Illinois. Most people that are not fans of either would say that right now that Iowa State is the better team.

Except according to NET ranking,
Winning @wisconsin is more difficult than beating Kansas.


Which even further demonstrates the stupidity of net
 

You guys are reading WAY WAY WAY too much into who we beat and who we got beat by.

In the NET score, I think it makes very little difference.

What does make a difference is your offensive and defensive efficiency. (or net efficiency).

Iowa scores 118.5 points per 100 possessions.
We score 109.7

Iowa's defense allows 100.7 points per possession.
We allow 97.5.

They have a 17.8 spread between those numbers.
We have a 12.21 point spread in those numbers.

Then you need to adjust for how many games you have beaten an opponent by 10 or more.
Iowa has 10 opponents they beat by more than 10. Move them up above the efficiency rankings a couple spots.
Minnesota only has 6 wins by 10 or more. Probably a little below average for our efficiency ranking. Keep us the same or move down a couple spots.


Our RPI which measures wins against opponents and strength of schedule is better than Iowa's so the strength of schedule or the number of wins against certain opponents is not the reason why Iowa is way above us.
Wins and SOS have almost nothing to do with NET score.
It's efficiency, and then a bonus for beating teams by 10 or more.
 

Lets look closer at the wins and the losses. Gophers do not have a win that even comes close to Kansas or Texas Tech.

WTF are you talking about??

Your beloved NET rankings indicates the Gophers have:

ROAD WIN over #13 NET team (Wis)
HOME WIN over #24 NET team (Neb)
HOME WIN over #30 NET team (IA)

Your beloved NET rankings indicates Cyclones have:

HOME win over #17 NET team (Kan)
ROAD win over #18 NET team (Texas Tech)
HOME win over #35 NET team (Mississippi)


I mean its the ranking system you seem to love and defend, my man.


Most people that are not fans of either would say that right now that Iowa State is the better team.

I think you're right, most fans agree that right now Iowa State is likely the better team.

The point isn't that. The point is that two teams appearing so closely matched on resumes and records and accomplishments so far this year are separated by 36 spots in some predictive/efficiency formula. That formula basically has one team (ISU) sitting around a 4 seed and the other team (MN) barely in the 68-team field, if at all.

Iowa State and Minnesota are both 15-5 with quite similar resumes (Gophers actually have the best win via NET rankings). But, the Cyclones have won "more pretty" so they are 36 spots higher. That's poppycock.

The RPI seems a better indicator as ISU is #25 and Minnesota is #30.
 
Last edited:



You guys are reading WAY WAY WAY too much into who we beat and who we got beat by.

In the NET score, I think it makes very little difference.

What does make a difference is your offensive and defensive efficiency. (or net efficiency).

Iowa scores 118.5 points per 100 possessions.
We score 109.7

Iowa's defense allows 100.7 points per possession.
We allow 97.5.

They have a 17.8 spread between those numbers.
We have a 12.21 point spread in those numbers.

Then you need to adjust for how many games you have beaten an opponent by 10 or more.
Iowa has 10 opponents they beat by more than 10. Move them up above the efficiency rankings a couple spots.
Minnesota only has 6 wins by 10 or more. Probably a little below average for our efficiency ranking. Keep us the same or move down a couple spots.


Our RPI which measures wins against opponents and strength of schedule is better than Iowa's so the strength of schedule or the number of wins against certain opponents is not the reason why Iowa is way above us.
Wins and SOS have almost nothing to do with NET score.
It's efficiency, and then a bonus for beating teams by 10 or more.

Your last two sentences identify the problem. I can see using NET in a seeding situation as a tiebreaker, but not much else.
 


I'm doing a comparison of NET vs Ken Pom vs RPI.

KenPom and NET are very close, but to try to find out what makes them different, here are some outliers.

Here are teams in top 100 NET who have a NET higher than KP and RPI.

NET KP rank RPI Team Conf
75 90 120 USC P12
80 100 127 South Florida Amer
86 97 119 Georgia SEC
89 121 106 Utah P12
94 115 159 Vanderbilt SEC


Four teams in the top 100 that have NET scores better than their RPI or KP scores, that actually reflect closer to their RPI than KP (a rarity) are:

NET KP rank RPI TEAM
45 62 54 Hofstra (11 wins by double digits)
54 69 57 Memphis (8 wins by double digits)
84 127 92 Penn (7 wins by double digits)
99 132 107 North Texas (9 wins by double digits)

What would have these teams ranked higher than KP or their RPI?
Here are their SOS's.

Team - SOS SOS-NC
Hostra - 138 174
Memphis - 311 339
Penn - 235 239
North Texas - 41 50
 

Teams hurt by NET. (Worse NET score than RPI or Ken Pom). SOS and Non-Conf SOS posted as well.


NET KP RPI NCAA SOS SOS-NC
74 60 44 New Mexico St. 115 170
58 50 39 Creighton 8 14
33 25 21 Florida St. 20 32 (9 wins by double digits)
59 51 56 Clemson 24 88
71 63 70 ETSU 208 287



Something about these teams are getting hurt by NET ranking outside of efficiency and RPI factors.
 
Last edited:



You guys are reading WAY WAY WAY too much into who we beat and who we got beat by.

In the NET score, I think it makes very little difference.

What does make a difference is your offensive and defensive efficiency. (or net efficiency).

Iowa scores 118.5 points per 100 possessions.
We score 109.7

Iowa's defense allows 100.7 points per possession.
We allow 97.5.

They have a 17.8 spread between those numbers.
We have a 12.21 point spread in those numbers.

Then you need to adjust for how many games you have beaten an opponent by 10 or more.
Iowa has 10 opponents they beat by more than 10. Move them up above the efficiency rankings a couple spots.
Minnesota only has 6 wins by 10 or more. Probably a little below average for our efficiency ranking. Keep us the same or move down a couple spots.


Our RPI which measures wins against opponents and strength of schedule is better than Iowa's so the strength of schedule or the number of wins against certain opponents is not the reason why Iowa is way above us.
Wins and SOS have almost nothing to do with NET score.
It's efficiency, and then a bonus for beating teams by 10 or more.

I think a really big issue is they are double counting margin of victory.

If efficiency ratings play a big role, then margin of victory already is measured because points per possession measures margin of victory give or take one possession in a game.



The other big issue is that efficiency treats every possession the same. In fact, not all possessions matter equally. Possession in the Iowa and Michigan game matter a lot more than a possession against Santa Clara but those possessions weight equally when measuring efficiency.

Likewise, in a game like Illinois or North Carolina A&T the game wasn’t in doubt the last 10 minutes of a game, but all those possessions matter in efficiency rankings. Basically the moral of the story is beat everyone by 50 when you can because you can fool efficiency rankings.


Even if they have an adjusted points per possession or defensive points per possession, it is just an arbitrary calculation created by someone. And you can say it’s mathematically calculated, but then again the score of the Minnesota Wisconsin game actually happened.
 
Last edited:

WTF are you talking about??

Your beloved NET rankings indicates the Gophers have:

ROAD WIN over #13 NET team (Wis)
HOME WIN over #24 NET team (Neb)
HOME WIN over #30 NET team (IA)

Your beloved NET rankings indicates Cyclones have:

HOME win over #17 NET team (Kan)
ROAD win over #18 NET team (Texas Tech)
HOME win over #35 NET team (Mississippi)


I mean its the ranking system you seem to love and defend, my man.




I think you're right, most fans agree that right now Iowa State is likely the better team.

The point isn't that. The point is that two teams appearing so closely matched on resumes and records and accomplishments so far this year are separated by 36 spots in some predictive/efficiency formula. That formula basically has one team (ISU) sitting around a 4 seed and the other team (MN) barely in the 68-team field, if at all.

Iowa State and Minnesota are both 15-5 with quite similar resumes (Gophers actually have the best win via NET rankings). But, the Cyclones have won "more pretty" so they are 36 spots higher. That's poppycock.

The RPI seems a better indicator as ISU is #25 and Minnesota is #30.

I do not love NET but there is not a perfect system and it is better than RPI. It is not the sole tool used to select teams. I look at Kenpom, the polls and my eyes. The Gophers are hurt in NET by two bad Kenpom losses, poor performances by the non con except for Washington. People really do not understand efficiency and how it spells winning by playing well. People do not understand pace of play either and that has nothing to do with efficiency ppp. Instead of complaining, what is it that you want to use . Can not pick RPI because coaches wanted it out with all its flaws. It was overly generous to wins like Santa Clara and then worse, not recognizing a power conference team that actually was above .500 in conference with several top 25 wins. RPI was deeply flawed but i suppose since likes us, you like it. Sure BC and Illinois have a nice win or two, everyone knows anyone can beat anyone but their body of work is truly awful. For someone who think winning is all that matters then those two teams do not win very often because they play poorly most of the time. I have us as a 7-8 seed right now. Play great, win the rest and we would have a 2 seed. Get to 13-7 and we will be a 5. It was not me that voted for the changes, it was the coaches themselves who wanted good play against good teams rewarded. The way they laid out the quads stink in my opinion.
 

I do not love NET but there is not a perfect system and it is better than RPI. It is not the sole tool used to select teams. I look at Kenpom, the polls and my eyes. The Gophers are hurt in NET by two bad Kenpom losses, poor performances by the non con except for Washington. People really do not understand efficiency and how it spells winning by playing well. People do not understand pace of play either and that has nothing to do with efficiency ppp. Instead of complaining, what is it that you want to use . Can not pick RPI because coaches wanted it out with all its flaws. It was overly generous to wins like Santa Clara and then worse, not recognizing a power conference team that actually was above .500 in conference with several top 25 wins. RPI was deeply flawed but i suppose since likes us, you like it. Sure BC and Illinois have a nice win or two, everyone knows anyone can beat anyone but their body of work is truly awful. For someone who think winning is all that matters then those two teams do not win very often because they play poorly most of the time. I have us as a 7-8 seed right now. Play great, win the rest and we would have a 2 seed. Get to 13-7 and we will be a 5. It was not me that voted for the changes, it was the coaches themselves who wanted good play against good teams rewarded. The way they laid out the quads stink in my opinion.

But our two bad losses and non conf. schedule are already baked in thru the sos, and our sos are very similar - 32 vss. 36.
Also, havent heard much negative about quadrants so would be interesting to hear your thoughts on that.
 
Last edited:

Instead of complaining, what is it that you want to use .

A combination of measures, of course. RPI may not be elegant enough for some purists but it measures the two most fundamental components for selection: winning and quality of opponents. It also has the feature of not relying on initial assumptions. If measures are highly correlated, then using any one of them will do. But, if there is a substantial discrepancy between measures (note the example of North Carolina State in Face the Facts lists), then combined measures should be considered.
 

A combination of measures, of course. RPI may not be elegant enough for some purists but it measures the two most fundamental components for selection: winning and quality of opponents. It also has the feature of not relying on initial assumptions. If measures are highly correlated, then using any one of them will do. But, if there is a substantial discrepancy between measures (note the example of North Carolina State in Face the Facts lists), then combined measures should be considered.

They do use a combination of measures. NET isn't the end all be all. There are six on the team sheets.

NC State's are at least somewhat correlated. Here are their rankings in the six systems used by the selection committee.

ESPN BPI - 21
Sagarin - 27
ESPN Strength of Record - 29
NET - 29
Pomeroy -32
KPI - 62

Not all the ranking systems rely on assumptions. NET doesn't rely on any data from previous years. For the ones that do those initial assumptions are no longer part of the calculation at this point.

https://kenpom.com/contact.php

When is the influence of preseason ratings removed from the system?
As of 2018, it is 73 days from the date of the first game. This ends up being sometime around January 20th. However, the influence of the preseason ratings is gradually reduced between the first week of the season and this point, so that the influence of the preseason ratings is minimal in mid-January.
 

A combination of measures, of course. RPI may not be elegant enough for some purists but it measures the two most fundamental components for selection: winning and quality of opponents. It also has the feature of not relying on initial assumptions. If measures are highly correlated, then using any one of them will do. But, if there is a substantial discrepancy between measures (note the example of North Carolina State in Face the Facts lists), then combined measures should be considered.

And several measures will be used and thank goodness RPI will not be one of them.
 

Plus NC State has the chance to put it all to rest tonight by beating UVA tonight, a top 4 team by any measure.
 


Tuesday's NET ranking.

I’m not actually sure the gophers are much better than 52,

I am certain that based on results to this point they should be rated higher than some of the teams you mentioned.
Or at least near them.
 

They do use a combination of measures. NET isn't the end all be all. There are six on the team sheets.

NC State's are at least somewhat correlated. Here are their rankings in the six systems used by the selection committee.

ESPN BPI - 21
Sagarin - 27
ESPN Strength of Record - 29
NET - 29
Pomeroy -32
KPI - 62

Not all the ranking systems rely on assumptions. NET doesn't rely on any data from previous years. For the ones that do those initial assumptions are no longer part of the calculation at this point.

https://kenpom.com/contact.php

When is the influence of preseason ratings removed from the system?
As of 2018, it is 73 days from the date of the first game. This ends up being sometime around January 20th. However, the influence of the preseason ratings is gradually reduced between the first week of the season and this point, so that the influence of the preseason ratings is minimal in mid-January.

Thanks for clarifying when the initial assumptions are dropped. Your list of ratings for NC State indicates that 5 of them are highly correlated. #2 through #5 are virtually indistinguishable. For all practical purposes, #2 through #5 are all measuring the same thing. If that is the case, using those 6 measures together reduces the influence of #6 to almost nothing. A better system would take one of the first five (or perhaps #1 and one of the other four) and combine that with #6.
 

The issue is relying on efficiency so heavily in a system that is not on an equal playing field. The reason these rating are more reliable in the nba is because the competition is more level (they play the same teams). Also they play 82 games, have a 24 second shot clock, play 48 minutes. Point being there is a lot more data to analyze and the data you get usually actually matters so you get a more accurate depiction of what those numbers mean. Not saying that efficiency is a bad thing, there are just times when its more beneficial than others. Now as far as this rating system goes...lol... Using other metrics to make your own, then also using those same metrics as another differentiator on top of the metric they just made. Sound confusing, yea super redundant and makes zero sense mathematically.
 

The best teams use OE and DE even more intensely per player in combination and who it is against. Look at the top 10 Kenpom and look at the best teams and coaches and they correlate very well. K, Izzo, Belien, Bennett etc.. all believe heavily in efficiency. Then they compare it to common opponent. Look at who throttled BC and look at the efficiency numbers and look at ours. Seeking good shots, taking care of the ball, playing great defense to limit good shots is a fantastic formula to win. Any coach worth his salt aims for the consistentcy of those three things. Rebounding is a no brainer as it ends the possession or keeps it going. Even fans hate watching high turnover basketball and it drives coaches nuts.
 

Thanks for clarifying when the initial assumptions are dropped. Your list of ratings for NC State indicates that 5 of them are highly correlated. #2 through #5 are virtually indistinguishable. For all practical purposes, #2 through #5 are all measuring the same thing. If that is the case, using those 6 measures together reduces the influence of #6 to almost nothing. A better system would take one of the first five (or perhaps #1 and one of the other four) and combine that with #6.

Using 6 metrics would hopefully protect teams from falling victim to an outlier metric. I guess as builtbadger says there is no perfect metric and the coaches asked for this additional one. Also, havent heard any coach complain about NET to this point. Think I personally am to the point where I see no value that these metrics provide to the committee, they're too flawed. However, have never been accused of being the sharpest knife in the drawer, so there is that.
 

The best teams use OE and DE even more intensely per player in combination and who it is against. Look at the top 10 Kenpom and look at the best teams and coaches and they correlate very well. K, Izzo, Belien, Bennett etc.. all believe heavily in efficiency. Then they compare it to common opponent. Look at who throttled BC and look at the efficiency numbers and look at ours. Seeking good shots, taking care of the ball, playing great defense to limit good shots is a fantastic formula to win. Any coach worth his salt aims for the consistentcy of those three things. Rebounding is a no brainer as it ends the possession or keeps it going. Even fans hate watching high turnover basketball and it drives coaches nuts.

Wait, you’re saying you want teams to play well?
 

The best teams use OE and DE even more intensely per player in combination and who it is against. Look at the top 10 Kenpom and look at the best teams and coaches and they correlate very well. K, Izzo, Belien, Bennett etc.. all believe heavily in efficiency. Then they compare it to common opponent. Look at who throttled BC and look at the efficiency numbers and look at ours. Seeking good shots, taking care of the ball, playing great defense to limit good shots is a fantastic formula to win. Any coach worth his salt aims for the consistentcy of those three things. Rebounding is a no brainer as it ends the possession or keeps it going. Even fans hate watching high turnover basketball and it drives coaches nuts.

If a possession leads to an offensive rebound, does that rebound end the possession and start a new one? I only ask because the Gophers are an elite offensive rebounding team, to the point that players aren't afraid to put up shots because they know there's a chance Jordan Murphy or Oturu will come down with it. Hell, Jordan Murphy himself seems willing to throw up a shot just so that he can can grab an offensive board and put back..

Do those type of sequences hurt our offensive efficiency?
 

The best teams use OE and DE even more intensely per player in combination and who it is against. Look at the top 10 Kenpom and look at the best teams and coaches and they correlate very well. K, Izzo, Belien, Bennett etc.. all believe heavily in efficiency. Then they compare it to common opponent. Look at who throttled BC and look at the efficiency numbers and look at ours. Seeking good shots, taking care of the ball, playing great defense to limit good shots is a fantastic formula to win. Any coach worth his salt aims for the consistentcy of those three things. Rebounding is a no brainer as it ends the possession or keeps it going. Even fans hate watching high turnover basketball and it drives coaches nuts.

Pretty much agree with all of this...Its a great coaching tool and it is a fantastic formula to win, but its not the only one....sometimes nitty gritty tough defense get to the foul line, make the game ugly is what you want (this gophers team), many teams play to their competition but just know how to win..all im saying is that sometimes ax+b just equals c and you dont need to do calculus to complicate things and prove it

now how would i do it...quick thought keep the old system and figure out who should be in the field.....there are always a few outliers so lets make everyone happy and decide the last four in by doing some sort efficiency thing.....now once we have the field set we could rank these teams off a combination off few things (coaches rankings, media rankings and efficiency rankings vs field) and just come up with the average...seems fair enough to me
 

Using 6 metrics would hopefully protect teams from falling victim to an outlier metric. I guess as builtbadger says there is no perfect metric and the coaches asked for this additional one. Also, havent heard any coach complain about NET to this point. Think I personally am to the point where I see no value that these metrics provide to the committee, they're too flawed. However, have never been accused of being the sharpest knife in the drawer, so there is that.

You will definitely hear about it on Selection Sunday when some team believing they are deserving gets left out because of NET.

What I have not heard much talk of with the new metric is the old stand-by of mid-majors not getting the chance at high level wins because the big boys won't play them. It was an issue before and may be even more of one now with the power conferences shrinking their out of conference schedules. So far, it looks like the NET highly favors the high majors. My one example is Oklahoma State being a really bad team, but 70 in the NET the last I saw a couple of days ago. That's enough evidence to me that the metric is flawed.
 




Top Bottom