NET Rankings are Flawed

[NoCreativity]

Hawkeyenation is dead. If you guys want to see some real comedy go to Hawkeyereport.

If you can sift through all the un-moderated Covid, Political, and people dropping F-bomb talk there's actually people on there that think both Murray's are coming back next year and Virginia and Utah State were signature wins.

 

[cykadelic2]

People can complain about Iowa's ranking in NET all they want, but they are ranked about the same in Torvik and Kenpom as well. It seems most power rankings like there numbers for whatever reason.

All 3 of those use a net efficiency component which rewards margin of victory and compared to ISU, TOE has more wins with big MOVs.

Previously posted that since NET was implemented, the Selection Committee's final seeding and selection have correlated most closely with ESPN's SOR rankings which doesn't include MOV. Also, NET is primarily used as a sorting tool to ID Quadrant wins and losses and those results are far more relevant than the NET rankings themselves.

 

[NorthCyd]

All 3 of those use a net efficiency component which rewards margin of victory and compared to ISU, TOE has more wins with big MOVs.

Previously posted that since NET was implemented, the Selection Committee's final seeding and selection have correlated most closely with ESPN's SOR rankings which doesn't include MOV. Also, NET is primarily used as a sorting tool to ID Quadrant wins and losses and those results are far more relevant than the NET rankings themselves.

Yeah, I don't really know how all the different models work, but that makes sense.

I get what you are saying about the NET rankings just sorting the teams to categorize what tier a team belongs in, but I do find the fundamental logical flaw in that system amusing. You take a team like Iowa that will award other teams a T1 victory for their high ranking. But then come seeding time you throw that ranking out the window saying they really aren't that good and they get a lower seed for not having any T1 victories. Yet teams who beat Iowa still get a T1 victory for beating a team that was overranked.

 

[cykadelic2]

Yeah, I don't really know how all the different models work, but that makes sense.

I get what you are saying about the NET rankings just sorting the teams to categorize what tier a team belongs in, but I do find the fundamental logical flaw in that system amusing. You take a team like Iowa that will award other teams a T1 victory for their high ranking. But then come seeding time you throw that ranking out the window saying they really aren't that good and they get a lower seed for not having any T1 victories. Yet teams who beat Iowa still get a T1 victory for beating a team that was overranked.

I agree that if NET is overly weighted for net efficiency/MOV. then perhaps it shouldn't be utilized as the Committee's sorting tool for Quad analysis. But I think that concern is mitigated due to recent correlation of final selection/seeding primarily to SOR and KPI metrics which don't incorporate net efficiency/MOV.

BTW, ISU is now #16 (4 seed) in ESPN SOR and TOE is #40 (10 seed). I see Lunardi now has both as 7 seeds which is complete nonsense given recent Committee seeding behavior.

 

[qwerty]

I agree that if NET is overly weighted for net efficiency/MOV. then perhaps it shouldn't be utilized as the Committee's sorting tool for Quad analysis. But I think that concern is mitigated due to recent correlation of final selection/seeding primarily to SOR and KPI metrics which don't incorporate net efficiency/MOV.

BTW, ISU is now #16 (4 seed) in ESPN SOR and TOE is #40 (10 seed). I see Lunardi now has both as 7 seeds which is complete nonsense given recent Committee seeding behavior.

Maybe they are following WarrenNolan.com. He predicts Iowa only loses one more rest of season . . . . Yeah, that's not happening.

 

[CyLyte2]

Hawkeyenation is dead. If you guys want to see some real comedy go to Hawkeyereport.

If you can sift through all the un-moderated Covid, Political, and people dropping F-bomb talk there's actually people on there that think both Murray's are coming back next year and Virginia and Utah State were signature wins.

Ironic. There are people here who think the same.

 

[isuno1fan]

I was told this thread wasn't butt hurt about Iowa's NET (or other ranking).

 

[VeloClone]

I was told this thread wasn't butt hurt about Iowa's NET (or other ranking).

I guess I think there is a difference between "butt hurt" and "laughing at." Personally, I am just laughing.

 

[CY88CE11]

I guess I think there is a difference between "butt hurt" and "laughing at." Personally, I am just laughing.

Also, there's a proximity bias at work. Iowa is easy to pick out because we know them, we beat them, our local reporters tell us how good they are for only losing to good teams by 12. So, we know they're clearly not a top 25 team based on their resume, and when NET shows them there, it jumps off the page.

 

[VeloClone]

Also, there's a proximity bias at work. Iowa is easy to pick out because we know them, we beat them, our local reporters tell us how good they are for only losing to good teams by 12. So, we know they're clearly not a top 25 team based on their resume, and when NET shows them there, it jumps off the page.

We have had people call foul when there is a list posted and it includes Big 12 schools/programs and Iowa. "LOL, when did Iowa join the Big 12?" Maybe look at the list as schools/programs we compete against every year...

 

[atlantacyclone]

That happens when you quickly lose 5 games…

 

[cyclones500]

Yeah, I don't really know how all the different models work, but that makes sense.

I get what you are saying about the NET rankings just sorting the teams to categorize what tier a team belongs in, but I do find the fundamental logical flaw in that system amusing. You take a team like Iowa that will award other teams a T1 victory for their high ranking. But then come seeding time you throw that ranking out the window saying they really aren't that good and they get a lower seed for not having any T1 victories. Yet teams who beat Iowa still get a T1 victory for beating a team that was overranked.

I gave this a Like, because I can see the supposed contradiction of it. I do think as we're closer to season's end there "should" be fewer outlier cases like that. (I say "should" since no system is flawless.) The "weights" should become more accurate by that point.

If we had a perfect analytic for selection, the committee could just look at the final NET ranking and use it as the S-curve and we wouldn't need the quadrants. (It would be redundant). Alas, we don't...

 

[awd4cy]

So now we have Iowa 1 spot ahead of us even though we are 6-4 in Q1 games and Iowa is 1-5. Makes absolutely no sense.

 

[EnkAMania]

So now we have Iowa 1 spot ahead of us even though we are 6-4 in Q1 games and Iowa is 1-5. Makes absolutely no sense.

If Iowa and Creighton slip, we are down 2 quad 1 wins

 

[clone52]

So now we have Iowa 1 spot ahead of us even though we are 6-4 in Q1 games and Iowa is 1-5. Makes absolutely no sense.

The must REALLY love efficiency ratings.

Don't worry, though. While NET did really well on on picking teams in the tournament, they didn't do great on seeds.

OSU was a 2 seed, but net says 4 seed.
Loyola was an 8 seed, but net said 3 seed.
Villanova was a 5 seed, but net said 3 seed.
Texas was a 3 seed, but net said 5 seed.
WVU was a 3 seed, but net said 6 seed.
UCLA was an 11 seed, but net said a 4 seed!!!

 

[isutrevman]

If Iowa and Creighton slip, we are down 2 quad 1 wins

You could make that same argument for any other team.

 

[Cyclones125]

I'm only referencing this because its all I have access to, but if you go back through ESPN's BPI we had a bunch of very low game scores in victories against crap teams during non-conference. For example, we actually got a better rating in our blowout home loss to TCU than our victory against Alabama State. I know its not the same rating system as NET but there's enough similarities to where this is probably why NET doesn't like us either.

 

[AuH2O]

If Iowa and Creighton slip, we are down 2 quad 1 wins

Don't worry, here's how it will play out:
- Ohio State is up 92-72 with 3 minutes to play and OSU empties the bench
- The Murrays and Toussaint press the hell out of those walk-ons, JorBo hits a few threes
- Iowa loses 96-87
- Iowa goes up 12 spots in the NET because of that offensive efficiency

 

[FriendlySpartan]

Is all this net concern due to thinking you might not make the tourny? Because I wouldn't be worried about that with your resume. As mentioned above NET rankings are meaningless for seeding and thus kinda meaningless in general. You guys are having an awesome year with a new coach, enjoy it and don't worry about what the computer says there is a committee for a reason not just an algorithm.

 

[isuno1fan]

Is all this net concern due to thinking you might not make the tourny? Because I wouldn't be worried about that with your resume. As mentioned above NET rankings are meaningless for seeding and thus kinda meaningless in general. You guys are having an awesome year with a new coach, enjoy it and don't worry about what the computer says there is a committee for a reason not just an algorithm.

It's due to butt hurt about Iowa.