Net Rankings and Team Data sheets

[vtpmt81]

https://extra.ncaa.org/solutions/rpi/Stats Library/NET Team Sheets - Games through March 6, 2019.pdf

As discussed before the season RPI is nowhere to be seen. Looks like the NCAA Evaluation Tool (NET), Kevin Pauga Index (KPI), Strength of Record (SOR), Basketball Power Index (BPI), Pomeroy (POM), and Sagarin (SAG) rankings are given for each team.

Pretty interesting stuff. Seems like some Quadrant 1 wins are separated from others according to these data sheets. Going to be pretty interesting to see if that or bad losses are what causes bubble teams to get left out.

 

[GE Nole]

https://extra.ncaa.org/solutions/rpi/Stats Library/NET Team Sheets - Games through March 6, 2019.pdf

As discussed before the season RPI is nowhere to be seen. Looks like the NCAA Evaluation Tool (NET), Kevin Pauga Index (KPI), Strength of Record (SOR), Basketball Power Index (BPI), Pomeroy (POM), and Sagarin (SAG) rankings are given for each team.

Pretty interesting stuff. Seems like some Quadrant 1 wins are separated from others according to these data sheets. Going to be pretty interesting to see if that or bad losses are what causes bubble teams to get left out.

Yeah, this year they have separated out “tier a” quad 1 wins and “tier b”

 

[lerario]

So what does NET actually measure? For instance, UNC is rated higher across the board by every Index and power ranking, and has way better scheduling numbers than UK, Tenn, and Houston, but is rated behind all of them in NET. What is NET capturing or emphasizing that the others aren’t?

 

[GE Nole]

The NCAA refuses to make their actual formula public. But generally speaking, it measures who you played, where played them, and how many points you beat them by (or lost by).

This is from December, so the example rankings are old. But the explanation is accurate. Will answer most of your questions.

https://www.google.com/amp/s/www.to...-tournament-florida-state-seminoles-fsu-hoops

 

[Montana81]

I don’t think Gonzaga or Houston have any business being in the discussion for a 1 seed but here we are.

Tennessee resume is still lacking too tbh.

Strength of schedule doesn’t seem to be enough of a factor.

Looking at the date sheets it should be Virginia, duke, Kentucky and UNC. Michigan State is the other team I had to look at but more of their quad 1 wins are tier B than the other 4.

 

[lerario]

I don’t think Gonzaga or Houston have any business being in the discussion for a 1 seed but here we are.

Tennessee resume is still lacking too tbh.

Strength of schedule doesn’t seem to be enough of a factor.

Looking at the date sheets it should be Virginia, duke, Kentucky and UNC. Michigan State is the other team I had to look at but more of their quad 1 wins are tier B than the other 4.

I agree on the Zags due to their pathetic schedule but the fact that all of the ratings that adjust for opponent still have them so high will lock them in.

 

[hailtoyourvictor]

So what does NET actually measure? For instance, UNC is rated higher across the board by every Index and power ranking, and has way better scheduling numbers than UK, Tenn, and Houston, but is rated behind all of them in NET. What is NET capturing or emphasizing that the others aren’t?

I don’t know for sure, but rating systems like Kenpom are more of a predictive model trying to show how good a team is / should be. NET (and formerly to a much greater extent RPI) aim to measure body of work and resume.

 

[Montana81]

Kentucky actually has a pretty strong case as the 2nd 1 seed. Them, duke and UNC all are 7-3 in tier a games but Kentucky has more 4 more quad 1 wins in tier b where as duke and UNC only have 1. Though at least 1 of duke and UNC will get 1 more tier A win before the end of the season. Kentucky will likely get 1 more tier b wins. Florida is sitting at 32 right now and even if they lose to us on Saturday they’ll probably move up

 

[BeerPoisoning]

So what does NET actually measure? For instance, UNC is rated higher across the board by every Index and power ranking, and has way better scheduling numbers than UK, Tenn, and Houston, but is rated behind all of them in NET. What is NET capturing or emphasizing that the others aren’t?

NET is trash... If you’re asking my opinion. It basically tosses out human-conscious opinion and makes everything metrical. For example, it weighs in off/def efficiency. Let’s look at the UNC/Virginia game.....

EYE Test: Within 3 points with under 2 minutes to play. You would say that either team could have won and that it was a competitive match-up.

NET Test: Wow, UNC only scored 23 field goals in 80 possessions and allowed Virginia to shoot 55% for the game. It was a home game. What a terrible performance!

Tons of other examples to be made and if you’re looking for all the technical measures, I’m sure Google won’t let you down. But that is basically how I sum it up. I think that advanced stats and metrics are cool to look at. But viewing a painting and watching a painting be done are two different stories.

 

[GE Nole]

NET is trash... If you’re asking my opinion. It basically tosses out human-conscious opinion and makes everything metrical. For example, it weighs in off/def efficiency. Let’s look at the UNC/Virginia game.....

EYE Test: Within 3 points with under 2 minutes to play. You would say that either team could have won and that it was a competitive match-up.

NET Test: Wow, UNC only scored 23 field goals in 80 possessions and allowed Virginia to shoot 55% for the game. It was a home game. What a terrible performance!

Tons of other examples to be made and if you’re looking for all the technical measures, I’m sure Google won’t let you down. But that is basically how I sum it up. I think that advanced stats and metrics are cool to look at. But viewing a painting and watching a painting be done are two different stories.

NET doesn’t care about your FG%. The efficiency is just how many points divided by how many possessions. So 70 points scored in a 70 possession game equals 1 point per possession for the efficiency.

The biggest problem by far with NET is that the efficiency data isn’t adjusted for opponent strength. So beating Wake 85-50 boosts you more than beating UNC 73-72.

 

[GE Nole]

Kentucky actually has a pretty strong case as the 2nd 1 seed. Them, duke and UNC all are 7-3 in tier a games but Kentucky has more 4 more quad 1 wins in tier b where as duke and UNC only have 1. Though at least 1 of duke and UNC will get 1 more tier A win before the end of the season. Kentucky will likely get 1 more tier b wins. Florida is sitting at 32 right now and even if they lose to us on Saturday they’ll probably move up

Depends how bad y’all beat them. I don’t think it’s likely UF moves up if y’all win, especially if you win by more than 10.

 

[ThroughBlue]

So what does NET actually measure? For instance, UNC is rated higher across the board by every Index and power ranking, and has way better scheduling numbers than UK, Tenn, and Houston, but is rated behind all of them in NET. What is NET capturing or emphasizing that the others aren’t?

Maybe because we beat North Carolina. Just a guess though

 

[uncfan in ky]

Maybe because we beat North Carolina. Just a guess though

Do you think LSU should be a 1 seed over UK?

 

[uncfan in ky]

Why is the win vs Syracuse & One of the NC state wins for UNC considered a quad 2 win?

 

[lerario]

Maybe because we beat North Carolina. Just a guess though

Lol

 

[EvilMonkeyInTheCloset]

Iowa fans are getting worried that if they lose their next two games, they will be out of the NCAA tournament.

Too many outside factors would have to unmiraculously occur for Iowa to miss the tournament.....

 

[GE Nole]

Why is the win vs Syracuse & One of the NC state wins for UNC considered a quad 2 win?

Probably because you played them at home. The location of the game matters for what qualifies in each quadrant. Quad 1 games have to be a team ranked 1-30 at home but only 1-75 on the road.

 

[iowahawkeyes69]

Our sheet is awful outside of 7-0 vs quad 2. 4 quad 1 wins but 2 vs the worst 2 teams in the BIG. How the hell are Indiana and Penn State quad 1 wins. So that counts the same as Virginia Tech beating Duke? That’s stupid

 

[uncfan in ky]

Our sheet is awful outside of 7-0 vs quad 2. 4 quad 1 wins but 2 vs the worst 2 teams in the BIG. How the hell are Indiana and Penn State quad 1 wins. So that counts the same as Virginia Tech beating Duke? That’s stupid

I agree. Ridiculous to count UNC beating Gonzaga at home & UK winning at Missouri as the same type of win.

 

[GE Nole]

Our sheet is awful outside of 7-0 vs quad 2. 4 quad 1 wins but 2 vs the worst 2 teams in the BIG. How the hell are Indiana and Penn State quad 1 wins. So that counts the same as Virginia Tech beating Duke? That’s stupid

It actually doesn't. The Quad one is broken up into two tiers. Beating a team in the top 15 (believe) is a Tier A, while other Quad one wins are Tier B.

 

[Montana81]

If the resumes are close head to head should matter. Which is why I’d give duke the nod over Uk and Uk the nod over UNC.

Especially when you consider duke has only lost 1 game at full strength.

That said I think all 3 of those teams have a case over the other 2. Hopefully it just works itself out over the next 10 days. But odds are someone at the top is gonna have a reasonable argument for why they should have been seeded where instead of what they were given.

I just don’t think Tennessee has a very strong argument for a 1 seed over either of UNC or Kentucky at this point. That’s my biggest beef with a lot of bracket projections at the moment. Well, that and gonzaga somehow being a lock for a 1 despite an awful schedule and fewer quad 1 wins than most of the top 25

 

[iowahawkeyes69]

It actually doesn't. The Quad one is broken up into two tiers. Beating a team in the top 15 (believe) is a Tier A, while other Quad one wins are Tier B.

I did not realize that. Thank you

 

[Montana81]

Depends how bad y’all beat them. I don’t think it’s likely UF moves up if y’all win, especially if you win by more than 10.

They moved up 3 spots after losing to a lower rated LSU at home. Seems reasonable that they would move up 2 even in a loss to a higher ranked UK on the road.

Though you’re right it could depend on how bad we beat them if we beat them. They’ll be desperate.

 

[Wildcat-in-STL]

I agree. Ridiculous to count UNC beating Gonzaga at home & UK winning at Missouri as the same type of win.

And to think...... UK easily handling UNC is counted the same as UNC beating 18-12 Clemson

 

[BeerPoisoning]

NET doesn’t care about your FG%. The efficiency is just how many points divided by how many possessions. So 70 points scored in a 70 possession game equals 1 point per possession for the efficiency.

The biggest problem by far with NET is that the efficiency data isn’t adjusted for opponent strength. So beating Wake 85-50 boosts you more than beating UNC 73-72.

Efficiency is indirectly related to FG%
Look up offensive efficiency rankings and then look up 2PT / 3PT% — They don’t identically lineup but it’s a very similar list. You score points by making field goals so they go hand in hand. Free throws and turnovers are the only factors that alter the difference.

Actually NET doesn’t give a boost for blowout wins, margin is capped at 10. So that 85-50 win is virtually 60-50.

 

[GE Nole]

Efficiency is indirectly related to FG%
Look up offensive efficiency rankings and then look up 2PT / 3PT% — They don’t identically lineup but it’s a very similar list. You score points by making field goals so they go hand in hand. Free throws and turnovers are the only factors that alter the difference.

Actually NET doesn’t give a boost for blowout wins, margin is capped at 10. So that 85-50 win is virtually 60-50.

Completely false on the margin of victory. Yes, that specific component (MOV) is capped at 10.

Efficiency, which is a completely separate component, is uncapped. If you win 100-0, your net efficiency will be off the charts. This is why NCSU is in the mid 30s on NET but down around 90 on RPI. They crushed their cupcakes (which were super cakey). Frankly, it seems kind of redundant to have both MOV and net efficiency as separate components, but that’s the ncaa for you.

As for the first part, free throws are a pretty big part. Some teams have 25+ free throws in a game. That’s a lot of extra points. Offensive rebounds also help. Bottom line is, it’s points per possession scored and points per possession allowed. Doesn’t matter if a team shoots 30% or 60%.

 

[lurkeraspect84]

UK now leads with 11 Quad 1 wins. Virginia, Michigan State, LSU, and Kansas are tied at 10.

 

[ThroughBlue]

Do you think LSU should be a 1 seed over UK?

Do you?

 

[uncfan in ky]

Do you?

It’s not polite to answer a question with a question.

 

[sgrooms]

It’s not polite to answer a question with a question.

It's Socratic.

 

[mikeinknoxville]

Good discussion

I admit I don't follow the rankings and how they are determined.

But I am for waiting until the end of the conference tournaments before weighing in.

The news out of LSU could factor in.

As a UT fan all I can think about right now is Auburn.

 

[ThroughBlue]

It’s not polite to answer a question with a question.

It’s not polite to compare apples to oranges

 

[uncfan in ky]

It’s not polite to compare apples to oranges

Let’s see, LSU has the same record as UK, LSU has a better conference record than UK & LSU beat UK in Rupp.
Again, do you think LSU deserves a better seed than UK?

 

[lurkeraspect84]

Let’s see, LSU has the same record as UK, LSU has a better conference record than UK & LSU beat UK in Rupp.
Again, do you think LSU deserves a better seed than UK?

In the SECT sure.

 

[uncfan in ky]

In the SECT sure.

But why not the NCAAT?

 

[lurkeraspect84]

I thought it should be obvious, but their only ranked/best win OOC was against #24 Furman. Their only OOC Quad 1 win was against St. Mary's.

They've earned the higher SECT seed, but not the NCAAT.

 

[ThroughBlue]

Let’s see, LSU has the same record as UK, LSU has a better conference record than UK & LSU beat UK in Rupp.
Again, do you think LSU deserves a better seed than UK?

UNC starts getting their pub when they beat a Zionless duke. If dukes losses don’t hurt them without Zion then UNC win shouldn’t help them without Zion.

 

[uncfan in ky]

UNC starts getting their pub when they beat a Zionless duke. If dukes losses don’t hurt them without Zion then UNC win shouldn’t help them without Zion.

What the hell does UNC have to do with my question?