OT / General Interest: NCAA Bubble and Tourney Impact Games March 9 and 10

[jncuse]

Now we just need someone to do as you and many of us agreed - create a hybrid ranking 50 pct NET and 50 pct RPI then let’s see what it looks like. Simply the quadrants to home, away and neutral with records vs top 50, top 100, top 150 and then 150 plus as means to look at performance against the rest of the country.

I wouldn’t change the quad system back to top 50, top 100. I would keep it the same as now — I don’t see it as the problem here. The underlying rankings that feed the quad groups are.

The quad system to me is much better now — road games are much harder and I am fine with the quads being extended or shortened for that. Also allows for more q1 and q2 oops for mid majors via road games — although that has been negated by power conferences getting more opps as well.

I would change the factor in NET that seemingly gives extra value to road wins. Inherently efficiency systems already adjust for road / home to begin with, so it’s double counting in my view — my guess is the NET expands the value of margin variances from expectation on the road by 40%. Bring that back to flat.

But 50/50 would solve lots of things. I won’t run all the quad records but I will show what the new NET/RPI hybrid would look like this year.

 

[longtimefan]

Team A - Michigan St NET #24 (Tourney Lock)
Team B - Indiana NET #97 (Needs to win B10)

Michigan St because of the Baylor win is clearly above them in terms of selection. That is fine - but the gap should not be huge

NET is doing crazy things. This is two teams from the same conference, same W-L record, with similar level of quality victories and no stupid losses.

Don't feel bad for Indiana, but just pointing it out.
They don't have a bad resume by the traditional analysis, but getting NET crushed. Opposite

I knew the answer--only because CBS ran a side-by-side graphic prior to yesterday's MSU-IU game. It was jarring to see Mich St with ~30 NET and Indiana mired in the 90's.

Dan Bonner had a disparaging comment on the NET, specifically the use of efficiency data.

 

[Orangezoo]

I wouldn’t change the quad system back to top 50, top 100. I would keep it the same as now — I don’t see it as the problem here. The underlying rankings that feed the quad groups are.

The quad system to me is much better now — road games are much harder and I am fine with the quads being extended or shortened for that. Also allows for more q1 and q2 oops for mid majors via road games — although that has been negated by power conferences getting more opps as well.

I would change the factor in NET that seemingly gives extra value to road wins. Inherently efficiency systems already adjust for road / home to begin with, so it’s double counting in my view — my guess is the NET expands the value of margin variances from expectation on the road by 40%. Bring that back to flat.

But 50/50 would solve lots of things. I won’t run all the quad records but I will show what the new NET/RPI hybrid would look like this year.

No disagreement and accomplishes the same thing if you remove the weighting for road wins in the NET

 

[jncuse]

(In Progress)

Houston 1,2 = 1.5
Purdue 2, 1 = 1.5
UConn 3, 5 = 4
Tennessee 5.3 = 4
Arizona 4,8 = 6
UNC 7,6 = 6.5
Auburn 6,13 = 9.5
Alabama 8.11 = 9.5
Iowa St 9,14 = 11.5
Dayton 21, 4 = 12.5
Marquette 13, 15 = 14
Kansas 18, 12 = 15
San Diego St 20, 10 = 15
Baylor 14. 17 = 15.5
Creighton 11,22 = 16.5
Duke 10, 25 = 17.5
Illinois 15, 20 = 17.5
Boise St 23, 26 = 24.5
Clemson 26, 23 = 24.5
Kentucky 19, 32 = 25.5
Wisconsin 22, 29 = 25.5
BYU 12, 42 = 27
Gonzaga 16, 38 = 27
Colorado 27, 31 = 29
St Mary's 17, 52 = 34.5
Texas 24, 51 = 37.5
Michigan St 25,60 = 42.5

 

[orangenauburn]

(In Progress)

Houston 1,2 = 1.5
Purdue 2, 1 = 1.5
UConn 3, 5 = 4
Tennessee 5.3 = 4
Arizona 4,8 = 6
UNC 7,6 = 6.5
Auburn 6,13 = 9.5
Alabama 8.11 = 9.5
Iowa St 9,14 = 11.5
Dayton 21, 4 = 12.5
Marquette 13, 15 = 14
Kansas 18, 12 = 15
San Diego St 20, 10 = 15
Baylor 14. 17 = 15.5
Creighton 11,22 = 16.5
Duke 10, 25 = 17.5
Illinois 15, 20 = 17.5
Boise St 23, 26 = 24.5
Clemson 26, 23 = 24.5
Kentucky 19, 32 = 25.5
Wisconsin 22, 29 = 25.5
BYU 12, 42 = 27
Gonzaga 16, 38 = 27
Colorado 27, 31 = 29
St Mary's 17, 52 = 34.5
Texas 24, 51 = 37.5
Michigan St 25,60 = 42.5

What if Syracuse beats NC State (Q2?), Duke (Q1), Virginia (Q1?), and lose close to UNC (Q1)

Or beats NC State, Duke, lose to Virginia

 

[bpo57]

(In Progress)

Houston 1,2 = 1.5
Purdue 2, 1 = 1.5
UConn 3, 5 = 4
Tennessee 5.3 = 4
Arizona 4,8 = 6
UNC 7,6 = 6.5
Auburn 6,13 = 9.5
Alabama 8.11 = 9.5
Iowa St 9,14 = 11.5
Dayton 21, 4 = 12.5
Marquette 13, 15 = 14
Kansas 18, 12 = 15
San Diego St 20, 10 = 15
Baylor 14. 17 = 15.5
Creighton 11,22 = 16.5
Duke 10, 25 = 17.5
Illinois 15, 20 = 17.5
Boise St 23, 26 = 24.5
Clemson 26, 23 = 24.5
Kentucky 19, 32 = 25.5
Wisconsin 22, 29 = 25.5
BYU 12, 42 = 27
Gonzaga 16, 38 = 27
Colorado 27, 31 = 29
St Mary's 17, 52 = 34.5
Texas 24, 51 = 37.5
Michigan St 25,60 = 42.5

You're reminding me how much I hated the RPI. KU 12? Dayton 4? Yikes.

 

[jncuse]

What if Syracuse beats NC State (Q2?), Duke (Q1), Virginia (Q1?), and lose close to UNC (Q1)

Or beats NC State, Duke, lose to Virginia

Scenario #1 - I would suspect we get in, no matter what the score is against UNC.

Scenario #2 - Maybe. Depends on a number of factors - whats going on around us, what the committee thinks about us. But I would guess miss is more likely by some degree.

The good thing about beating Duke on Thursday, if we do, is that we get in the minds of the committee early in the process.

 

[jncuse]

You're reminding me how much I hated the RPI. KU 12? Dayton 4? Yikes.

They are both imperfect

The NET can be abused by Q4 blowouts where teams don't let up against 300+ cupcakes.
The RPI can be abused if you can avoid Q4 level games and find easily winnable Q3 games instead. But play a number of 280+ teams and you will get dinged.

That is why I think combining them would be better moving forward, because they reward opposite things- hate the fact that real cupcake games are becoming more and more prevalent in OOC.

 

[bpo57]

They are both imperfect

The NET can be abused by Q4 blowouts where teams don't let up against 300+ cupcakes.
The RPI can be abused if you can avoid Q4 level games and find easily winnable Q3 games instead. But play a number of 280+ teams and you will get dinged.

That is why I think combining them would be better moving forward, because they reward opposite things- hate the fact that real cupcake games are becoming more and more prevalent in OOC.

I hear ya and agree. But that 4 ranking of Dayton is one of the craziest things I've ever seen from one of these ranking services.

 

[jncuse]

I hear ya and agree. But that 4 ranking of Dayton is one of the craziest things I've ever seen from one of these ranking services.

They shouldn't be #4. The NET seems to do a much better job at identifying the top 10 than RPI -- but its start to get wacky around #25.

Looking at the RPI components its driven by their schedule ranking which seems abnormally high at #22 for an A-10 team (in part because of their #9 ranking for OOC).

- 5 losses only (only 3 teams from good conferences have less) combined with a good SOS gets them up to #4.

Interestingly KP says their schedule is #89 in the country vs #22... much different. And NET tends to follow KP principles, in terms of measuring schedule.

 

[bpo57]

They shouldn't be #4. The NET seems to do a much better job at identifying the top 10 than RPI -- but its start to get wacky around #25.

Looking at the RPI components its driven by their schedule ranking which seems abnormally high at #22 for an A-10 team (in part because of their #9 ranking for OOC).

- 5 losses only (only 3 teams from good conferences have less) combined with a good SOS gets them up to #4.

Interestingly KP says their schedule is #89 in the country vs #22... much different. And NET tends to follow KP principles, in terms of measuring schedule.

I watch the A-10 more than most as my son graduated from St. Joe's a few years back and my brother went to Loyola many moons ago. Dayton has not distinguished themselves at all in the conference. They never led in their last game (vs VCU) in either half - on their home floor - only to prevail in OT. Meanwhile they lost to Richmond, Loyola, George Mason and VCU all on the road. The A-10 is generally dismissed as a weak conference yet Dayton is given a pass for those conference losses. I know I'm beating a dead horse but that doesn't make sense to me.

 

[SUfanNC]

Men's Bracketology: Duke plus Cooper Flagg makes Blue Devils early overall No. 1

Bracketology begins the off season projections with a familiar name at the top. Duke, led by generational recruit Cooper Flagg, starts our 2024-25 bracket cycle on top.

www.espn.com

Two wins as of now per Joey B. I’m sure it is depending on other teams.

 

[jncuse]

I watch the A-10 more than most as my son graduated from St. Joe's a few years back and my brother went to Loyola many moons ago. Dayton has not distinguished themselves at all in the conference. They never led in their last game (vs VCU) in either half - on their home floor - only to prevail in OT. Meanwhile they lost to Richmond, Loyola, George Mason and VCU all on the road. The A-10 is generally dismissed as a weak conference yet Dayton is given a pass for those conference losses. I know I'm beating a dead horse but that doesn't make sense to me.

Not trying to give any hype to Dayton, but those 4 road losses, even in the A-10, are still very "protected" metric wise in that they are 71, 76, 84, 85. All Q2, and very close to Q1. If you get enough good wins and wins in general they will not really be used against you.

The current system "protects" road losses in big conferences and even in good mid majors. - its very hard to get a bad loss on the road in a power conference. ACC only has 1 bad loss opportunity on the road... Big East 2, SEC 2, B12 1, BIG 0.

Even in the A-10, 9 of its 15 members are protected as Q1 or Q2 road losses.

 

[jncuse]

Men's Bracketology: Duke plus Cooper Flagg makes Blue Devils early overall No. 1

Bracketology begins the off season projections with a familiar name at the top. Duke, led by generational recruit Cooper Flagg, starts our 2024-25 bracket cycle on top.

www.espn.com

Two wins as of now per Joey B. I’m sure it is depending on other teams.

Ironically 2 wins get us back to where we were after the Clemson game, where some said beat Clemson and then do something of quality in the ACC. The NC St game at least would be a Q2 game.

And I'm still left in the same conclusion if we beat Duke or Clemson -- it depends. Depends on others, depends on the committee.

 

[baggerbob]

Ironically 2 wins get us back to where we were after the Clemson game, where some said beat Clemson and then do something of quality in the ACC. The NC St game at least would be a Q2 game.

And I'm still left in the same conclusion if we beat Duke or Clemson -- it depends. Depends on others, depends on the committee.

Beating someone 3 times is hard especially with this team.
Expect Judah to play hero ball and a lot of one on one.
Would have preferred playing Louisville, rather then NC St.

 

[bpo57]

Not trying to give any hype to Dayton, but those 4 road losses, even in the A-10, are still very "protected" metric wise in that they are 71, 76, 84, 85. All Q2, and very close to Q1. If you get enough good wins and wins in general they will not really be used against you.

The current system "protects" road losses in big conferences and even in good mid majors. - its very hard to get a bad loss on the road in a power conference. ACC only has 1 bad loss opportunity on the road... Big East 2, SEC 2, B12 1, BIG 0.

Even in the A-10, 9 of its 15 members are protected as Q1 or Q2 road losses.

And yet Loyola only has two road losses in the conference and beat Dayton head to head.