NET Quad system comparison (4 schools)

[JonathanW]

I looked at how 4 schools (UK, Duke, Auburn, Illinois) have done this year in the Quad system, comparing them graphically/visually. One problem with the Quad system is that breaking 362 teams into just 4 groups. Is a Win at #1 the same as a Win at #75? Of course not, but they both count equally (as Q1 wins). Similarly, are Loses at #135 and at #136 really different? Of course not, but using the Quad system they are.

Unfortunately, I can't get the formatting within this page to show what I've done in Excel, but below is my attempt as explaining it.

First of all, we will drop the Q4 games, they all had 5-7 of them, all at home, and they all won.
If you standardize the Neutral and Road games in a way that you can overlay them with the Home games, and then break each Quad in half, then that improves the precision and seems to make things more comparable.

RECORDS vs:

	Top half of Quad 1	Lower half of Quad 1	Top half of Quad 2	Lower half of Quad 2	Top half of Quad 3	Lower half of Quad 3
Team A	2 - 2	3 - 1	5 - 1	1 - 2	4 - 0	3 - 0
Team B	2 - 5	3 - 1	5 - 1	1 - 1	3 - 0	4 - 0
Team C	5 - 2	0 - 4	1 - 1	1 - 0	5 - 1	3 - 0
Team D	1 - 5	0 - 2	7 - 0	1 - 0	6 - 0	3 - 0

So, the question really is, what do you VALUE MORE, quality WINS, or bad LOSES?

UK has the 1 Q3 loss, but it is a top-half Q3 loss, and we do not have any lower-half Q2 losses (all 3 others have at least 1). So, looking at the last 3-4 columns Team C is a little worse than B & D, but not much. And it is arguable if Team C is worse than A since A has 2 lower half Q2 losses.
Next let's look at the first 2 columns. Team C is easily the best in top-half Q1, followed by A; B and D did bad here. But Team C did really bad in lower-half Q1, as did D, where as B and A did well.
Basically, vs very good teams, C has the best chance of winning, followed by A, then B, and D has a bad chance.
But vs "just ok" teams, D and B are the best, followed by A and C.

A=Duke
B=Illinois
C=UK
D=Auburn

 

[FiveStarCat]

My biggest takeaway from this is there is no chance Auburn should be seeded ahead of us at this point. And thats ignoring the fact that we beat them on their home floor. We should be at least a seed line distance from them.

 

[BlueBomb]

OP doin' work.

 

[moses1uk]

Duke should never get a break playing in a joke league .

 

[JwUKFan11]

Duke should never get a break playing in a joke league .

Easy pick to get "upset" in a bracket. Free money.

 

[kyeric]

first thing i see is we squandered a TON of opportunities in the lower half of Q1. Just having the Florida loss back and the UNCW debacle and it's a completely different picture.

 

[Creed Bratton]

first thing i see is we squandered a TON of opportunities in the lower half of Q1. Just having the Florida loss back and the UNCW debacle and it's a completely different picture.

and the A&M loss too.

 

[gharding07]

I looked at how 4 schools (UK, Duke, Auburn, Illinois) have done this year in the Quad system, comparing them graphically/visually. One problem with the Quad system is that breaking 362 teams into just 4 groups. Is a Win at #1 the same as a Win at #75? Of course not, but they both count equally (as Q1 wins). Similarly, are Loses at #135 and at #136 really different? Of course not, but using the Quad system they are.

Unfortunately, I can't get the formatting within this page to show what I've done in Excel, but below is my attempt as explaining it.

First of all, we will drop the Q4 games, they all had 5-7 of them, all at home, and they all won.
If you standardize the Neutral and Road games in a way that you can overlay them with the Home games, and then break each Quad in half, then that improves the precision and seems to make things more comparable.

RECORDS vs:

	Top half of Quad 1	Lower half of Quad 1	Top half of Quad 2	Lower half of Quad 2	Top half of Quad 3	Lower half of Quad 3
Team A	2 - 2	3 - 1	5 - 1	1 - 2	4 - 0	3 - 0
Team B	2 - 4	3 - 0	5 - 1	1 - 1	3 - 0	4 - 0
Team C	5 - 2	0 - 4	1 - 1	1 - 0	5 - 1	3 - 0
Team D	1 - 5	0 - 2	4 - 1	2 - 1	6 - 0	3 - 0

So, the question really is, what do you VALUE MORE, quality WINS, or bad LOSES?

UK has the 1 Q3 loss, but it is a top-half Q3 loss, and we do not have any lower-half Q2 losses (all 3 others have at least 1). So, looking at the last 3-4 columns Team C is a little worse than B & D, but not much. And it is arguable if Team C is worse than A since A has 2 lower half Q2 losses.
Next let's look at the first 2 columns. Team C is easily the best in top-half Q1, followed by A; B and D did bad here. But Team C did really bad in lower-half Q1, as did D, where as B and A did well.
Basically, vs very good teams, C has the best chance of winning, followed by A, then B, and D has a bad chance.
But vs "just ok" teams, D and B are the best, followed by A and C.

A=Duke
B=Illinois
C=UK
D=Auburn

The committee is "supposed" to value quality wins more. And a teams strength of schedule. Who did you play, who did you beat. The committee uses the NET a lot. So, here is what they SHOULD be looking at when comparing those 4 teams.
3 best wins 3 worst losses NET SOS
UK: (5-6 quad 1) #6, 7, 10 #55, 88, 115 #58
Auburn: (1-7 quad 1) #7, 33, 47 #26, 33, 69 #62
Illinois: (5-6 quad 1) #22, 23, 43 #52, 73, 94 #40
Duke: (5-3 quad 1) #13, 22, 24 #44, 120, 124 #74

Trying to be objective, it should be right now: Kentucky, Duke, Illinois, Auburn in seeding those 4.
**Oh and we're 5-2 in quad 1A games, thats 4th best in the country.

T-Rank - Customizable College Basketball Tempo Free Stats - T-Rank

T-Rank College Basketball Ratings and Rankings. And fun.

barttorvik.com

 

[JonathanW]

The committee is "supposed" to value quality wins more. And a teams strength of schedule. Who did you play, who did you beat. The committee uses the NET a lot. So, here is what they SHOULD be looking at when comparing those 4 teams.
3 best wins 3 worst losses NET SOS
UK: (5-6 quad 1) #6, 7, 10 #55, 88, 115 #58
Auburn: (1-7 quad 1) #7, 33, 47 #26, 33, 69 #62
Illinois: (5-6 quad 1) #22, 23, 43 #52, 73, 94 #40
Duke: (5-3 quad 1) #13, 22, 24 #44, 120, 124 #74

Trying to be objective, it should be right now: Kentucky, Duke, Illinois, Auburn in seeding those 4.
**Oh and we're 5-2 in quad 1A games, thats 4th best in the country.

T-Rank - Customizable College Basketball Tempo Free Stats - T-Rank

T-Rank College Basketball Ratings and Rankings. And fun.

barttorvik.com

Like a golfers handicap, which is based on your best rounds, not your worst rounds, essentially on a good day how well could you shoot.

 

[wmk0002]

I'm seeing auburn as 8-0 in Q2... Worst loss is @App St which is #69 (Niiiice) currently. Am I missing something there? Obviously 1-7 Q1 is not good especially when combined with a low level Q2 loss.

 

[The_Answer1313]

I mean I know that I've railed on the ridiculousness of the Quad system.........

I will say this tho.

Each committee member DOES contain a team sheet. This team sheet has all of the results on it on a game by game level. So technically even tho they very much break the records into Quads, they could easily be looking at more than this.

We just don't know. And when you are trying to seed a bunch of teams it makes logical sense for a human to look at simple comparisons vs looking at things on a deeper level. But it's very possible "some" do look at it on a game by game basis. When I do my seed predictions I use the Quads as an initial way to separate teams but then I always look at the actual ranking of these teams they played. I'd assume that at least some others do it similar in that committee room..

The thing about Auburn being ahead of us............honestly maybe they should. This is where humans not looking at things like margin of victory come into play. There's a reason they are ahead of us in NET. Also a reason they would be currently favored on a neutral court over us if we played right this second.

Humans make this process a very flawed one.

 

[Rideordiewildcat]

I looked at how 4 schools (UK, Duke, Auburn, Illinois) have done this year in the Quad system, comparing them graphically/visually. One problem with the Quad system is that breaking 362 teams into just 4 groups. Is a Win at #1 the same as a Win at #75? Of course not, but they both count equally (as Q1 wins). Similarly, are Loses at #135 and at #136 really different? Of course not, but using the Quad system they are.

Unfortunately, I can't get the formatting within this page to show what I've done in Excel, but below is my attempt as explaining it.

First of all, we will drop the Q4 games, they all had 5-7 of them, all at home, and they all won.
If you standardize the Neutral and Road games in a way that you can overlay them with the Home games, and then break each Quad in half, then that improves the precision and seems to make things more comparable.

RECORDS vs:

	Top half of Quad 1	Lower half of Quad 1	Top half of Quad 2	Lower half of Quad 2	Top half of Quad 3	Lower half of Quad 3
Team A	2 - 2	3 - 1	5 - 1	1 - 2	4 - 0	3 - 0
Team B	2 - 4	3 - 0	5 - 1	1 - 1	3 - 0	4 - 0
Team C	5 - 2	0 - 4	1 - 1	1 - 0	5 - 1	3 - 0
Team D	1 - 5	0 - 2	4 - 1	2 - 1	6 - 0	3 - 0

So, the question really is, what do you VALUE MORE, quality WINS, or bad LOSES?

UK has the 1 Q3 loss, but it is a top-half Q3 loss, and we do not have any lower-half Q2 losses (all 3 others have at least 1). So, looking at the last 3-4 columns Team C is a little worse than B & D, but not much. And it is arguable if Team C is worse than A since A has 2 lower half Q2 losses.
Next let's look at the first 2 columns. Team C is easily the best in top-half Q1, followed by A; B and D did bad here. But Team C did really bad in lower-half Q1, as did D, where as B and A did well.
Basically, vs very good teams, C has the best chance of winning, followed by A, then B, and D has a bad chance.
But vs "just ok" teams, D and B are the best, followed by A and C.

A=Duke
B=Illinois
C=UK
D=Auburn

Dook lost to #120 and #124 two teams with losing records. #124 GATech was a Quad 3 loss for them they must have just moved up out of it recently, barely. Convenient.😀

 

[TucsonCat]

You guys think worry Quad 1 wins way too much.

I've never ever heard a committee member say one team had 6 quad 1 wins and the other 5 to explain seedings. They're a tool, a blunt tool, to start the process. Give some guidance when evaluating home and away victories.

But once the field is selected, the committee pretty much compares every team against each other. And it's s not blind resumes like on ESPN. They know if you beat Houston by 15 or UC Irvine by 1 on the road.

 

[The_Answer1313]

But the kicker is NET already does this. It knows that you beat Houston by 15 or UC Irvine by 1 on the road. It’s basically your ranking

Even if the committee did look at it in more detail than we give credit for, this stuff is already figured out for them.

 

[TheDude73]

I looked at how 4 schools (UK, Duke, Auburn, Illinois) have done this year in the Quad system, comparing them graphically/visually. One problem with the Quad system is that breaking 362 teams into just 4 groups. Is a Win at #1 the same as a Win at #75? Of course not, but they both count equally (as Q1 wins). Similarly, are Loses at #135 and at #136 really different? Of course not, but using the Quad system they are.

Unfortunately, I can't get the formatting within this page to show what I've done in Excel, but below is my attempt as explaining it.

First of all, we will drop the Q4 games, they all had 5-7 of them, all at home, and they all won.
If you standardize the Neutral and Road games in a way that you can overlay them with the Home games, and then break each Quad in half, then that improves the precision and seems to make things more comparable.

RECORDS vs:

	Top half of Quad 1	Lower half of Quad 1	Top half of Quad 2	Lower half of Quad 2	Top half of Quad 3	Lower half of Quad 3
Team A	2 - 2	3 - 1	5 - 1	1 - 2	4 - 0	3 - 0
Team B	2 - 4	3 - 0	5 - 1	1 - 1	3 - 0	4 - 0
Team C	5 - 2	0 - 4	1 - 1	1 - 0	5 - 1	3 - 0
Team D	1 - 5	0 - 2	4 - 1	2 - 1	6 - 0	3 - 0

So, the question really is, what do you VALUE MORE, quality WINS, or bad LOSES?

UK has the 1 Q3 loss, but it is a top-half Q3 loss, and we do not have any lower-half Q2 losses (all 3 others have at least 1). So, looking at the last 3-4 columns Team C is a little worse than B & D, but not much. And it is arguable if Team C is worse than A since A has 2 lower half Q2 losses.
Next let's look at the first 2 columns. Team C is easily the best in top-half Q1, followed by A; B and D did bad here. But Team C did really bad in lower-half Q1, as did D, where as B and A did well.
Basically, vs very good teams, C has the best chance of winning, followed by A, then B, and D has a bad chance.
But vs "just ok" teams, D and B are the best, followed by A and C.

A=Duke
B=Illinois
C=UK
D=Auburn

Awesome look at this my man. But sadly the committee won’t look at it even 1/4 as much as you did (although they should).