NET rankings

[redmen9194]

The Big East and the Big 12 have all of their teams in the top 100.

[Savannah Jay]

At least 4 if not more, with good seeds? It looks like Houston is probably on the right side of the bubble and then what? It'd take a great SMU, Cinci, or Temple run to get 4th team in, but ruining good seeds in the process.

No it wouldn't ruin good seeds.... The AAC can get 4 teams finishing with 13+ wins. Last year they had 4 with 13+ wins and also a 11 and 10 win team(that depending the team can make the tourney). You can do that with a 12 team/18 game schedule.

Mid-Major Man,

All that winning! So much winning! Yeah, your conference had a team go 0-18 and another 3-15 (and a couple 6-12 teams, to boot), inflating win totals of other teams.

How did that work out? Only 4 of the 12 teams in the league received bids (that's 33.3%, stats guy) with 3, 7, 9, and 11 seeds. You paint this seeding as a "doomsday" scenario for the Big East (except for the part where we will get more than 4 teams in even though we are "saddled" with 2 fewer teams). Literally, you are on here saying how terrible it would be to get 6-8 teams (that would be 60-80% of our conference, which is a lot better than 33.3%) in the tournament because they'd have bad seeds yet that's your conference every year, except far fewer bids.

You shit all over your own credibility with almost every most, Mid-Major Man! Can I call you 3M?

[GumbyDamnit!]

PC up to 81 in the Net. 10 teams in the top 81. Zero poison games left from here on out. This is crazy.

Looking at it though, it seems you get overly punished for home losses. Marquette and CU both dropped a bunch with the home losses last night. To win this league and maximize Net ranking you have to win at home and try to steal a couple here and there on the road.

[stever20]

PC up to 81 in the Net. 10 teams in the top 81. Zero poison games left from here on out. This is crazy.

Looking at it though, it seems you get overly punished for home losses. Marquette and CU both dropped a bunch with the home losses last night. To win this league and maximize Net ranking you have to win at home and try to steal a couple here and there on the road.

I think the drop for home teams is because home losses count 1.4 instead of 1.... Also, we're still pretty early on..

Creighton last night loses at home... going into the game their record was 7.8-2.2 .780. It's now 7.8-3.6. .684. That's a pretty huge drop. For Marquette from .766 to .667.

You would think as we go along, the home losses get penalized less as the denominators get bigger. But right now, it's like you missed an assignment early in the quarter and got a zero.

[Fieldhouse Flyer]

Looking at it though, it seems you get overly punished for home losses. Marquette and CU both dropped a bunch with the home losses last night. To win this league and maximize Net ranking you have to win at home and try to steal a couple here and there on the road.

I agree. I noticed this too in December, and posted the following:

Biggest Takeaways from the NCAA Tournament Selection Committee's Top 16 Reveal – Kerry Miller, Bleacher Report – February 9, 2019

Those of us in the bracketology industry were curious to see how the committee would use its new ranking tool, and these are the five biggest takeaways from the reveal:

1. Quadrant 1 Wins Are Still Huge
2. Conference Records Don't Matter
3. Player Absences Are Being Considered
4. Road Records Are Important
5. The Committee Loves NET

4. Road Records Are Important

Prior to the top 16 reveal, Muir spoke with Matt Norlander of CBSSports.com. He made it clear that road records—which have always been a key discussion point—are extra important this year.

"We recognize now, even more so than prior, road wins and crediting road wins," he said. "That's something we heard from our coaching brethren, the National Association of Basketball Coaches saying, 'We want this factored in as you continue to compare teams and trying to select and seed.'"

Muir mentioned Kansas' poor road record factored into the struggle to properly seed the Jayhawks. And Kansas is one of just three teams (with Purdue and Iowa State) in the top 16 that doesn't have at least two more road wins than losses.

Duke and Tennessee are the only two teams that are undefeated on the road this season, and they were the top two overall seeds. Even though Nevada has yet to play a Quadrant 1 game this season, the Wolf Pack sneaked in as a No. 4 seed, no doubt thanks to their 11-1 record in road and neutral games.

What is a little confusing about that emphasis on road warriors is that Villanova (5-2 on the road) and LSU (5-1) were left out of the top 16. I thought both of those four-loss teams would make the cut. But with 19 teams reasonably in the running for those 16 spots, a few deserving resumes had to be left out.

5. The Committee Loves NET

There were a couple of noteworthy differences between the committee's seedings and the NET rankings. But unless there's something about a resume that jumps off the page—such as a bunch of Quadrant 1 wins or an unsightly strength of schedule — it seems like the committee is more or less going to default to the NET rankings.

But unless there's something about a resume that jumps off the page—such as a bunch of Quadrant 1 wins or an unsightly strength of schedule—it seems like the committee is more or less going to default to the NET rankings.

Twelve of the 16 teams have a NET ranking within two spots of where the committee seeded them, and three of those teams are in the same spot on both lists. The No. 1 seeds are all in the top four in NET, and all eight Nos. 1 and 2 seeds are in the top nine in NET.

We'll have to wait until Selection Sunday to see if that adherence to NET carries all the way to the bubble, but it does seem like the committee is more hesitant to disregard NET than it was RPI.

One of the big unknowns about the top-secret NET Rankings formula is the extent that the win margin affects the NET Ranking. I expect that it is rather higher than the NCAA would like to be known publicly.

We do know that credit for win margin is capped at a 10-point victory, but we may never find out the relative value of a 10-point win vs. a 5-point win - and it could be significant - which will keep NET Rankings formula top-secret, as the NCAA does not want to be seen as encouraging running up the score.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

It seems you get overly punished for home losses.

The corollary to this is that you get overly rewarded for road wins, which should turn out to be of considerable benefit to any strong basketball team in a mid-major conference who could get 8 or 9 road wins in conference play - something that would be very unlikely in a P6 basketball conference.

[stever20]

what folks fail to realize with the NET is this-
while yes there is the 10 point max. for MOV

there is also the effeciency part of the ratings. And that is not capped at all. So there is a difference between a 10 point win and a 30 point win.

[stever20]

It seems you get overly punished for home losses.

The corollary to this is that you get overly rewarded for road wins, which should turn out to be of considerable benefit to any strong basketball team in a mid-major conference who could get 8 or 9 road wins in conference play - something that would be very unlikely in a P6 basketball conference.

It feels having watched so far this year like the punishment for a home loss is far greater than the reward for a Road win.

Lets take a look at something...

Say your adjusted record going into a game is like Creighton last night. 7.8-2.2. So .780.

Road win- takes them to 9.2-2.2. So .807
home loss- takes them to 7.8-3.6. So .684

That's crazy.

[kayako]

Wow, Xavier and St. John's got absolutely hammered.

[Hall2012]

Wow, Xavier and St. John's got absolutely hammered.

Meanwhile Wisconsin loses at home to Illinois and only drops 3 spots from 26 to 29...wtf?

Other "P5's" to lose at home yesterday:
-Georgia Tech - dropped 4 spots
-Wake Forest - dropped 7 spots
-North Carolina - dropped 3 spots
-Iowa State - dropped 5 spots
-Texas - dropped 9 spots

Yet Xavier and St. John's both dropped 16 spots, and it's not like they lost to bad teams.

I'm sure it's all based on how close the teams around you are, but one can be forgiven for looking at that and asking some questions.

[stever20]

Wow, Xavier and St. John's got absolutely hammered.

Meanwhile Wisconsin loses at home to Illinois and only drops 3 spots from 26 to 29...wtf?

Other "P5's" to lose at home yesterday:
-Georgia Tech - dropped 4 spots
-Wake Forest - dropped 7 spots
-North Carolina - dropped 3 spots
-Iowa State - dropped 5 spots
-Texas - dropped 9 spots

Yet Xavier and St. John's both dropped 16 spots, and it's not like they lost to bad teams.

I'm sure it's all based on how close the teams around you are, but one can be forgiven for looking at that and asking some questions.

St John's got beat by 21 points. That's going to cost you.

The problem with all the ones you bring up-
GT/WF/UNC- were in that 90-105 range already
ISU/Tex- Iowa St was 84 going into yesterday, Texas was 67.

As far as Wisconsin- looking-
was 7-4.2 .625
now 7-5.6 .555

So dropped .070 for adjusted record

and even regular record went from 9-5 (.643) to 9-6 (.600)

Xavier on the other hand
was 8.8-2.2 .800
now 8.8-3.6 .709

so proportionally a bigger drop Regular record went from 12-3 (.800) to 12-4 (.750) so more there

Also you have the point difference 1 vs 12.