NET Rankings

[adoraz]

NET rankings, the NCAA's main consideration for the Tournament this year, will be released later today.

https://twitter.com/marchmadness/status ... 0753573889

Appears that 50% of the rating will just be whether you win/lose (just like RPI), while the other half is based on efficiency/margin of victory/loss. Seems like a fair system.

SJU will almost definitely be rated worse in NET than RPI (which we're currently #6) due to winning a few games by small margins. Obviously after 5 games though, margins in any given game are meaningless and simply winning is most important. Same for virtually any metric at this point, there's not enough data (games) for the ratings to matter.

Still, will be fun to see how the league does.

[Fieldhouse Flyer]

Appears that 50% of the rating will just be whether you win/lose (just like RPI) . . .

The RPI formula:

The RPI comprises a team's winning percentage (25%), its opponents' winning percentage (50%), and the winning percentage of those opponents' opponents (25%).

The opponents' winning percentage and the winning percentage of those opponents' opponents both comprise the strength of schedule (SOS). Thus, the SOS accounts for 75% of the RPI calculation and is 2/3 its opponents' winning percentage and 1/3 its opponents' opponents' winning percentages.

Related thread: Bye Bye RPI – 75 posts dated August 22 to September 4, 2018

[stever20]

NET rankings, the NCAA's main consideration for the Tournament this year, will be released later today.

https://twitter.com/marchmadness/status ... 0753573889

Appears that 50% of the rating will just be whether you win/lose (just like RPI), while the other half is based on efficiency/margin of victory/loss. Seems like a fair system.

SJU will almost definitely be rated worse in NET than RPI (which we're currently #6) due to winning a few games by small margins. Obviously after 5 games though, margins in any given game are meaningless and simply winning is most important. Same for virtually any metric at this point, there's not enough data (games) for the ratings to matter.

Still, will be fun to see how the league does.

The 50% part though is also measured by your opponents and your SOS...

Then looks like 25% is efficiency which isn't capped....

Then 25% is actual record, adjusted record, and MOV.

[stever20]

and they are out....
St John's 29
Creighton 36
Villanova 38
Butler 45
Georgetown 76
DePaul 78
Seton Hall 85
Marquette 99
Xavier 101
Providence 115

[adoraz]

and they are out....
St John's 29
Creighton 36
Villanova 38
Butler 45
Georgetown 76
DePaul 78
Seton Hall 85
Marquette 99
Xavier 101
Providence 115

Thanks, so comparing with RPI (which you posted earlier):

RPI
St John's 6
Creighton 24
Villanova 39
Butler 41
Seton Hall 45
Georgetown 49
Marquette 78
Providence 79
DePaul 80
Xavier 129

Lower is better, just adding up all the numbers:
NET total: 702 (average=70)
RPI total: 570 (average=57)

So, basically this means we've won a lot of close games. The pessimist would say we've gotten lucky. The optimist would say we've been clutch in closing out games. For SJU at least, clearly Ponds has been clutch in the closing seconds of games. I'd like to think other teams also already know how to close out games (due to experience from last year) and thus are winning a lot of these nail bitters.

Either way, we're doing solid so far in the rankings. There are a ton of mid majors still in the top 50, and they'll filter out as the year progresses. Radford for example won't be #22 a few weeks from now.

[kayako]

Predictably hilarious, but that was expected this early in the season.

1. Ohio State
10. Loyola Marymount
12. Belmont

https://www.cbssports.com/college-basketball/news/ncaas-new-net-rankings-debut-ohio-state-shockingly-no-1-with-duke-no-6-and-kentucky-no-61/

[billyjack]

Using a formula that isn't transparent is asinine. I felt this way from the moment they first proposed NET, not just because it looks like the Friars initially get hammered by it. RPI had flaws, but the math was in front of us.

Basing team rankings on margin of victory (or efficiency) to me is a bad idea. In every other sport, the object is to win games. Except European soccer uses point differential as a tiebreaker. And NASCAR has a point system. And the NFL has its like 7th tiebreaker based on points... i remember it coming into play once, in 1979, when the Bears ran up like a 62-6 win over the St Louis Football Cardinals, which put Chicago in over Washington.

[stever20]

Predictably hilarious, but that was expected this early in the season.

1. Ohio State
10. Loyola Marymount
12. Belmont

https://www.cbssports.com/college-basketball/news/ncaas-new-net-rankings-debut-ohio-state-shockingly-no-1-with-duke-no-6-and-kentucky-no-61/

Ohio St has 2 games against top 50ish teams on the road vs Cincy and Creighton. won those by 8 and 9 points. Other 4 games by 10+ points- including a Samford team that is 6-1 vs D1 schools. They're #12 in the RPI right now....

Also why shouldn't Kentucky be 61 right now? They have the loss to Duke and 5 wins against bad teams. To argue they should be higher right now is frankly a joke.

[XUFan09]

Using a formula that isn't transparent is asinine. I felt this way from the moment they first proposed NET, not just because it looks like the Friars initially get hammered by it. RPI had flaws, but the math was in front of us.

Basing team rankings on margin of victory (or efficiency) to me is a bad idea. In every other sport, the object is to win games. Except European soccer uses point differential as a tiebreaker. And NASCAR has a point system. And the NFL has its like 7th tiebreaker based on points... i remember it coming into play once, in 1979, when the Bears ran up like a 62-6 win over the St Louis Football Cardinals, which put Chicago in over Washington.

I agree that the lack of transparency is foolish. I am also skeptical of the NET rating quality, though the sample size is too small to tell right now.

However, concerning the other part, I disagree. First, the point is still to win games. You increase the probability of winning a game by extending the lead and not giving your opponent a chance. If you can't extend the lead and instead opponents can make the game close, it reflects on you as a team. Or, if you fall behind early and can't effectively go on a run to catch up, it reflects on you as a team too.

In the end, your MOV-based efficiency is a measure of how tough of an out you are. It demonstrates how well you can extend leads you have and how well you can come back from deficits. Essentially, it shows how hard it is to beat you. That's important because NET (which is only partially an efficiency ranking) is used to evaluate opponents, not the team itself. Providence's NET rating doesn't matter to Providence; it matters to Providence's opponents. It helps in determining how good a win over the Friars was or how understandable a loss to them was.

Also, the example of other leagues is a non sequitur, as they are not subject to at-large selections for playoff spots. Record is still used to seed teams for conference tournaments and the auto-bid goes to the winner in those tournaments, not the highest rated team. However, it is figuring out the other half of the field that requires more refined measures.

[Edrick]

Because the rankings are such a joke, people have been trying to actually reverse engineer that mess.

Here’s this guys attempt...

https://twitter.com/recspecs730/status/ ... 24736?s=21

—-
Nate Silver was the most blunt about it on Twitter - ‘These are the worst rankings I've ever seen in any sport, ever. NCAA needs to go completely back to the drawing board.’