NET Rankings

The home for Big East hoops

Re: NET Rankings

Postby Hall2012 » Mon Mar 11, 2019 4:45 pm

XUFan09 wrote:
Hall2012 wrote:LOL Seton Hall's NET ranking went down after beating Villanova. $50 says changing the name on their jerseys to "Ohio State" or "Clemson" would make it skyrocket.


Or home court actually matters. On Kenpom, the game was a virtual coin flip, and Seton Hall only won by 4. I imagine the NET formula followed some similar pattern. I don't like that we don't know how NET is calculated, but we do know it is in part an efficiency metric. If a game result doesn't deviate much from what an efficiency metric expects, then the metric won't adjust much.

Anyway, NET is primarily used to evaluate the quality of one's opponents. The fact that Villanova is #25 is far more important to Seton Hall than the fact that they themselves are #62.


And that would be fine if it were just a tool for calculating SOS, but if it's used as presented - being a major factor in NCAA Tournament selection seeding, similar to RPI - then efficiency metrics are being far too heavily considered. Results are what should matter, not efficiency.
Seton Hall Pirates
Big East Tournament Champions: 1991, 1993, 2016
Big East Regular Season Champions: 1992, 1993, 2020
Hall2012
 
Posts: 3462
Joined: Mon Sep 30, 2013 3:04 pm

Re: NET Rankings

Sponsor

Sponsor
 

Re: NET Rankings

Postby XUFan09 » Mon Mar 11, 2019 5:00 pm

Hall2012 wrote:
XUFan09 wrote:
Hall2012 wrote:LOL Seton Hall's NET ranking went down after beating Villanova. $50 says changing the name on their jerseys to "Ohio State" or "Clemson" would make it skyrocket.


Or home court actually matters. On Kenpom, the game was a virtual coin flip, and Seton Hall only won by 4. I imagine the NET formula followed some similar pattern. I don't like that we don't know how NET is calculated, but we do know it is in part an efficiency metric. If a game result doesn't deviate much from what an efficiency metric expects, then the metric won't adjust much.

Anyway, NET is primarily used to evaluate the quality of one's opponents. The fact that Villanova is #25 is far more important to Seton Hall than the fact that they themselves are #62.


And that would be fine if it were just a tool for calculating SOS, but if it's used as presented - being a major factor in NCAA Tournament selection seeding, similar to RPI - then efficiency metrics are being far too heavily considered. Results are what should matter, not efficiency.


Read my second paragraph again. It is about results. Seton Hall beat the #25 team in the nation on their home court. That's what matters. Villanova lost to the #62 team in the nation on the road. That's what matters. Efficiency is used to determine how tough a team's opponents are to play, not how well a team has done personally.

The NET is a major factor in NCAA seeding because it's used to judge who teams have played. The rankings are going to correlate strongly with seeding because of that, but that shouldn't be a surprise because on an aggregate level, a NET ranking somewhat approximates who a team beat and who they lost to. But correlation doesn't equal causation. Note that RPI was used in the exact same way, but people never seemed to get that because of they treated correlation as causation.
Gangsters in the locker room
XUFan09
 
Posts: 1463
Joined: Sun Jul 07, 2013 5:07 pm

Re: NET Rankings

Postby XUFan09 » Mon Mar 11, 2019 5:06 pm

The caveat to all of this is that in addition to the standard things on the team sheets, individual Committee members can use whatever metrics they feel are relevant. So a member could use number of road wins or he could even use RPI (since it's still tracked) or he could use last 10 games (even though it's no longer an official metric). So a member could use direct NET ratings as a personal preference, but that's just one of many possible personal preferences. The official metrics and the week-to-week discussions of team's performances are what drive the process and seeding.
Gangsters in the locker room
XUFan09
 
Posts: 1463
Joined: Sun Jul 07, 2013 5:07 pm

Re: NET Rankings

Postby Hall2012 » Mon Mar 11, 2019 5:41 pm

XUFan09 wrote:
Read my second paragraph again. It is about results. Seton Hall beat the #25 team in the nation on their home court. That's what matters. Villanova lost to the #62 team in the nation on the road. That's what matters. Efficiency is used to determine how tough a team's opponents are to play, not how well a team has done personally.

The NET is a major factor in NCAA seeding because it's used to judge who teams have played. The rankings are going to correlate strongly with seeding because of that, but that shouldn't be a surprise because on an aggregate level, a NET ranking somewhat approximates who a team beat and who they lost to. But correlation doesn't equal causation. Note that RPI was used in the exact same way, but people never seemed to get that because of they treated correlation as causation.


True, correlation does not equal causation, but it does leave the possibility open. And without any data in front of me (this would be a difficult one to test with so many common factors between RPI rating and S curve seeding) I would still bet that there was a cause-effect relationship. However I would also bet that the relationship has diminished over time. We'll see as it relates to the NET rating.

I can't say I know in what ways the committee considered RPI (opponent quality or a direct ranking - probably an element of both) so I'll have to take your word there. My issue is mostly just getting used to change - in the RPI system beating a team ranked significantly above you usually resulted in a boost. NET seems to require a blowout to generate any sort of positive movement.
Seton Hall Pirates
Big East Tournament Champions: 1991, 1993, 2016
Big East Regular Season Champions: 1992, 1993, 2020
Hall2012
 
Posts: 3462
Joined: Mon Sep 30, 2013 3:04 pm

Re: NET Rankings

Postby adoraz » Wed Mar 13, 2019 11:39 pm

Looks like the magic numbers I've been using may have been dead on.

Updated how many BE+BET wins I think each team needs to make the Tournament, as well as my odds of them reaching that number:

Villanova (#25, 11-2)- LOCK
Marquette (#29, 10-2)- LOCK
Seton Hall (#62, 9-9)- needs 0 more to likely get in (100% odds of winning 0 more) TRENDING UP
St. John's (#66, 8-10, 1-0)- needs 0 more to likely get in (100% odds) TRENDING UP
Creighton (#54, 9-9)- needs 2 more to likely get in (25% odds) TRENDING UP
Xavier (#70, 9-9)- needs 2 more to likely get in (25% odds) TRENDING UP
Georgetown (#77, 9-9)- needs 2 more to likely get in (25% odds) TRENDING UP

Crazy how the entire bubble is trending UP. One week ago, we legitimately could've had 2 teams dancing.

With their big win tonight, Johnnies most likely are IN. As I've said before, adding 1 win to the "magic number" should guarantee a bid (while hitting the magic number "likely" gets teams in), so if Johnnies or Seton Hall win tomorrow then they shouldn't need to be concerned with the bubble any longer.

As for everyone else... to be honest I was lazy in evaluating because we don't know yet who will play in the semis. Creighton might be able to get in with 1, but I think they need 2 to feel decent about their chances. Georgetown and Xavier could get in with 2 as well, though I wouldn't feel as confident. Again, depends on who they play.

3 wins of course would clinch the auto bid.
Johnnies
adoraz
 
Posts: 1955
Joined: Sat Nov 15, 2014 5:13 pm

Re: NET Rankings

Postby stever20 » Thu Mar 14, 2019 12:43 am

You really want to say Creighton gets in with just the win over Xavier.... I think it would hinge on how the bubble does go- are there any more bid thieves, etc.

I'd for sure add Providence to your list. If they beat Villanova and then Creighton/Xavier- they'd be 20-15.

I think we can say for sure though-
if Providence, Creighton, Xavier, or Georgetown loses tomorrow- they're out of contention for an NCAA bid. All 4 need at least 1 win here.
stever20
 
Posts: 13488
Joined: Wed Sep 11, 2013 1:43 pm

Re: NET Rankings

Postby stever20 » Thu Mar 14, 2019 3:27 pm

stever20 wrote:You really want to say Creighton gets in with just the win over Xavier.... I think it would hinge on how the bubble does go- are there any more bid thieves, etc.

I'd for sure add Providence to your list. If they beat Villanova and then Creighton/Xavier- they'd be 20-15.

I think we can say for sure though-
if Providence, Creighton, Xavier, or Georgetown loses tomorrow- they're out of contention for an NCAA bid. All 4 need at least 1 win here.

well, I gave Providence the kiss of death, didn't i?
stever20
 
Posts: 13488
Joined: Wed Sep 11, 2013 1:43 pm

Re: NET Rankings

Postby Fieldhouse Flyer » Fri Mar 15, 2019 9:17 am

On Sunday March 10 Fieldhouse Flyer wrote:
NCAA MEN'S BASKETBALL NET RANKINGS - ncaa.com - Games through Saturday, March 9th

26 - Villanova
29 - Marquette

54 – Creighton
61 - Seton Hall
62 – Butler
66 - St. John's
71 – Xavier
74 – Providence
76 – Georgetown

101 - DePaul


NCAA MEN'S BASKETBALL NET RANKINGS - ncaa.com - Games through Thursday, March 14th
25 - Villanova
28 - Marquette

52 – Creighton
59 - Seton Hall
64 – Butler
69 – Providence
70 – Xavier
72 - St. John's

82 – Georgetown
102 - DePaul


BUBBLE WATCH: Teams on NCAA bubble fear the bid thief - The Tacoma News Tribune - Friday March 15, 2019
ON THE RISE

Xavier:
The Musketeers (18-14) still look like they're a long shot to go dancing, but the win over Creighton gives some hope. If Xavier can beat Villanova on Friday, Selection Sunday will at least be worth watching.

FADING HOPES

Creighton:
The Bluejays (18-14) have had a lot of tough losses this season and suffered another one on Thursday in a 63-61 loss to Xavier. Selection Sunday is going to be a nail-biter.

Georgetown: The Hoyas (19-13) got thumped by Seton Hall 73-57 . They still have a slim shot to get in the field, but that wasn't a good last impression.

FRIDAY'S IMPORTANT GAMES

Florida vs. No. 9 LSU, Ohio State vs. No. 6 Michigan State, Xavier vs. No. 25 Villanova, Alabama vs. No. 4 Kentucky, and Temple vs. Wichita State.
User avatar
Fieldhouse Flyer
 
Posts: 1389
Joined: Tue Apr 01, 2014 5:11 am

Re: NET Rankings

Postby Fieldhouse Flyer » Sat Mar 16, 2019 9:41 am

NCAA MEN'S BASKETBALL NET RANKINGS - ncaa.com - Games through Friday, March 15th
26 - Villanova
28 - Marquette

52 – Creighton
58 - Seton Hall
64 – Butler
68 – Xavier
70 – Providence
72 - St. John's

80 – Georgetown
102 - DePaul

2018-19 BIG EAST STANDINGS - Games through Friday, March 15th

#25 Villanova (13-5) 24-9
#23 Marquette (12-6) 24-9

Seton Hall (9-9) 20-12
Georgetown (9-9) 19-13
Creighton (9-9) 18-14
Xavier (9-9) 18-15

St. John's (8-10) 21-12
Providence (7-11) 18-15
Butler (7-11) 16-16
DePaul (7-11) 15-15
User avatar
Fieldhouse Flyer
 
Posts: 1389
Joined: Tue Apr 01, 2014 5:11 am

Re: NET Rankings

Postby Fieldhouse Flyer » Sun Mar 17, 2019 9:18 am

NCAA MEN'S BASKETBALL NET RANKINGS - ncaa.com - Games through Saturday, March 16th
26 - Villanova
28 - Marquette

53 – Creighton
57 - Seton Hall
64 – Butler
67 – Xavier
70 – Providence
73 - St. John's

82 – Georgetown
102 - DePaul


NCAA tournament bubble breakdown: Projecting who's in and who's out- Jeff Eisenberg, Yahoo Sports - March 16, 2019
LAST FOUR IN

ST. JOHN’S
(21-12, 8-10, NET 72, KenPom 79)

Q1: 5-7

Q2: 5-3

Best wins: Marquette (2), Villanova, VCU, at Creighton

Q3, Q4 losses: 2 (Georgetown, DePaul)

When St. John’s led Villanova deep into the second half on Jan. 8, the Johnnies appeared to be on track to make the NCAA tournament with ease. They were minutes away from improving to 15-1 overall and 3-1 in the Big East. Those days feel like a long time ago now given how the second half of the season has gone. A talented but erratic St. John’s team dropped 11 of its final 18 games to fall perilously close to the bubble. The argument in favor of St. John’s is a 10-10 record against the top two quadrants, including a total of three wins over Big East frontrunners Marquette and Villanova. The argument against the Johnnies is their soft non-conference strength of schedule, bloated computer numbers and sub.-500 record in a weaker-than-usual Big East.


The five toughest decisions facing the NCAA tournament selection committee - Jeff Eisenberg, Yahoo Sports - March 17, 2019


NCAA tournament Bubble Watch: Which teams could be left out - John Gasaway, ESPN Insider - March 17, 2019
BIG EAST

Locks:
Marquette, Villanova, Seton Hall

Work to do: St. John's, Creighton, Georgetown, Xavier

St. John's Red Storm
St. John's was projected as a No. 11 seed going into the Marquette game, and Bubble Watch supposes, if a few more bubble teams had won a few more games this week, there could be a discussion on whether this team is now in real trouble. Chris Mullin's men were blown out 86-54 by Marquette, and the NET ranking attached to the Red Storm even going into a lopsided loss already was in the mid-60s. Now factor in that, in the eyes of the rating system, the Johnnies actually were playing a "home" game at Madison Square Garden when they were hammered by 32 points. Meaning this NET ranking could get ugly. Still, there also are two things going in this team's favor. First, even after the debacle against the Golden Eagles in the Big East tournament, St. John's is still 3-2 against Marquette and Villanova for the season. Second, well, the committee does have to get to 68 somehow. (Updated: March 14)

Creighton Bluejays
A two-point loss to Xavier in the Big East tournament probably leaves Creighton just short of the tournament field. The Bluejays covered a good deal of ground to make it that far, and a win against the Musketeers would have changed this picture considerably. Instead, CU has stopped playing at 18-14, leaving it difficult to point to any one facet of the profile that recommends Greg McDermott's team at the expense of other at-large candidates. Creighton's NET ranking is in the 50s, which is fine but not great, a description that also might apply to the 3-10 Quad 1 record and the 13-3 mark in Quads 2, 3 and 4. It was a spirited turnaround from 13-13, but it doesn't appear to be quite enough. (Updated: March 14)

Georgetown Hoyas
Georgetown gave this thing a run for its money when pretty much no one in mid-February saw the run coming, but it appears the Hoyas have come up short. Losing by 16 in your conference tournament quarterfinals (to Seton Hall) when you're being shown as "next four out" material does not customarily presage receiving an at-large bid. The all-freshman backcourt of James Akinjo and Mac McClung gives the Hoyas all kinds of hope for the future, and who knew before the season that this would be the fastest-paced team in Big East play? Alas, it just wasn't enough, apparently, to get into the field this time. (Updated: March 14)

Xavier Musketeers
Xavier needed that game against Villanova, and the Musketeers led for roughly 28 minutes of regulation before the Wildcats took the proceedings to overtime. You know the rest: Travis Steele's men came up short, and the same can almost certainly be said for XU's shot at at-large bid. At 18-15, the Musketeers are a respectable 4-9 against Quad 1 but just 8-6 against Quads 2 and 3. That, plus a lackluster NET ranking meant this team had to have that win against Nova. It almost happened. (Updated: March 15)
User avatar
Fieldhouse Flyer
 
Posts: 1389
Joined: Tue Apr 01, 2014 5:11 am

PreviousNext

Return to Big East basketball message board

Who is online

Users browsing this forum: Bing [Bot] and 47 guests