NET 2019-2020 Season

ramster · Unread post by **ramster** » 4 years ago

CBS take on the initial NET Rankings

https://www.cbssports.com/college-baske ... ego-state/

Interesting that there are only 4 undefeated teams left out of 353 and all 4 are in the NET Top 20. Looks like NET is overrating based on zero loses.

Top 20
1. Ohio State
2. Kansas
3. San Diego State
4. Butler
5. Baylor
6. Gonzaga
7. Auburn
8. Louisville
9. Stanford
10. Dayton
11. Maryland
12. West Virginia
13. Duke
14. Wichita State
15. Arizona
16. Memphis
17. Oregon
18. Penn State
19. Liberty
20. Duquesne

ramster · Unread post by **ramster** » 4 years ago

Bleacher Report Assessment of first NET Rankings release

https://www.google.com/amp/s/syndicatio ... s.amp.html

Across the board, the NET loves mid-major teams with either a zero or a one in the loss column a whole lot more than KenPom does. The one exception to that rule is Gonzaga, which ranks sixth both on KenPom and in the NET. But isn't Gonzaga always the exception when we're talking about mid-majors?

Aside from the Zags, there are 13 mid-major teams who fit the undefeated or one-loss description. Here's a look at how much better they rank in NET than in KenPom, in ascending order of magnitude:

Dayton (No. 10 in NET, No. 14 in KenPom, +4)
SMU (88 NET, 96 KP, +8)
Memphis (16 NET, 28 KP, +12)
Wichita State (14 NET, 34 KP, +20)
San Diego State (3 NET, 32 KP, +29)
Eastern Michigan (147 NET, 178 KP, +31)
Duquesne (20 NET, 60 KP, +40)
Kent State (59 NET, 102 KP, +43)
UTEP (70 NET, 114 KP, +44)
Richmond (22 NET, 69 KP, +47)
Northern Iowa (28 NET, 76 KP, +48)
Liberty (19 NET, 71 KP, +52)
George Mason (67 NET, 121 KP, +54)

Unread post by **SGreenwell** » 4 years ago

It wouldn't surprise me if NET is too heavily ranking teams with 0 losses right now - isn't it meant more as an end-of-year snapshot or ranking? Meaning, it isn't a metric like Kenpom that's using the results from players last year to attempt ranking the teams going forward. It's just meant to be a metric of how a team is doing *this* year. It wouldn't surprise me that, if for most of these teams, their "true" talent level is somewhere in the middle of the NET and Kenpom rankings.

For me, Duquesne is the hardest of these teams to peg. Their schedule is garbage, but it's hard for even an average team to play consistently enough to rack up a 9-0 record. (Meaning, you get off shooting nights, you drop a game you should have won, etc.) They only lost by 8 in an exhibition vs. West Virginia, but coaches all handle that sort of game differently, so I don't know how "valuable" that result is. I don't think we'll know until A-10 play whether they're legitimately good, or if they simply had a good stretch at the beginning of the year helped by a soft schedule.

Blue Man · Unread post by **Blue Man** » 4 years ago

I'm just confused as to how they start a NET type ranking. It's tough to know who you played if everyone starts at the same place.

Like a win off of NJIT is the same as a win against Kentucky. It must be?

To me that says that NET will literally be meaningless until mid-Feb at the earliest.

rjsuperfly66 · Unread post by **rjsuperfly66** » 4 years ago

SGreenwell wrote: ↑4 years ago It wouldn't surprise me if NET is too heavily ranking teams with 0 losses right now - isn't it meant more as an end-of-year snapshot or ranking? Meaning, it isn't a metric like Kenpom that's using the results from players last year to attempt ranking the teams going forward. It's just meant to be a metric of how a team is doing *this* year. It wouldn't surprise me that, if for most of these teams, their "true" talent level is somewhere in the middle of the NET and Kenpom rankings.

For me, Duquesne is the hardest of these teams to peg. Their schedule is garbage, but it's hard for even an average team to play consistently enough to rack up a 9-0 record. (Meaning, you get off shooting nights, you drop a game you should have won, etc.) They only lost by 8 in an exhibition vs. West Virginia, but coaches all handle that sort of game differently, so I don't know how "valuable" that result is. I don't think we'll know until A-10 play whether they're legitimately good, or if they simply had a good stretch at the beginning of the year helped by a soft schedule.

I think you are on the right track with that...

KenPoms formula is far from public, but he does say it takes about 20 games to remove preseason influence.

If you assume every game to be worth 5%, you can figure where a team would be roughly ranked without preseason influence if you have the right data points.

Duquesnes preseason 109 is hurting their regular season 58.

Rhody83 · Unread post by **Rhody83** » 4 years ago

Richmond’s win over Vandy gets weaker.

Rhody83 · Unread post by **Rhody83** » 4 years ago

NYGFan_Section208 · Unread post by **NYGFan_Section208** » 4 years ago

Rhody83 wrote: ↑4 years ago

Isn't it always?

Rhody83 · Unread post by **Rhody83** » 4 years ago

This improves Dayton’s neutral court win over Saint Marys.
Saint Marys up 26 in neutral court game against Arizona State.

ramster · Unread post by **ramster** » 4 years ago

Rhody83 wrote: ↑4 years ago

Richmond and VCU lost versus the Point Spread while St Bonaventure and LaSalle won.

Richmond -4 points. Richmond won by 3 - lost PS
VCU -5.5 points. VCU won by 5 - lost PS
St Bonaventure -14 points. St Bonaventure won by 17 - won PS
LaSalle -12.5 points. LaSalle won by 14 - won PS

KevanBoyles · Unread post by **KevanBoyles** » 4 years ago

Scratching my head at the NET this morning. We beat Western Kentucky, who was ranked In the high 80s, and we stay the same on Sunday and lose a couple of spots on Monday. UCONN beats a team ranked 237 (after beating Saint Peters who was ranked over 250) and gains seven spots. Both were home games. I understand they won by more than 20 points, but still it doesn’t make any sense.

Rhody83 · Unread post by **Rhody83** » 4 years ago

RF1 · Unread post by **RF1** » 4 years ago

The full official NET rankings are always listed here:

https://www.ncaa.com/rankings/basketbal ... t-rankings

Rhody83 · Unread post by **Rhody83** » 4 years ago

Dayton 14
Richmond 54
VCU 58
SLU 61
Duquesne 66
G Mason 72
URI 80
Davidson 99

RhodyRams12 · Unread post by **RhodyRams12** » 4 years ago

Rhody83 wrote: ↑4 years ago

San Diego St #1

Obadiah · Unread post by **Obadiah** » 4 years ago

KevanBoyles wrote: ↑4 years ago Scratching my head at the NET this morning. We beat Western Kentucky, who was ranked In the high 80s, and we stay the same on Sunday and lose a couple of spots on Monday. UCONN beats a team ranked 237 (after beating Saint Peters who was ranked over 250) and gains seven spots. Both were home games. I understand they won by more than 20 points, but still it doesn’t make any sense.

Two possible reasons. It was home game and URI was favored to win. The game went to OT and the rating was negatively impacted by the margin of victory. Also, our SOS took a hit as our OOC opponents had a bad wee.

PeterRamTime · Unread post by **PeterRamTime** » 4 years ago

Our NET keeps going down

Duquesne dropped 46 places after a neutral site loss to UAB.

Unread post by **ATPTourFan** » 4 years ago

Shows win % is a large factor.

Rhodymob05 · Unread post by **Rhodymob05** » 4 years ago

NET seems like its flawed.

Unread post by **TruePoint** » 4 years ago

Any metric you could design would be flawed right now, 1/3 of the way thru the season. Even when we get to the end, there will be nits to pick with any metric, but at this point it is borderline useless. Just let it play out and if it feels wrong in March we can make a bunch of noise about it.

SmartyBarrett · Unread post by **SmartyBarrett** » 4 years ago

The fact that Duquesne dropped almost 50 spots due to one loss shows that there's not a large enough sample size to put much stock into NET right now. Same with most of the other metrics. Doesn't stop me from stalking them like crazy, though.

theblueram · Unread post by **theblueram** » 4 years ago

Sad thing is, conference play starts in about two weeks so the NET rankings actually do mean something right now.

Unread post by **TruePoint** » 4 years ago

The impact of OOC games will continue to evolve even after we enter conference play as OOC opponents sink or float during their conference schedules. The truth is that until every game has been played, these metrics are just snap shots. They don’t count for anything until they are finalized and the committee is looking at them on the final team sheets. Obviously once you get to a certain point in the season, you can get a more solid idea about where everything will wind up.

The best way I can think of to visualize this is just to simplify and think about winning percentage. Let’s say you play your first game of thirty and win it. Your second game is worth .50 against your winning percentage: if you win, your winning percentage is 1.000 and if you lose it is .500. Then a few months later you play your thirtieth game and whether you win it or lose it will impact your winning percentage by only about .03. THIS DOES NOT MEAN THAT YOUR SECOND GAME IS WORTH .47 MORE THAN YOUR LAST GAME. At the end, all 30 of your games are worth .03, even though your second game moved your winning percentage by more at the time due to the small sample size. This is the same thing with these metrics, only they are way way more dynamic than winning percentage because they are trying to factor in the strength of your opponents plus a bunch of other efficiency stats and whatever. The more data there is, the more precise the measurement is going to be (and therefore the more accurate it is).

If you bought a brand new car and then checked the “average MPG” gauge after you drove it home from the dealership in rush hour traffic and it tells you that you that your car averages 9 MPG, chances are you didn’t buy the world’s biggest gas guzzler and if you check it again after 1,000 miles you’ll get a more accurate number, and more accurate still after 10,000, and even more still after 100,000. At this point in the basketball season, the data just isn’t there to produce an accurate ranking no matter what inputs you use in your model.

reef · Unread post by **reef** » 4 years ago

We really need these last 2 games to end OOC 9-3 then maybe a 12-6 puts us at 21-9 heading to the conf tournament guessing that’s on the bubble ??

rambone 78 · Unread post by **rambone 78** » 4 years ago

Depends on who we lose to.....I'm beginning to think not only do we need to go 9-3 OOC, but 12-6 in conference might not be enough.

Wrong side of the bubble is my estimation now. NIT worthy only.

A big run in the conference tourney might do it, but not really sure how much the committee values that nowadays.

Our margin for error is very small compared to Dayton's. Our NET has to be in the 40's tops to make it.

P5's can get in with NET rating in the 50's or so...but poor old mid majors usually get left out with that.

Rhodymob05 · Unread post by **Rhodymob05** » 4 years ago

We need our wins to get wins

rambone 78 · Unread post by **rambone 78** » 4 years ago

A NET of 80 will not get it done.

I know it will be adjusted as the season goes on, but we're a long way from any conversation about the NCAA tournament.

I hate to say it, but that loss to WV could be the difference. We needed that Q1 win. 0-3 currently no good

Yeah I know it's still early, before I get roasted for the above comments.

Rhody83 · Unread post by **Rhody83** » 4 years ago

rambone 78 wrote: ↑4 years ago A NET of 80 will not get it done.

I know it will be adjusted as the season goes on, but we're a long way from any conversation about the NCAA tournament.

I hate to say it, but that loss to WV could be the difference. We needed that Q1 win. 0-3 currently no good

Yeah I know it's still early, before I get roasted for the above comments.

20 games plus A10 tournament games. The discussion on where Rhody’s NET is today is meaningless.
Finish in the Top 3 of the A10 with some good wins and get to semi finals of A10T. Then they are in.
Finishing 4th in A10 spells either bubble or trouble.

PeterRamTime · Unread post by **PeterRamTime** » 4 years ago

rambone 78 wrote: ↑4 years ago Depends on who we lose to.....I'm beginning to think not only do we need to go 9-3 OOC, but 12-6 in conference might not be enough.

Wrong side of the bubble is my estimation now. NIT worthy only.

A big run in the conference tourney might do it, but not really sure how much the committee values that nowadays.

Our margin for error is very small compared to Dayton's. Our NET has to be in the 40's tops to make it.

P5's can get in with NET rating in the 50's or so...but poor old mid majors usually get left out with that.

Yeah six conference losses is a lot of losses.

Also why it really sucked how Richmond and Duquesne lost those games the other day.

If they had finished the OOC unscathed, they would have brought those pretty NETs into conference play, they still have one more game each.

We want as many A-10 teams in the top 100 as possible.

Unread post by **TruePoint** » 4 years ago

Agreed I was hoping for Duquesne and Richmond to make it into conference play before turning back into pumpkins, but you can’t take away the wins they already got.

rambone 78 · Unread post by **rambone 78** » 4 years ago

Well we all knew those pretty NET's were fool's gold anyway.

But yeah they could at least have waited a couple more games......

PeterRamTime · Unread post by **PeterRamTime** » 4 years ago

On top of all that, the A-10 has already gone above and beyond.

I have a good feeling our top four wins will get better as the year goes on.

Bama, NT, PC, and WKU.

Especially Bama and PC since they play nothing, but quality teams from here on out.

Same goes for other A-10 teams who have beaten power conference foes.

rhodysurf · Unread post by **rhodysurf** » 4 years ago

PeterRamTime wrote: ↑4 years ago On top of all that, the A-10 has already gone above and beyond.

I have a good feeling our top four wins will get better as the year goes on.

Bama, NT, PC, and WKU.

Especially Bama and PC since they play nothing, but quality teams from here on out.

Same goes for other A-10 teams who have beaten power conference foes.

PC jumped 30 NET spots beating texas. They will steal a few games they have no business doing so in and their net will jump to be top 75 I think.

reef · Unread post by **reef** » 4 years ago

We may need to go 13-5 to feel good about ourselves then 12-6 may leave a lot of work to be done at conf tournament

KevanBoyles · Unread post by **KevanBoyles** » 4 years ago

We are now 69 in the NET. Up 15 from 84. Nice bump.

ramster · Unread post by **ramster** » 4 years ago

Anyone know how to see how the conferences compare in strength on NET? How does the A-10 compare in the 32 D1 Conferences?

As we close out OOC Play:

Duquesne gets hammered from 65 to 105 dropping 40 slots with neutral court loss to Marshall
Richmond goes from 51 to 77 with loss at Alabama
VCU moves up from 64 to 52 with home win vs Loyola Maryland

Dayton is ranked #12 followed by VCU #52, Saint Louis #60, URI #69, George Mason #71, Richmond #77, Davidson #94

West Virginia #9
Maryland #16
LSU #41
Alabama #65
North Texas #111
WKU #122

Rutgers #37 - of interest since they have been bad for so long
DePaul #39 - much better than expected in BE
St Johns #49 - much better than expected in BE
UConn #70
PC a disaster #129

rambone 78 · Unread post by **rambone 78** » 4 years ago

Win by 10 or more.

That's the key.

Win at Brown by double digits, NET gets better.

Beating Bethune-Cookman at home gains VCU 17 spots?

Could Alabama get to 35?

Unread post by **TruePoint** » 4 years ago

I think VCU gained 12 spots, not 17 right? But the VCU jump, like URI’s, likely has to do with more than just their result yesterday. As your opponents win or lose, that impacts your ranking also.

rambone 78 · Unread post by **rambone 78** » 4 years ago

I used to be good at math lol.

Unread post by **TruePoint** » 4 years ago

ramster wrote: ↑4 years ago Anyone know how to see how the conferences compare in strength on NET? How does the A-10 compare in the 32 D1 Conferences?

As we close out OOC Play:

Duquesne gets hammered from 65 to 105 dropping 40 slots with neutral court loss to Marshall
Richmond goes from 51 to 77 with loss at Alabama
VCU moves up from 64 to 52 with home win vs Bethune-Cookman

Dayton is ranked #12 followed by VCU #52, Saint Louis #60, URI #69, George Mason #71, Richmond #77, Davidson #94

West Virginia #9
Maryland #16
LSU #41
Alabama #65
WKU #122

Rutgers #37 - of interest since they have been bad for so long
DePaul #39 - much better than expected in BE
St Johns #49 - much better than expected in BE
UConn #70
PC a disaster #129

B27037B5-04D3-47BE-9402-41F2E3609577.png

Don’t forget North Texas, which is #111. Beating them on a neutral court is our second best win right now as much attention as the PC game gets and how much we’ve talked about WKU.

Rhody83 · Unread post by **Rhody83** » 4 years ago

ramster, great job on pulling that altogether.
Just one correction - VCU beat Loyola MD and SLU beat Bethune-Cookman.
I agree with TP that VCU’s gain is probably attributable to two items - their previous opponents’ wins (like LSU beating Liberty) and teams ranked ahead of them losing/dropping (like Richmond).

Rhodymob05 · Unread post by **Rhodymob05** » 4 years ago

Crimson tide are winners of five out of their last six, with a big match-up vs Florida coming up.

Unread post by **TruePoint** » 4 years ago

I have so much respect for Nate Oats that I was always hopeful that our beating them early would pay off later as he got control of that program and his team improved over the course of the year. He has some talent to work with there. Good to see it starting to come together.

rambone 78 · Unread post by **rambone 78** » 4 years ago

Alabama with Oats will be really good in a year or two.

This year, they may end up like us, a borderline NCAA tourney team.

If so, our win against them could be a difference maker?

Blue Man · Unread post by **Blue Man** » 4 years ago

NET must count every game the same until later in the year.

Like at the beginning Kentucky and NJIT are equally ranked, so rankings can only go up based on a win/by how much.

The longer the season goes on, the more accurate it becomes.

If PC and Bama can turn their years around, along with a strong A10 showing, we might have enough of a shot to be an at-large.

But we'll need to take 1 against Dayton, and maybe 1 against VCU too.

Section104 · Unread post by **Section104** » 4 years ago

I still think we're in great shape for an at-large if we avoid losing to bad A10 teams and pickup wins against STL, Richmond, and VCU. I don't think beating Dayton is a requirement.

There's so much parity this year that everyone is losing games and as a whole not many outside the top 25 have tremendous resumes.

RF1 · Unread post by **RF1** » 4 years ago

Ponder this: Duquesne dropped more than 80 NET spots in one week after losing to UAB and Marshall on neutral courts.

Blue Man · Unread post by **Blue Man** » 4 years ago

Section104 wrote: ↑4 years ago I still think we're in great shape for an at-large if we avoid losing to bad A10 teams and pickup wins against STL, Richmond, and VCU. I don't think beating Dayton is a requirement.

There's so much parity this year that everyone is losing games and as a whole not many outside the top 25 have tremendous resumes.

Fair. We're still going to need that "marquee win" though. The unfortunate thing for our schedule (or fortunate depending on your POV), is that outside of VCU and Dayton - we don't have many top 100 road games that we could afford to lose in that scenario.

As of right now George Mason and Davidson are the only 2 that I could look at and say that we can survive.

Granted, if by some miracle we go 8-1 at home, with a loss to UD or VCU, and 5-4 on the road (losses to UD, VCU, GM, Davidson) - I would think 13-5, 9-3 still requires a bit more work in the A10 tourney.

Thanks to PC and Bama sucking, along with the WKU injury - our margin for error is a lot smaller.

Unread post by **ATPTourFan** » 4 years ago

RF1 wrote: ↑4 years ago Ponder this: Duquesne dropped more than 80 NET spots in one week after losing to UAB and Marshall on neutral courts.

Shows us that the net ranking gives a lot of weight to winning percentage and efficiency. They were way too high due to 100% win percentage. They also played awful in there two losses which hurt their efficiency.

Rhodymob05 · Unread post by **Rhodymob05** » 4 years ago

URI currently has a top 50 RPI.

Login • Register