Vegas aside, who gives a shit about Texas Tech’s resume. If someone loses to Texas Tech, even though Texas Tech is the ~70th best team, that should count as Q3 because Texas Tech has a bad resume?
That’s not the point of the system.
OFFICIAL NET Thread - 2022/23
- Thread starter fatsam98
- Start date
That’s not the point of the system.
Imagine defending a team that cant beat anyone with a pulse
But whoopee they beat a 369 team by 50
This seems like a team Minnesota could beat.
Who the **** cares about kenpom...im glad he plays games..in glad there is a live game
How did that kenpom work with Rutgers last year
Well on the one hand we've got a computer that calculates using the score of every single game, and on the other hand we have arguments put together in the brains of people that have thrown away 99% of the available information. If you think the #182 is better than the #70, this late in the season (barring some bizarre scenario like a whole team was injured or something) you're just wrong. You can write 10,000 words about how Texas Tech hasn't beaten anyone and Minnesota managed to beat Ohio St. It doesn't matter. It's just wrong.
It worked fine.
Because in my opinion, actual (non-style driven) results should define who the “70th best team” is for Selection and seeding purposes - not “potential”.
For whatever reason, in addition to losing as expected to elite teams, TT is 0-3 in attempts against at large / bubble teams currently projected in the bottom half of the field or worse. They lost 2 of 3 of those games by double digits and none were road games (one home). This current data is far more meaningful in my opinion than how well they dominated awful Q4 teams compared to their peers who also won their Q4 games.
Actual results say it hasn’t been that hard, to date for bubbler types to beat TT thus far so why should wins against them factor in at a premium as Q1 caliber?
Dude, you better take this Rutgers nonsense to the Texas Tech board!
Your entitled to your opinion, but how can you say it’s completely off based to challenge this? Your choosing” to value Kenpom (not officially used by the NCAA for anything) leaps and bounds ahead of the pure results based system that truly was the Bible of Selection Sunday only a mere 5 years ago. But that system (whether you like it or not - it’s factually accurate to say that it was widely accepted by the NCAA as the tool to define good and bad wins for many many years), ranks TT at 177 which would be the equivalent of a Q4 loss at home and Q3 road for sorting. It’s not like there isn’t a legit basis to disagree with your argument is all I’m saying.
It’s not about TT. Hardly any mid majors in the at large mix. Quad system wins for major conference teams will play a huge role in selection Sunday this year. This is a huge issue to watch.
Don't like it, don't read it.
(1) You keep saying this (the bolded) but the scores of the games are part of the results. Kenpom and every other computer metric are also based on "actual results", they just aren't throwing away the information that you would prefer they throw away
(2) We aren't talking about selection and seeding, we are talking about evaluating Texas Tech for the purposes of evaluating the resumes of other teams who have played games against Texas Tech. I've never seen anyone seriously propose directly using the NET to pick the field. It was obvious based on last years seeding and selection that they did not do that (again NET had even less correlation to seeding than other computer ratings, i.e. Bart/Kenpom/etc. that the NCAA is supposedly not using).
Because it's not being challenged in any way that actually makes sense. It's just people looking at the list and saying "I don't like this". If you want to challenge a model like Kenpom you need to find some systematic way that it is flawed. You need to find more information that it should be including (you are arguing for considering less information which is basically an impossible way to make a model better). And you need to demonstrate that your changes actually improve the predictive accuracy of the model. No one is doing any of that. What you guys doing is akin to reading some scientific paper and saying "well this seems wrong to me". That's nice, but it's also completely meaningless.
Well, again, no one is suggesting we use Kenpom or NET to directly seed or select the field. That's about resumes. Resumes are (1) a subjective thing and (2) you can make arguments for throwing out information when considering resumes.
Also, the RPI, even for what it's designed to do, is terrible. It isn't actually even a fitted model, it's just a formula that someone made up in 1981 to be simple to calculate and understand. Even on the scale of models that throw out all information besides who one the game, it's not good. The fact that the NCAA accepted it is irrelevant.
Somehow people think discussion of the NET itself is bad in the NET thread and would prefer it have one post per day listing Rutgers' ranking that day. A number you can easily look up yourself in other places.
If only we could have more substantive, useful posts like
"Who cares …..
Let’s get this back on track."
and fewer posts about the philosophy or theory of the NET rankings here in the NET thread.
If only we could have more substantive, useful posts like
"Who cares …..
Let’s get this back on track."
and fewer posts about the philosophy or theory of the NET rankings here in the NET thread.
I'm curious if the removal of raw win margin has reduced the overall weight of actually winning games in the NET this year. Will be very interesting to see where TTech ultimately falls out if they go winless the rest of the way - is it possible for an 10-21 team to be in the NET top 100? Or 11-20 if they beat a team like Oklahoma or Oklahoma St?
There's some bootstrapping happening in the B12... TTech is rated highly in because they play a lot of highly rated teams, but those teams are rated highly in part because they play TTech.
Last year we saw some similar strangeness, though, so maybe raw in margin isn't the culprit.
45 - Oklahoma St (15-15)
69 - Kansas St (14-17)
73 - WVU (16-17)
Those were games that counted as Q1 road wins, and none of them won more games than they lost.
There's some bootstrapping happening in the B12... TTech is rated highly in because they play a lot of highly rated teams, but those teams are rated highly in part because they play TTech.
Last year we saw some similar strangeness, though, so maybe raw in margin isn't the culprit.
45 - Oklahoma St (15-15)
69 - Kansas St (14-17)
73 - WVU (16-17)
Those were games that counted as Q1 road wins, and none of them won more games than they lost.
If Texas Tech keeps losing they're not going to count as a Q1 win for anyone when all is said and done. They're right at the cusp of becoming a Q2 road win right now (74).
If they win some and stay in top 75 then maybe you could argue they are worthy of being a low-end Q1 road win.
As you always say, seems slightly too early to be up in arms about this stuff in January
So that might be why last year saw OklaSt, KState, and WVU so high up without the wins to back it up?
I know we have said (maybe earlier in this thread?) that the raw win margin was almost like double dipping - you got credit for outplaying a team and also for beating them by X points - but without it, is the model inflating teams that simply struggle to win games?
NET isn't meant to be predictive like kenpom, and the selection committee is supposed to reward accomplishment not potential - so why is NET pushing teams up the list that don't really have many accomplishments? It feels like beating up on bad teams overcomes actually compiling wins against good teams.
Teams with 9+ losses and their NET ranking, and their best 5 NET wins:
27 - Ohio State (11-9)..... 20 Rutgers, 41 Iowa, 44 @Northwestern, 74 TTech, 81 Cincy
51 - Oklahoma State (11-9).... 10 Iowa St, 25 WVU, 29 Sam Houston St, 65 Oklahoma, 142 (N) Wichita St
61 - Oregon (12-9).... 11 Arizona, 41 Utah, 45 @Utah, 63 Colorado, 71 Washington St
63 - Colorado (12-10).... 2 (N) Tennessee, 42 (N) Texas A&M, 61 Oregon, 71 Wash St, 91 Yale
65 - Oklahoma (11-9).... 25 WVU, 50 (N) Florida, 66 (N) Seton Hall, 74 TTech, 98 Nebraska
66 - Seton Hall (11-9).... 7 UConn, 20 @Rutgers, 39 (N) Memphis, 85 SJU, 101 Butler
72 - Michigan (11-9).... 38 Maryland, 44 Northwestern, 60 Penn St, 62 (N) Pittsburgh, 141 Ohio
75 - TTech (10-10).... 140 E Wash, 159 LTech, 182 NW State, 233 Nicholls, 245 Georgetown
Some thoughts:
- Ohio State is suspect at 27 with only 3 wins against teams better than 70th (and their fourth is an inflated TTech).
- Okla St (51) and Michigan (72) each have just 4 wins against teams better than 140th?
- Seton Hall only has 4 wins better than 100th.
- TTech's best win is vs #140, yet they are 75th?
Looking at 2021, there were also some weird anomalies:
NET Top 75 teams that were 2 or fewer games over .500:
38 - Maryland (16-14)
44 - Penn State (11-14)
49 - Duke (13-11)
56 - Stanford (14-13)
57 - Seton Hall (14-13)
61 - Kentucky (9-16)
64 - Utah (12-13)
66 - Auburn (13-14)
67 - Indiana (12-15)
72 - Georgetown (13-13)
Kentucky was 61 and only won 36% of their games?
NET Top 75 teams that were 2 or fewer games over .500:
38 - Maryland (16-14)
44 - Penn State (11-14)
49 - Duke (13-11)
56 - Stanford (14-13)
57 - Seton Hall (14-13)
61 - Kentucky (9-16)
64 - Utah (12-13)
66 - Auburn (13-14)
67 - Indiana (12-15)
72 - Georgetown (13-13)
Kentucky was 61 and only won 36% of their games?
I guess it’s really a matter of which flaws your more bothered by. For the intended purpose of sorting (quad analysis), the RPI flaw of over valuing gaudy mid major records, impacted far fewer teams that end up in At Large contention. For this reason I thought it worked better than NET has thus far overall.
Kennesaw St at 16-6 in the top 75 RPI would’ve benefitted maybe one At Large team (Indiana) as an example. It didn’t prop up a whole conference of contenders with 2 more potential Q1 and Q2 wins for each other team apiece.
There was some scheduling luck, but at the same time, the system rewarded major conference teams for trying to schedule better mid-majors. I didn’t think that was such a bad thing.
Michigan to this point - is RPI 90 which I think is well placed for where they really ought to be. They have a bad loss, but they also have a collection of wins over bubble caliber teams to offset it some.
Last edited:
Yeaaa any team with a losing record shouldn't be a Q1 win
Great to go deep. But multiples pages comparing Min to TT is a waste of time to me.
Jamille Reynolds is back, too. Went for 16pts/5rb in his second game back, after only playing 12 minutes vs. USF.
I want to argue about this.. but that would really be a waste of time. Though then someone could complain about our meta-argument, thus taking it one level deeper. In conclusion, Minnesota sucks.
I see TTech has jumped up to 69.... but, why on earth is Washington St a place ahead of them at 68 with a 10-13 record and only 1 Q1 win?
Edit: Also not sure how TTech jumped 6 spots by beating an LSU team that was ranked 136th by single digits?
Edit: Also not sure how TTech jumped 6 spots by beating an LSU team that was ranked 136th by single digits?
That’s really a mess. And I wonder how much it’s supported by the offeff that is 8 in kenpom
When you look at their last 10 games , you wonder how they can be top 40 nationally in offeff for the year…and not just 8
The reality , is if they get their adt together quick at home, they can get to 10-10 and sneak in …they have 3 of 4 at home thsr they must get three of four to get to 6-8….
Being 5-9 is too much of a mountain to climb to get to 10-10
And it feels like with so much parity …there will be few to no teams below .500 getting a bid (as it should be )
How in he world is OSU 29?!?!?! That is ridiculous( I guess it only helps us)
Lets beat Iowa today and both RU and IU would be ranked
I think part of the reason OSU is still high is the win over us, and playing us tough at our place. They're also probably getting a bump from playing Purdue close. But, man, I get that being competitive against a very good program should be rewarded in some way... but actually WINNING games needs to have more weight in a tool designed to measure accomplishment.
18-4 Clemson at 59 while 11-10 Ohio State at 29. When records are not valued at all
who cares who a computer says is favorite.
Rutgers was 80 something in the net last year while they were running train on a bunch of schools with great rankings
it comes down to winning games, not what some efficiency says. Clemson actually has more q1 wins than Ohio State and more Q1/2 wins.
Clemson is 3-2/7-2 while Ohio State is a woeful 2-8/5-9
I would go with Clemson here. Ohio State has lost 7 of 8...sorry that is pathetic, any ranking system having a team with 10 losses, and 7 of 8 losses should not be ranked 29
Rutgers was 80 something in the net last year while they were running train on a bunch of schools with great rankings
it comes down to winning games, not what some efficiency says. Clemson actually has more q1 wins than Ohio State and more Q1/2 wins.
Clemson is 3-2/7-2 while Ohio State is a woeful 2-8/5-9
I would go with Clemson here. Ohio State has lost 7 of 8...sorry that is pathetic, any ranking system having a team with 10 losses, and 7 of 8 losses should not be ranked 29
the issue seems to be every year the NET is having some schools get caught in a feedback loop...see Connecticut, Florida Atlantic, Ohio State, Clemson, Rutgers last year....there are enough outliers where you can say win loss record is not being factored highly enough in the team value index. I get it,, its a body of work thing, but geez OSU could sustain 2 more losses and be 11-12 and still be ranked in the top 40
oh another big joke is St Marys at 6...the Gaels have exactly one win over a school projected in the tournament....San Diego State plus 2 quad 3 losses and the are FREAKING 6 in the NET
oh another big joke is St Marys at 6...the Gaels have exactly one win over a school projected in the tournament....San Diego State plus 2 quad 3 losses and the are FREAKING 6 in the NET
Latest posts
-
-
-
BB Recruiting Rivals Rankings Week: Rutgers finishes with No. 2 class
- Latest: -RUFAN4LIFE-
