Nah, I mean that they are equal after doing all the adjustments. What you are describing is already part of every computer rating system.
OFFICIAL NET Thread - 2022/23
- Thread starter fatsam98
- Start date
Also, to be clear, I'm not saying that 30 games into a season you need to use prior year data to make decent predictions. I'm just saying that the relevance of prior expectations slowly fades as you get more current year data and never completely reaches zero. The current season is obviously far more important once you have a good sample of games.
Rutgers down a spot to 39. Northwestern moves to 38. Illinois tumbles from 23 to 43. Penn State beats Quinnipiac by 9 and falls from 37 to 49. Maryland jumps from 28 to 23.
Nc State beat Louisville by 12 and fell from 42 to 56.
To keep from dropping RU needs to win by at least 15 tonight
To keep from dropping RU needs to win by at least 15 tonight
Penn State's resume took a hit outside of their game vs. Quinnipiac
Illinois (who they beat by 15) dropped from 37 -> 49 after a beating vs. (then 65) Mizzou on a neutral court
Butler (who they beat by 6) dropped from 53 -> 67 after a beating at (then 51) Creighton
NC State's resume also dragged them down a bit:
Butler (who they beat by 15) dropped from 53 -> 67
Vandy (who they beat by 4) dropped from 138 -> 147
Several teams start too high in the rankings based on name and last year's performance. When those teams underperform, people act shocked. Maybe they aren't as good as advertised. Teams lose players, there are injuries, and players don't develop.
This is NET, not kenpom or sagarin. No one starts based on name or last year's performance.
NET is based entirely on current season actuals - which means early season you see teams beat up on cupcakes and soar to the top, then lose their next game and plummet. We're still less than halfway into the season, so a bad loss can still have double-digit swings. NET is much more volatile, because it doesn't have a cushion of prior season data.
It's like a kid who gets a 100% on his first math quiz that is mostly review material from the prior year.... are they an A student? At that moment, yeah. If they get a 50% on their next quiz, though, that GPA drops fast.
As the season progresses, reality starts to emerge as more and more data points are added.
The thing we're seeing now is that some teams are starting to run into reality (Iowa, Illinois, etc) and their NET is dropping as a result.... but that has ripple effects. Teams that pointed to Illinois or Iowa as one of their best wins (or worst loss) are also dropping, because the model no longer sees those teams as strong as it did.
Each big upset is going to immediately impact the teams who played... but also impact the teams that had previously played either of those teams... and subsequently impact the teams those secondary teams had previously played, etc.
Later in the year, there's much more stability, and a single bad result won't have as strong of a ripple effect.
Each big upset is going to immediately impact the teams who played... but also impact the teams that had previously played either of those teams... and subsequently impact the teams those secondary teams had previously played, etc.
Later in the year, there's much more stability, and a single bad result won't have as strong of a ripple effect.
I thought the NET was no different than Kenpom, Sagarin, etc. I admit I'm not a basketball expert.
NET has two parts, the exact weightings are not disclosed:I thought the NET was no different than Kenpom, Sagarin, etc. I admit I'm not a basketball expert.
(1) net efficiency margin. This is extremely similar to Kenpom, but it doesn't use any prior expectations. Kenpom has a preseason component baked in for most of the season.
(2) team value index. This is some RPI-like thing, but it's exact formula is undisclosed. I wouldn't be entirely surprised if it were just the actual RPI
(1) seems to be more important than (2)
And 1 seems to correct for the flaws of 2 (you don’t see teams getting as penalized for demolishing the 350th NET team the way RPI worked. They don’t necessarily move up - but as long as they win easily they aren’t taking the huge hits we used to see after a dominating win) but creates a new major flaw (rewarding meaningless style points).
The blended average flaw from RPI was carried over to NET but perhaps isnt as blatant in the formula output. With RPI - beating 4 teams in the 170s could be the same as notching 2 top 25 wins and 2 dominant wins in the 300s (average RPI of the 4 teams in each sample coming to around the same amount). Very flawed.
so the NET which worked against us in so many ways is now beyond giving and I really think the coaching staff knows it, they have to. Now its about just getting those 5-6 quality conference wins and avoiding a bad conference loss since we already have one in Temple
Last year we were not only losing to Lafayette and UMass but squeaking by Lehigh and putting up mediocre performances against NJIT and Merrimak and the like. Just replacing those performances with complete tail-kickings of all our low major opponents is making a huge difference.
With that OSU win we're probably top 20 NET after tonight.
Meaning if the refs didn't eff up so bad we rightfully had a W against OSU on the road. I think with that W we'd have been top 20 NET tomorrow.
Oh, yeah, maybe. I didnt understand your post at first.. thought you were saying OSU had a big win tonight or something
Yeah I didn't explain myself very well. That road W would have been such a big star on the resume.
This. Exactly. The constant schedule crying is so ridiculous. NET will be perfectly fine even with a ton of horrible luck this season
RU with a MASSIVE jump in NET after obliterating Bucknell. Up 12 spots to #27 as of today.
That is why you win by 35, it separates the NETmen from the NETboys.
Wow even more than I thought. Great news and that's with a Q4 L and not having that OSU(15) W.
I almost feel bad for all the Sisters of the Poor schools.
Love seeing Michigan second to last in Big Ten NET
The average NET conference standings is interesting nearing the end of non conference play.
1) Big 12 31.70 (10 teams)
2) B1G 57.00 (14 teams)
3) SEC 74.79 (14 teams)
4) Big East 85.18 (11 teams)
5) Mountain West 93.82 (11 teams)
6) Pac 12 97.12 (12 teams)
7) West Coast 113.90 (11 teams)
8) Always Cheat C 114.60 (15 teams)
9) AAC 137.82 (11 teams)
10) CUSA 148.64 (11 teams)
1) Big 12 31.70 (10 teams)
2) B1G 57.00 (14 teams)
3) SEC 74.79 (14 teams)
4) Big East 85.18 (11 teams)
5) Mountain West 93.82 (11 teams)
6) Pac 12 97.12 (12 teams)
7) West Coast 113.90 (11 teams)
8) Always Cheat C 114.60 (15 teams)
9) AAC 137.82 (11 teams)
10) CUSA 148.64 (11 teams)
B1G is weighted down because Minnesota's albatross NET ranking of 248.
The averages they are doing don’t really make sense. I know they don’t have access to it, but the proper way is really to average the ratings, not the rankings. The gaps are not close to even, the closer you are to the middle the more packed together everything is.
Every year the fans of Minny are critical of Rutgers. They say things like Rutgers is an easy win and, at best, an average team. There are Minny fans as recent as this week still thinking they are not the worst team in the B1G. Some think Rutgers, Northwestern, and Nebraska are bad.
Honestly I don't care much for Minnesota fans. I feel like they're snobby for no reason.
I think the sport reference guys do a better gauge of conference rankings and the SRS.
1) Big 12 15.99
2) B1G 13.21
3) SEC 11.25
4) Big East 11.18
5) Pac 12 11.18
6) MWC 8.34
7) ACC 7.52
8) WCC 7.19
9) AAC 6.53
10) CUSA 1.96
11) A10 1.43
12) WAC 0.36
CUSA and WAC have replaced Missouri Valley and Southern Conference as the next 6 this year with transfer portal and conference tealignments 22-23.
I'd say it's more like very proud, with a chip on their shoulders as flyover state that makes them overly defensive.
I think they are snobby because they fear Rutgers and Maryland would push them further down the pecking order of the B1G when both teams hit their stride in all sports. It's the same reason Northwestern, Indiana, Iowa, and Illinois come at us.