NET and KenPom Tracker 23-24 (SU = 84 3/9/24)
- Thread starter Sbucketball
- Start date
atlanticuse
Starter
- Joined
- Jul 8, 2022
- Messages
- 1,689
- Like
- 2,824
probably a million explanations - I found this one quickly: https://www.si.com/college/2018/08/22/ncaa-net-ranking-system-rpi-replacement-explained
Probably whining by teams who didn't get in. Or Joe Lunardi was frustrated that Syracuse kept sneaking in every year off the bubble.
orangefog
Certified Message Board Genius
- Joined
- Aug 26, 2011
- Messages
- 17,814
- Like
- 58,260
It just seems odd when you beat teams head to head that don't have any more impressive wins than what we have and they are significantly higher.
hoopsupstate
Living Legend
- Joined
- Aug 14, 2011
- Messages
- 12,917
- Like
- 29,550
People are underrating Wake. They are good since their big man became eligible. They might pummel us inside.
The only positive thing you can say about the #5 kenpom
ranked team is they have 5 losses against quality
opponents. Wins against Morehead State, Indiana State,
S Alabama, Mercer, Arkansa State, Eastern Ky, Liberty,
and Oregon just don't seem all that impressive.
8-5 with this resume doesn't make them the 5th best
team out there.
ranked team is they have 5 losses against quality
opponents. Wins against Morehead State, Indiana State,
S Alabama, Mercer, Arkansa State, Eastern Ky, Liberty,
and Oregon just don't seem all that impressive.
8-5 with this resume doesn't make them the 5th best
team out there.
atlanticuse
Starter
- Joined
- Jul 8, 2022
- Messages
- 1,689
- Like
- 2,824
as we've seen, a team with a 5'8" guard can pummel us inside.
see: McLeod.
Fly Rodder
Hall of Pretty Good
- Joined
- Mar 26, 2012
- Messages
- 10,835
- Like
- 32,383
CardiacCuse
All Conference
- Joined
- Oct 18, 2018
- Messages
- 3,202
- Like
- 6,146
all of their losses were pretty close - most within 10 points
and all of their wins have been blowouts
their overall plus minus must be really good.
seems the NET really punishes teams for getting blownout and really rewards teams for blowing out creampuffs
I guess the NET is looking for those teams that the top tier cant blowout.
it makes sense if NCAAB secretly wants to avoid blowouts in the NCAAT - if so, they have NET tuned perfectly.
It's crazy bc I think most fans have a totally different view and only care about raw win loss record, not margin of victory...its also crazy bc teams and players dont seem to care much about maxing out how much they win by, though they should.
coaches need to triple down to their players to avoid getting blown out at all costs, if they care about their NET ranking, imo
and also run up the score as much as possible, whenever possible...SU has often had wins in the bag and let the margin get closer...which has hurt them more than they realize in the NET, imo
Last edited:
Fly Rodder
Hall of Pretty Good
- Joined
- Mar 26, 2012
- Messages
- 10,835
- Like
- 32,383
From the college basketball subreddit on NET:
In other words, don't take plays off. And a good comment from that reddit thread: "It's a lot harder than people think to run up the score on inferior teams, especially as those margins are adjusted relative to how much other teams run up the score on the same teams."
He developed his version NET and posted the guts to Github. Go build your own ranking system.
Here are his Overall NET scatterplots and the ACC one (do note that the y-axis is log scale, which means that the teams in the mushy middle haven't done much to separate themselves). "The EFFICIENCY is the difference in the score per 100 possessions. The remaining component is the “Team Value Index” which only considers wins, opponents, and location." He explains that the efficiency and value are NOT weighted 50-50, they're more like "Efficiency ranking at 80% and the Value ranking at 20%." Which means how you beat teams relative to how other teams have played those teams is more important in predicting future outcomes.
Looking at the graph, SU is one of the better teams in the ACC by team value, but they're not very efficient with their possessions.
The author of the above is a contributor to the NC State SB Nation page and he has a solid write up of NET and how it works (and what it is used for). "Should teams run up the score? Yeah, probably. Well, let’s just say that it’s always a good time for a good possession. For the NET, it’s the same thing to win by 24 points in 60 possessions as it is to win by 32 points in 80 possessions. So you don’t have to go at a fast pace. But if you beat Green Bay by 10 and others are beating Green Bay by 40, you’re probably going to look bad to the NET."Short answer: The current NET does not cap the margin of victory.
Long answer: The old NET (2019-2020) had a "Scoring Margin" component, which was the average of your point differential, capped at 10 points per game. I don't think this was a major component, but it was there. The old NET also had a "Net Efficiency" metric, which was your advantage in points per possession. That was uncapped and not adjusted for strength of opponent. The new NET (2021-present) doesn't use the Scoring Margin and it does adjust the Efficiency for opponent strength. So there isn't diminishing returns, but you only get credit for beating bad teams to the extent that you beat those teams by MORE than what other teams beat those bad teams by.
ncaabball/README.md at main · bchare/ncaabball
Contribute to bchare/ncaabball development by creating an account on GitHub.github.com
In other words, don't take plays off. And a good comment from that reddit thread: "It's a lot harder than people think to run up the score on inferior teams, especially as those margins are adjusted relative to how much other teams run up the score on the same teams."
He developed his version NET and posted the guts to Github. Go build your own ranking system.
Here are his Overall NET scatterplots and the ACC one (do note that the y-axis is log scale, which means that the teams in the mushy middle haven't done much to separate themselves). "The EFFICIENCY is the difference in the score per 100 possessions. The remaining component is the “Team Value Index” which only considers wins, opponents, and location." He explains that the efficiency and value are NOT weighted 50-50, they're more like "Efficiency ranking at 80% and the Value ranking at 20%." Which means how you beat teams relative to how other teams have played those teams is more important in predicting future outcomes.
Looking at the graph, SU is one of the better teams in the ACC by team value, but they're not very efficient with their possessions.
Last edited:
Knicks411
Living Legend
- Joined
- Aug 29, 2011
- Messages
- 13,799
- Like
- 12,462
I don't know, this is just a guess, (and I'm not sure there is any real way to prove this, so this will be fun) but I tend to think the line of thought that "team x (usually the team you root for!) is undervalued by scoring margin because sometimes they let up in games they have in hand and the score is closer than it 'should be' at the end" is vastly overstated. Does it happen from time to time? I am sure it does. I am also sure there are times when the team getting beat down throws in the towel and starts chucking long 3's that lead to dunks on the other end that extend the lead. Or when you're getting blown out and the other team lets up and you make a run.
Additionally, sometimes when you're blowing a team out by 20 points through 34 or 35 minutes and they make a little run at the end to make it 13 or whatever, that's not just because you "let up". It might also be because you really weren't 20 points better than they were and they went on a run themselves because basketball is a game of runs and sometimes you're on the wrong side of it.
Like I said, this is all a gut feel, and maybe someone has done the research on this to prove me wrong, but I really think this effect is vastly overstated over the course of a 30 game regular season.
Edited: Cleaning the glass on the NBA side does this thing where they filter our garbage time stats (not exactly sure how they define it, will see if I can check.) I compared teams overall net rating (scoring margin per 100 possessions) to the net rating filtering out garbage time. The average of the absolute differences was about 0.35 points per 100 per team. The largest difference was 1 point per 100, which was Memphis. 21 of the 30 teams were under half a point per 100. Would be curious to see something similar done at the NCAA level.
Additionally, sometimes when you're blowing a team out by 20 points through 34 or 35 minutes and they make a little run at the end to make it 13 or whatever, that's not just because you "let up". It might also be because you really weren't 20 points better than they were and they went on a run themselves because basketball is a game of runs and sometimes you're on the wrong side of it.
Like I said, this is all a gut feel, and maybe someone has done the research on this to prove me wrong, but I really think this effect is vastly overstated over the course of a 30 game regular season.
Edited: Cleaning the glass on the NBA side does this thing where they filter our garbage time stats (not exactly sure how they define it, will see if I can check.) I compared teams overall net rating (scoring margin per 100 possessions) to the net rating filtering out garbage time. The average of the absolute differences was about 0.35 points per 100 per team. The largest difference was 1 point per 100, which was Memphis. 21 of the 30 teams were under half a point per 100. Would be curious to see something similar done at the NCAA level.
Last edited:
look at Lunardi.. Oregon is now on the bubble
They have beaten 6-7 USC/UCLA/Mich
Good game 1 over Georgia
Lost to Ala
Lost to SU
Lost to Santa Clara
SU has a win over them
Lost to Tenn/Gonz
Lost to Virg
But they are much higher in BPI and the like in the rankings
jncuse
I brought the Cocaine to the White House
- Joined
- Feb 19, 2012
- Messages
- 19,209
- Like
- 32,804
Getting fussed over the "as of now" bracket seedings this early in the year is just going to drag you into a time wasting hole. Just look at brackets as a fun exercise at this point, with little more meaning than that.
There is limited "meaningful" data in teams resumes due to a high amount of OOC cupcake games (or Q3 and Q4) for most. As a result doing an "as of now" selection often is a combination of your view of the team (independent of results) or focusing highly on a couple of games on a schedule to pick teams apart.
Of course that doesn't mean what has happened doesn't matter. Of course it does -- you want to build up as much as you can all year, and avoid stupid losses, but the real meat out of your resume is going to get formed in conference season.
Which is why its so important that your conference does good OOC, so it gives you more Q1 and Q2 opportunities in January and February,
There is limited "meaningful" data in teams resumes due to a high amount of OOC cupcake games (or Q3 and Q4) for most. As a result doing an "as of now" selection often is a combination of your view of the team (independent of results) or focusing highly on a couple of games on a schedule to pick teams apart.
Of course that doesn't mean what has happened doesn't matter. Of course it does -- you want to build up as much as you can all year, and avoid stupid losses, but the real meat out of your resume is going to get formed in conference season.
Which is why its so important that your conference does good OOC, so it gives you more Q1 and Q2 opportunities in January and February,
Sbucketball
All Conference
- Joined
- Mar 31, 2022
- Messages
- 2,254
- Like
- 8,057
We're at 77 in the NET, which is 1 higher than the day before the Duke game. No harm done, but obviously we need to capitalize on opportunities like that to get on the bubble. Some of the teams we beat are creeping a bit higher which is helpful. Really need Oregon to keep climbing since it's close to a Q1 win.
blust2i3d4
Starter
- Joined
- Dec 6, 2016
- Messages
- 1,688
- Like
- 1,500
2022 NCAA Tournament Bracket Predictions - Bracketology - College Basketball
CBS Sports is helping you get ready for March Madness with the latest news, picks, and predictions for the 2022 NCAA Basketball Tournament Bracket.
CBS picked SU as the 12th seed in their bracketology.
Last edited:
Sbucketball
All Conference
- Joined
- Mar 31, 2022
- Messages
- 2,254
- Like
- 8,057
With Oregon's big move of 11 spots, Cuse moved up 2 spots to 75. I'm curious as to whether Cuse will get a nice bump if Oregon moves up 1 more spot, making it a Q1 victory.
I have a feeling NET looks at gross number for calc purposes, not the quad impact? I could be wrong but would seem weird to factor a win over a #1 ranked team on the road the same as a win over #75
Sbucketball
All Conference
- Joined
- Mar 31, 2022
- Messages
- 2,254
- Like
- 8,057
What you're saying makes sense. It does make a difference to at least have a tally in the Q1 section when it comes down to the people actually selecting the field though, and it would be significant if we miss that Q1 win by one spot. But yeah you're probably right for the formula it wouldn't make a big difference.
Oh 100%
hoopsupstate
Living Legend
- Joined
- Aug 14, 2011
- Messages
- 12,917
- Like
- 29,550
Who was our last quad 1 win?
Nc State i think
jncuse
I brought the Cocaine to the White House
- Joined
- Feb 19, 2012
- Messages
- 19,209
- Like
- 32,804
Correct. Road win early February 2021, and it was barely a Q1 victory as NC ST was NET 73 at season end just below the #75 threshold.
Since then we have lost a staggering 24 consecutive Q1 games .. which would end at 23 if Oregon swings its status to Q1 by season end.
Went to Warren Nolan who has accessible game listing data back to 2021.
Last edited:
BomberOrange
2019-20 Iggy Co-Winner Reg Season Rcd
- Joined
- Sep 9, 2011
- Messages
- 1,917
- Like
- 3,591
When is the quad number set? End of season? I’ve always wondered if it’s when the game is played or if is where you end up.
Similar threads
