Recursive is a great word
NET has a lot to do with efficiency ratings and margin of victory is very adjacent to that.
As stated above, NET sorts teams. Then, the quadrants are an after-the-fact statistic derived from the sorts
2024-2025 MBB computer projections thread
- Thread starter Sigmapolis
- Start date
NET has a lot to do with efficiency ratings and margin of victory is very adjacent to that.
As stated above, NET sorts teams. Then, the quadrants are an after-the-fact statistic derived from the sorts
CycloneErik
Well-Known Member
I think the biggest thing is consistency - their Torvik game scores have all been 88 or above - and SDSU was the only one under 94. All but 4 games have been 98 - 100. It's completely crazy to even consider that
yep, quadrants are a way to basically understand the data differently. Usually easier for humans to understand "they're 5-5 against great teams, but 9-1 against good teams" than "they're .9423".
I always look at the Sagarin FB rankings that way - he puts teams' records vs his top 30 and top 10. Kind of helps separate the sheep from the goats, in tiers.
Drives me insane when I see people bring that up. NET is just another predictive metric similar to Kenpom, Torvik, BPI that ranks how good teams are. The only difference, and in my opinion, major issue with NET, is that we can't see the actual values of the NET; we only see the rankings.
When I see complaints to the effect of "we dropped from 4 to 7 after one game, stupid!" I wonder how one can say that when no one has any idea how far apart 4 and 7 are in terms of the actual NET value. I know the NCAA will never share the math, but I wish they'd share the number that determine the rankings.
Last edited:
Houston is currently in their easy stretch of conference games but have a couple difficult stretches of games. One coming up here next week and then the other starts in mid February.
In my mind, here is Iowa State's path to a Big 12 title and #1 seed in the tournament. No, this is not the only path they can take, but in my mind this seems attainable for this team.
1.) Win @WVU this weekend.
2.) Chalking up a win @Okie State (they are horrid)
3.) That leaves SIX road games (@ASU, @UCF, @KState, @Arizona, @Houston, @Kansas) Find a way to go 3-3 in those games.
4.) Protect Hilton and win out at home. Outside of the Arizona game, we should be 10+ point favorites in all remaining home games. I have a hard time seeing this team lose in Hilton.
That path is a 4 loss team. Big 12 champion and #1 seed in the tournament.
1.) Win @WVU this weekend.
2.) Chalking up a win @Okie State (they are horrid)
3.) That leaves SIX road games (@ASU, @UCF, @KState, @Arizona, @Houston, @Kansas) Find a way to go 3-3 in those games.
4.) Protect Hilton and win out at home. Outside of the Arizona game, we should be 10+ point favorites in all remaining home games. I have a hard time seeing this team lose in Hilton.
That path is a 4 loss team. Big 12 champion and #1 seed in the tournament.
OR... hear me out... we just win them all
Torvik has us as underdogs @ Zona, @ KU, @ Houston - and a 6-pt favorite over Zona in Ames
yeah, they've got 11 games remaining against Q1 teams, while we only have 7. We have 8 Q2/3 games left, while they have 4.
Hopefully that can result in them picking up a couple losses more than us down the stretch
Cyclonepride
Thought Police
I looked at our remaining schedule and I could see anywhere from 15-5 to 18-2.
The fact that they are #3 in NET shows just how flawed that formula is.We are 15/16 games into the season and Houston still does not have a Quad 1 win.
KP has them 3rd, and BT has them #1. So not just NET.
It's the challenge in any model that has to compare results against different opponents and extrapolate. Not to mention the inherent variability in performance game to game.
Everything matters when the committee's criteria is a friggin moving target...
Hey, CU, do you suppose you could win a game or two? You are currently 1/8 of our schedule played and we still have to play you a third time. You are dragging down our metrics...
Yeah I get that but it wasn't originally designed to fall in line with those models. It seems like over the last 2-3 seasons they have put more weight into the offensive efficiency numbers. It really contradicts the committees "who did they play/beat" narrative that they always fall back on.
i dont understand net, truly. Im sure you can figure out the 2 teams here.
Bolded are the obvious hardest remaining games. If they can steal 1/3 of those that'll be a huge. I saw somewhere they're battle tested in terms of playing the better competition actually better. So they step it up in big games. It is nice having an array of scorers on the team versus 1-2 and can easily be shut down if they have an off night. Clearly there are more scorers than just Gilbert and CJ. I love Jefferson's game a lot and Dishon took it personal that he didn't start against KU.
Last year it was all about how our non-con was weak even though our Conference schedule was rigorous.
Duke this year is the opposite. Strong Non-con with a weak conference schedule because the ACC is stinky doodoo. We'll we hear the nashing of teeth about this? We all know the answer because it's Duke.
