Bracketology

Status
Not open for further replies.
#51      
I'm not confused if you are referencing me. I am just stating facts. And the win percentage of your opponents is just another stupid thing the NET and formerly the RPI does. EDIT: Of course the poorly implemented NET rankings determine the quad positions of your opponents which may or may not effect the teams seed in the tournament.
The win percentage of your opponents does not affect the NET, that was an artefact of the RPI. The efficiency of your opponent matters for the adjustments, as beating a 90% efficient team by 20 adjusts higher compared to beating a 50% efficient team by 20.
 
#52      

danielb927

Orange Krush Class of 2013
Rochester, MN
If I am not mistaken, people largely understand this. They just take issue with how the Committee/NCAA/whatever are using the NET Rankings to define the Quads. For example, I find the idea that beating the #76 team at home is roughly analogous to beating the #136 team in their gym to be quite suspect. I think they would do well to "tighten up" the differences for home, away and neutral, with the answer being to be less harsh on teams for home victories - because I do like rewarding teams for winning away from home, where the NCAA Tournament is played ... I just don't think the jump down from 75 to 30 for that difference is reasonable.

If the goal is really apples-to-apples difficulty, the best metrics we have say the spreads should be wider, if anything.
 
#53      
The win percentage of your opponents does not affect the NET, that was an artefact of the RPI. The efficiency of your opponent matters for the adjustments, as beating a 90% efficient team by 20 adjusts higher compared to beating a 50% efficient team by 20.
They say they don't use opponent win percentage but every teams ranking is based on win percentage, and current team ranking is based on opponent. So its still there.
 
#54      
They say they don't use opponent win percentage but every teams ranking is based on win percentage, and current team ranking is based on opponent. So its still there.
For the third time, no, it's based on efficiency margin, which is not the same as win percentage. Which was the point of my initial post trying to clarify for those confused, which you've now doubled down on twice incorrectly.
 
#55      
what team do you think will be under seeded this year and are not getting enough respect? For this teams outside of the top 16
 
#56      
what team do you think will be under seeded this year and are not getting enough respect? For this teams outside of the top 16
Major conference team: Auburn

Mid-majors: Dayton, St. Mary's, New Mexico
 
#57      
For the third time, no, it's based on efficiency margin, which is not the same as win percentage. Which was the point of my initial post trying to clarify for those confused, which you've now doubled down on twice incorrectly.
Ok. I don't have access to the exact algo, but why did the NCAA publish this explanation which clearly shows win loss in the updated NET calculation? Its in the team value index. final edit: I'm done with this. we are off in the weeds here.
 

Attachments

  • NET_calc.png
    NET_calc.png
    804.7 KB · Views: 88
Last edited:
#58      
For the third time, no, it's based on efficiency margin, which is not the same as win percentage. Which was the point of my initial post trying to clarify for those confused, which you've now doubled down on twice incorrectly.
It doesn't Really matter what it's based on. They take liberties every year with the all the data they are supposed to use when seeding and making the bracket. It never turns out like everyone thinks it would. Just a lot of wasted conjecture on this thread in my opinion. It will be what it is and we'll find out on selection Sunday.
 
#60      
What I don’t see referenced or talked about is that we went 3-2 in B1G play without arguably our best player. Which includes the dreaded Maryland home loss. And a close road loss to Purdue. I thought the committee took these things into consideration when comparing/seeding teams.
 
#61      

Illini2010-11

Sugar Grove
What I don’t see referenced or talked about is that we went 3-2 in B1G play without arguably our best player. Which includes the dreaded Maryland home loss. And a close road loss to Purdue. I thought the committee took these things into consideration when comparing/seeding teams.
I would think the Maryland loss (if it were to become Quad 3) would be looked at by committee. But it is extremely unlikely that those 5 games will mean much during seeding in the grand scheme of things. The committee takes a lot into consideration (metrics, human factors, etc.).
 
#63      

Illini2010-11

Sugar Grove
I have a feeling Gonzaga's late season run will get them a 6 or 7 seed, which seems about right for them.
I guess it is possible (Saturday was a very important win for them). They were hovering around the 10/11 line before that dominant win over St. Mary's. The WCC tourney will be important for their final seeding. Win that, and I think you may be right. Get shocked early, then they could be a 9, 10, or 11 seed.
 
#64      
Ok. I don't have access to the exact algo, but why did the NCAA publish this explanation which clearly shows win loss in the updated NET calculation? Its in the team value index. final edit: I'm done with this. we are off in the weeds here.
They've never released the algorithm, they’ve just said in a general sense what goes into your rating (your record, strength of opponent, net efficiency, and game location), so there's a lot of conjecture of how they actually use that data. However one thing that we do know is that as of 2020 they removed opponent win percentage along with margin of victory as it wasn't having the effects they wanted. And we know this because they announced it.

So the governing thought process is that for any given game you play against an opponent you'll have a net efficiency based on pts for and against divided by possession. Those values are then adjusted for location to what they would theoretically be if the game was held on a neutral court, then that value is compared to the other modified efficiency values that opponent has had. And based on that differential your own strength goes up or down accordingly. It's important to note that this is an iterative process as you'll start with a test value (i.e. everyone is equal) and you'll get all teams strength as an output. You then take that output and put it right back in as your input, and you keep doing that until your values converge to where your input and output values for strength are identical (within an error you set for it).

In any case, that's the 1 minute read crash course of what is going on here at a very basic level. So opponent w-l record isn't actually accounted for in the algorithm itself and doesn't directly contribute to a team's NET rating. Instead NET reports out your wins and losses organized by NET rating strength of opponent per their adjusted efficiency numbers. Hope that helps explain things a bit more clearly
 
#65      
Ok. I don't have access to the exact algo, but why did the NCAA publish this explanation which clearly shows win loss in the updated NET calculation? Its in the team value index. final edit: I'm done with this. we are off in the weeds here.
Each team’s NET ranking is based partly on win percentage. But “strength of opponent” has nothing to do with your opponents’ NET ranking. “Strength of opponent” is based on opponents’ efficiency ratings (which are only a PART of the NET). Auburn, for example, has been efficient and beaten some efficient teams, which are viewed as “strong” regardless of their actual win-loss records. So they are high in the NET. However, the efficient teams they’ve beaten don’t necessarily have a high win percentage, and are therefore not very high in the NET, thus resulting in Auburn’s poor Q1 record.
 
#66      

Illini2010-11

Sugar Grove
They've never released the algorithm, they’ve just said in a general sense what goes into your rating (your record, strength of opponent, net efficiency, and game location), so there's a lot of conjecture of how they actually use that data. However one thing that we do know is that as of 2020 they removed opponent win percentage along with margin of victory as it wasn't having the effects they wanted. And we know this because they announced it.

So the governing thought process is that for any given game you play against an opponent you'll have a net efficiency based on pts for and against divided by possession. Those values are then adjusted for location to what they would theoretically be if the game was held on a neutral court, then that value is compared to the other modified efficiency values that opponent has had. And based on that differential your own strength goes up or down accordingly. It's important to note that this is an iterative process as you'll start with a test value (i.e. everyone is equal) and you'll get all teams strength as an output. You then take that output and put it right back in as your input, and you keep doing that until your values converge to where your input and output values for strength are identical (within an error you set for it).

In any case, that's the 1 minute read crash course of what is going on here at a very basic level. So opponent w-l record isn't actually accounted for in the algorithm itself and doesn't directly contribute to a team's NET rating. Instead NET reports out your wins and losses organized by NET rating strength of opponent per their adjusted efficiency numbers. Hope that helps explain things a bit more clearly
Explaining Fran Healy GIF by Travis
 
#67      
There is way too much math in this thread. Here's what I know, I will take our guys against the field all day, every day. That's what the Big Dance is, not the number next to your seed. Beat the crap out of the other 5 guys that walk out on the court to play you.
 
#68      
I'm old school, you gotta beat the best to be the best. I honestly think if a team on selection Sunday looks at seeding and who is in their bracket and starts worrying about future matchups (ex: UCONN or HOUSTON) they are already losing half the battle. I dont care if its 64 teams and Illinois is number 64. The way this team is playing and jell'n at the right time, we should walk onto any court thinking we are going to win and it's the other teams that shouldn't want to be anywhere near us in the brackets....period. I also tend to drink a lot of orange kool aid but this year...I just feel we have a group that really does believe in one another and that they can beat anyone. ILL baby!!!!!!!!!!!
 
#69      

Chad Fleck

Eureka, IL
I'm old school, you gotta beat the best to be the best. I honestly think if a team on selection Sunday looks at seeding and who is in their bracket and starts worrying about future matchups (ex: UCONN or HOUSTON) they are already losing half the battle. I dont care if its 64 teams and Illinois is number 64. The way this team is playing and jell'n at the right time, we should walk onto any court thinking we are going to win and it's the other teams that shouldn't want to be anywhere near us in the brackets....period. I also tend to drink a lot of orange kool aid but this year...I just feel we have a group that really does believe in one another and that they can beat anyone. ILL baby!!!!!!!!!!!
That should be their attitude for sure.

But us lowely message board fans can go ahead and agonize over minuit details.
 
#70      

DeonThomas

South Carolina
What I don’t see referenced or talked about is that we went 3-2 in B1G play without arguably our best player. Which includes the dreaded Maryland home loss. And a close road loss to Purdue. I thought the committee took these things into consideration when comparing/seeding teams.
You are correct.
 
#74      

danielb927

Orange Krush Class of 2013
Rochester, MN
Gonzaga is the early round landmine I'm hoping we can avoid right now. KenPom has them 15th, Torvik at 14th, and EvanMiya at 12th in the nation - indicative of a high 4 / low 3 seed in terms of actual quality.

Despite that, they were a bubble team as recently as last week, and I was worried about a 3/11 game against them.

They picked up their 2nd and 3rd Q1 wins and now seem solidly in. They've moved up to the 8 line on Bracket Matrix, which would be great - ideally they'll stick around at 8/9 and make things tough for a 1 seed in the second round. But if they move further up and grab a 6 seed, we could see them again.

The good news (I hope) is that they shouldn't play in Spokane, since that would be heavily disadvantaged towards the top-4 seed in their pod. But I still hope they muddle along and stick around the 7-10 range. Or, if we don't grab a 3, a silver lining of a 4 seed is that we also almost certainly wouldn't see them in the first 2 rounds.
 
#75      

chrisRunner7

Spokane, WA
Gonzaga is the early round landmine I'm hoping we can avoid right now. KenPom has them 15th, Torvik at 14th, and EvanMiya at 12th in the nation - indicative of a high 4 / low 3 seed in terms of actual quality.

Despite that, they were a bubble team as recently as last week, and I was worried about a 3/11 game against them.

They picked up their 2nd and 3rd Q1 wins and now seem solidly in. They've moved up to the 8 line on Bracket Matrix, which would be great - ideally they'll stick around at 8/9 and make things tough for a 1 seed in the second round. But if they move further up and grab a 6 seed, we could see them again.

The good news (I hope) is that they shouldn't play in Spokane, since that would be heavily disadvantaged towards the top-4 seed in their pod. But I still hope they muddle along and stick around the 7-10 range. Or, if we don't grab a 3, a silver lining of a 4 seed is that we also almost certainly wouldn't see them in the first 2 rounds.
I barely follow Gonzaga but odds are we won't play them... I doubt we wind up with a 3 seed. Would love to be wrong. Think we would have to either beat Purdue or Iowa plus win two games in the BTT.

With the crazy parity this year, I feel like there are landmines all over the bracket. Just looking at the 10 seeds on Bracket Matrix right now... I know Virginia has been garbage for a lot of the season but I still wouldn't like facing them:

10 Nebraska
10 Michigan State
10 Florida Atlantic
10 Virginia
 
Status
Not open for further replies.
Back