Bracketology 2024

forensicbuzz

21st Century Throwback Dad
Messages
8,803
Location
North Shore, Chicago
This is a binary approach to a probability problem. Even 1 vs. 16 isn't 100% as we have seen. 8 vs. 9 is essentially 50-50.
I would go as far as to say that in the 5/12 game, the 12 is the favorite. I think in most tournaments, there are an inordinate number of 12 seeds in the 2nd round.

Besides, this isn't about seeding, it's about seats at the table. the SEC/B12 each went from 8 to 2 pretty damn fast. People are talking about the top of each conference being there because they should be there and the teams like Pitt, Wake, UVa, etc. should be there (not as a play-in game) is because top-to-bottom, the teams in the ACC, at the end of the season, were stronger. GT was a 12 seed in the ACC tournament. We beat 3 of the 4 ACC teams (the top seeded ACC teams) this season giving them a Q3 loss. Anyone want to putt forth GT was a Q3 team in February and March? GT actually beat MSU early in the season, when we probably were a Q3 team. I'm not going to argue it, but I do think the ACC got jobbed.
 

Root4GT

Helluva Engineer
Messages
2,963
I would go as far as to say that in the 5/12 game, the 12 is the favorite. I think in most tournaments, there are an inordinate number of 12 seeds in the 2nd round.

Besides, this isn't about seeding, it's about seats at the table. the SEC/B12 each went from 8 to 2 pretty damn fast. People are talking about the top of each conference being there because they should be there and the teams like Pitt, Wake, UVa, etc. should be there (not as a play-in game) is because top-to-bottom, the teams in the ACC, at the end of the season, were stronger. GT was a 12 seed in the ACC tournament. We beat 3 of the 4 ACC teams (the top seeded ACC teams) this season giving them a Q3 loss. Anyone want to putt forth GT was a Q3 team in February and March? GT actually beat MSU early in the season, when we probably were a Q3 team. I'm not going to argue it, but I do think the ACC got jobbed.
I suspect the past 5 or so years the numbers are a bit different in the 4-13 to 6-11 range.

The all time seed-vs-seed records in the first round:​

SEED VS. SEED.W-LPCT.
1 vs. 16150-2.987
2 vs. 15141-11.928
3 vs. 14130-22.855
4 vs. 13120-32.789
5 vs. 1299-53.651
6 vs. 1194-58.618
7 vs. 10*92-59.609
8 vs. 974-78.487
 

stinger78

Helluva Engineer
Messages
4,113
I would go as far as to say that in the 5/12 game, the 12 is the favorite. I think in most tournaments, there are an inordinate number of 12 seeds in the 2nd round.

Besides, this isn't about seeding, it's about seats at the table. the SEC/B12 each went from 8 to 2 pretty damn fast. People are talking about the top of each conference being there because they should be there and the teams like Pitt, Wake, UVa, etc. should be there (not as a play-in game) is because top-to-bottom, the teams in the ACC, at the end of the season, were stronger. GT was a 12 seed in the ACC tournament. We beat 3 of the 4 ACC teams (the top seeded ACC teams) this season giving them a Q3 loss. Anyone want to putt forth GT was a Q3 team in February and March? GT actually beat MSU early in the season, when we probably were a Q3 team. I'm not going to argue it, but I do think the ACC got jobbed.
Not only did the ACC outperform this year, but they have most years over the last 10 years. Sure, there have been 3-4 years in that span that the conference was off, as were all others. There was never a reason to penalize the ACC for poor performance. Yet it has happened. Why?
 

lv20gt

Helluva Engineer
Messages
5,580
UNC making the Sweet 16 is not an argument that Pitt deserved to be in the tournament. That's an UNC accomplishment, not a Pitt one. Conferences don't play the games.

Kentucky losing to Oakland is funny but there is no argument that Kentucky shouldn't have been in the tournament so how is that relevant? Is the argument really Oakland beat Kentucky so Wake should have made it over Mississippi state?

People want to look at the results in the post season but somehow overlook Wake played in the NIT and lost in the second round, to an SEC team no less. So how is the "SEC"'s failure to make it out of the second round of the NCAAT somehow more of an argument for Wake being in the NCAAT than Wake not being able to make it out of the second round of the NIT? What about UVA losing by 25 to a double digit seed? If you want to make an argument that Pitt should have replaced a P5 team then why not a team that they had the same record with, beat on the road by 11 points, and who had probably the worse showing of any P5 team in the tournament? the obvious answer is that the desired conclusion isn't because a team that was worthy was left out, but rather you want to argue the ACC should have gotten more bids so GT would have gotten more money. Pitt replacing UVA doesn't achieve that goal. Everyone should want the ACC to have gotten more bids. Doesn't mean the ACC actually deserved it.

And what is with trying to use performances in past NCAATs as justification for teams being selected this year. How is that at all relevant? And if it is, then isn't that even more reason for UVA to have not been included given that this was there 4th loss to a double digit seed in the last 5 tournaments they appeared in? Obviously tongue in cheek, but why would what Syracuse did 4 years ago or something matter to any team this year?

At best you could argue that Pitt could have been included and it wouldn't have been a travesty. It's also not a travesty that they weren't there. They didn't have an overly strong resume, and probably aren't even the biggest snub.
 

stinger78

Helluva Engineer
Messages
4,113
UNC making the Sweet 16 is not an argument that Pitt deserved to be in the tournament. That's an UNC accomplishment, not a Pitt one. Conferences don't play the games.

Kentucky losing to Oakland is funny but there is no argument that Kentucky shouldn't have been in the tournament so how is that relevant? Is the argument really Oakland beat Kentucky so Wake should have made it over Mississippi state?

People want to look at the results in the post season but somehow overlook Wake played in the NIT and lost in the second round, to an SEC team no less. So how is the "SEC"'s failure to make it out of the second round of the NCAAT somehow more of an argument for Wake being in the NCAAT than Wake not being able to make it out of the second round of the NIT? What about UVA losing by 25 to a double digit seed? If you want to make an argument that Pitt should have replaced a P5 team then why not a team that they had the same record with, beat on the road by 11 points, and who had probably the worse showing of any P5 team in the tournament? the obvious answer is that the desired conclusion isn't because a team that was worthy was left out, but rather you want to argue the ACC should have gotten more bids so GT would have gotten more money. Pitt replacing UVA doesn't achieve that goal. Everyone should want the ACC to have gotten more bids. Doesn't mean the ACC actually deserved it.

And what is with trying to use performances in past NCAATs as justification for teams being selected this year. How is that at all relevant? And if it is, then isn't that even more reason for UVA to have not been included given that this was there 4th loss to a double digit seed in the last 5 tournaments they appeared in? Obviously tongue in cheek, but why would what Syracuse did 4 years ago or something matter to any team this year?

At best you could argue that Pitt could have been included and it wouldn't have been a travesty. It's also not a travesty that they weren't there. They didn't have an overly strong resume, and probably aren't even the biggest snub.
OK. I am going to assume you're being serious.

The argument, so to speak, is that the conference has consistently outperformed all others over the past 10 years - up to last year. Yet, there has been a narrative hatched that the ACC has fallen off and it affects polls and tournament invitations. That's the "argument."

No one - no one - is arguing whether or not UNC or UK deserve to be in the NCAAT. Absolutely no one. This is about equity of invites when a bunch of teams are relatively even. The ACC has been being shafted among those teams now for 5+ years. It is unfairly costing the ACC lots of dollars while (again) enriching the richest. Most of the money from the NCAAT is generated the first weekend when the most games are played. Whether these teams win or lose, they get a share.

This season, the SEC got 8 invites: UT, UK, Bama, and 5 various next-level teams. The B12 also got 8 invites: Houston, Iowa St, and BYU, and 3-4 other next-level teams. Well, the ACC had 3-4 next-level teams as well in addition to UNC, Dook, and Clemson. Namely, Pitt, Wake, UVA, NCSU, and Syracuse. Yet, Pit and Wake, particularly, were left to pound sand and UVA was given a play-in game. Why? The point isw, once you get past the top-shelf teams and are into the next-level teams that are in the same bucket, dispense those more equitably. The ACC, particularly, has shown its competitiveness year after year.
 

ESPNjacket

Helluva Engineer
Messages
1,531
OK. I am going to assume you're being serious.

The argument, so to speak, is that the conference has consistently outperformed all others over the past 10 years - up to last year. Yet, there has been a narrative hatched that the ACC has fallen off and it affects polls and tournament invitations. That's the "argument."

No one - no one - is arguing whether or not UNC or UK deserve to be in the NCAAT. Absolutely no one. This is about equity of invites when a bunch of teams are relatively even. The ACC has been being shafted among those teams now for 5+ years. It is unfairly costing the ACC lots of dollars while (again) enriching the richest. Most of the money from the NCAAT is generated the first weekend when the most games are played. Whether these teams win or lose, they get a share.

This season, the SEC got 8 invites: UT, UK, Bama, and 5 various next-level teams. The B12 also got 8 invites: Houston, Iowa St, and BYU, and 3-4 other next-level teams. Well, the ACC had 3-4 next-level teams as well in addition to UNC, Dook, and Clemson. Namely, Pitt, Wake, UVA, NCSU, and Syracuse. Yet, Pit and Wake, particularly, were left to pound sand and UVA was given a play-in game. Why? The point isw, once you get past the top-shelf teams and are into the next-level teams that are in the same bucket, dispense those more equitably. The ACC, particularly, has shown its competitiveness year after year.
There seems to be denial of the anti-ACC narrative the last 4 years. I don't know how. I don't consume much sports media and I haven't been able to avoid it.

There are something like 10 Peter Burns for each Danny Kanell. I know some of these people pretty well. They aren't playing a role for TV. They believe what they are saying.
 

stinger78

Helluva Engineer
Messages
4,113
There seems to be denial of the anti-ACC narrative the last 4 years. I don't know how. I don't consume much sports media and I haven't been able to avoid it.

There are something like 10 Peter Burns for each Danny Kanell. I know some of these people pretty well. They aren't playing a role for TV. They believe what they are saying.
I think MtnWasp hit on something when he concluded that lots of folks “worship” the linear metrics that are supposed to explain non-linear performance of complex systems. Descriptive metrics absolutely have their place, no doubt, but it is very difficult (many would say impossible) to predict the outcomes of complex systems.
 

MtnWasp

Ramblin' Wreck
Messages
978
OK, thanks. So here's where I was going.

Since the NET algorithm largely known and it awards for playing quality teams away (it looks to have similar implicit assumptions as RPI with different math), teams which schedule weak OOC SoS get what they deserve if they are bubble teams come selection time.

View attachment 15944
Thanks to NavyNuke for posting this info. From his pasted memo of explanation of the NET:

"The exact algorithm is unknown."

This speaks volumes. Why the opacity? Two obvious reasons are that, one, they do not want the algorithm scrutinized. That likely means that it is flawed to begin with. Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol. I think it is pretty cirtain that the selection committee can disregard the algorithm if they so choose. If that is the case then why have the algorithm to begin with? They have it as an excuse. They can state that they use the algorithm to suggest that they are scientific and unbiased. "We chose team 'X' over team "Y" because that is what the algorithm said to do." But then, if they disregard the model in a selection or seeding, they don't have to say anything because use of the algorithm is implied.

This is not scientific. Science does not hide procedures and methods from thorough review.

The NET algorithm employs (in some fashion) two nebulous "parts," the "Team value Index "TVI," and the "Adjusted Net efficiency rating." Each of these tools is not defined. Each takes into account the "quality" or "strength" of the opponent but never is it defined how quality or strength is derived. Again, there is absolutely no statement on how the initial power ratings are input into the system and how those guesses impact the entire process throughout the season.

Again, if the entire system depends on accurate initial inputs of power ratings, and there is no valid way to assess initial power ratings before games are played, then the entire system is an artifice. If the initial inputs are subjective crap, putting the crap input through a fancy computer model does not mean you get anything out of the other end that is not subjective crap.

This is at the core of scientism. Take a subjective, agendized hypothesis, piggy-back that info into some hazy mathematics that few understand or have the inclination to proof read to launder the subjective information to make it appear quantitative. But it isn't quantitative. It is just made to LOOK quantitative.
 

slugboy

Moderator
Staff member
Messages
11,471
Thanks to NavyNuke for posting this info. From his pasted memo of explanation of the NET:

"The exact algorithm is unknown."

This speaks volumes. Why the opacity? Two obvious reasons are that, one, they do not want the algorithm scrutinized. That likely means that it is flawed to begin with. Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol. I think it is pretty cirtain that the selection committee can disregard the algorithm if they so choose. If that is the case then why have the algorithm to begin with? They have it as an excuse. They can state that they use the algorithm to suggest that they are scientific and unbiased. "We chose team 'X' over team "Y" because that is what the algorithm said to do." But then, if they disregard the model in a selection or seeding, they don't have to say anything because use of the algorithm is implied.

This is not scientific. Science does not hide procedures and methods from thorough review.

The NET algorithm employs (in some fashion) two nebulous "parts," the "Team value Index "TVI," and the "Adjusted Net efficiency rating." Each of these tools is not defined. Each takes into account the "quality" or "strength" of the opponent but never is it defined how quality or strength is derived. Again, there is absolutely no statement on how the initial power ratings are input into the system and how those guesses impact the entire process throughout the season.

Again, if the entire system depends on accurate initial inputs of power ratings, and there is no valid way to assess initial power ratings before games are played, then the entire system is an artifice. If the initial inputs are subjective crap, putting the crap input through a fancy computer model does not mean you get anything out of the other end that is not subjective crap.

This is at the core of scientism. Take a subjective, agendized hypothesis, piggy-back that info into some hazy mathematics that few understand or have the inclination to proof read to launder the subjective information to make it appear quantitative. But it isn't quantitative. It is just made to LOOK quantitative.

Here’s a link to a Medium article (that is probably the article that GTNavyNuke is using)

The article has some complaints about NET, and they’re reasonable

You said this
This speaks volumes. Why the opacity? Two obvious reasons are that, one, they do not want the algorithm scrutinized. That likely means that it is flawed to begin with. Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol.


Why the opacity?
We don’t know. Lots of people keep formulas and algorithms secret. Lots of companies believe they have trade secrets. Keeping something secret doesn’t make it flawed.

Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol.
It’s said that the committee looks at NET, but they obviously make the decisions they want to. Blaming NET for Texas A&M being in the tournament makes no sense. If NET was the main criteria, then Pitt and Wake would have been in the tournament instead. South Carolina and UVA don’t get in with NET. Maybe it helped with seeding, but it’s clear that the committee moves teams around. If they used NET for seeding, UNC isn’t a 1 seed. The committee didn’t use NET all that much
 
Last edited:

ESPNjacket

Helluva Engineer
Messages
1,531
Thanks to NavyNuke for posting this info. From his pasted memo of explanation of the NET:

"The exact algorithm is unknown."

This speaks volumes. Why the opacity? Two obvious reasons are that, one, they do not want the algorithm scrutinized. That likely means that it is flawed to begin with. Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol. I think it is pretty cirtain that the selection committee can disregard the algorithm if they so choose. If that is the case then why have the algorithm to begin with? They have it as an excuse. They can state that they use the algorithm to suggest that they are scientific and unbiased. "We chose team 'X' over team "Y" because that is what the algorithm said to do." But then, if they disregard the model in a selection or seeding, they don't have to say anything because use of the algorithm is implied.

This is not scientific. Science does not hide procedures and methods from thorough review.

The NET algorithm employs (in some fashion) two nebulous "parts," the "Team value Index "TVI," and the "Adjusted Net efficiency rating." Each of these tools is not defined. Each takes into account the "quality" or "strength" of the opponent but never is it defined how quality or strength is derived. Again, there is absolutely no statement on how the initial power ratings are input into the system and how those guesses impact the entire process throughout the season.

Again, if the entire system depends on accurate initial inputs of power ratings, and there is no valid way to assess initial power ratings before games are played, then the entire system is an artifice. If the initial inputs are subjective crap, putting the crap input through a fancy computer model does not mean you get anything out of the other end that is not subjective crap.

This is at the core of scientism. Take a subjective, agendized hypothesis, piggy-back that info into some hazy mathematics that few understand or have the inclination to proof read to launder the subjective information to make it appear quantitative. But it isn't quantitative. It is just made to LOOK quantitative.
Efficiency ratings are close to universally understood. They are points per possession on offense and defense. They are adjusted for strength of schedule, the exact metric for that isn't going to vary much.

So then they take the efficiency rating, which already is adjusted for SOS, and adjust it again. TVI is a strength of schedule metric.

Then they look at Quad wins, which are lines in the sand of the NET rating, and yet again OOC SOS. They are really obsessed with OOC SOS. That's why Pitt with a NET of 40th is left out. I cannot explain South Carolina, with a terrible OOC SOS and a NET that shouldn't get them into the tournament ends up a 6 seed.

The metric is much better than the rest of their reasoning.
 

stinger78

Helluva Engineer
Messages
4,113
Here’s a link to a Medium article (that is probably the article that GTNavyNuke is using)

The article has some complaints about NET, and they’re reasonable

You said this
This speaks volumes. Why the opacity? Two obvious reasons are that, one, they do not want the algorithm scrutinized. That likely means that it is flawed to begin with. Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol.


Why the opacity?
We don’t know. Lots of people keep formulas and algorithms secret. Lots of companies believe they have trade secrets. Keeping something secret doesn’t make it flawed.

Two, The failure to open the algorithm for review means that they do not want to be held accountable for following the protocol.
It’s said that the committee looks at NET, but they obviously make the decisions they want to. Blaming NET for Texas A&M being in the tournament makes no sense. If NET was the main criteria, then Pitt and Wake would have been in the tournament instead. South Carolina and UVA don’t get in with NET. Maybe it helped with seeding, but it’s clear that the committee moves teams around. If they used NET for seeding, UNC isn’t a 1 seed. The committee didn’t use NET all that much
Thank you. Yes. So if decisions are independent, why the bias? Why not more equitably distribute the bids to the next-level teams? Particularly to a conference that has consistently outperformed the others. What’s up with that?
 
Top