|
|
Post by boxcariii on Nov 29, 2022 11:53:14 GMT -5
The RPI is bad, but every system is going to have flaws. It's been argued to death so I'll only add this when MBB "switched" to the NET. Many of the analytics are predictive. Creighton's AD at the time (Bruce Rasmussen) was on the MBB committee. He said one thing that stuck with me when it came to analytics. It was very important to him that results still mattered. As long as results still matter and it's not purely predictive, I don't really care what the system is. Additionally, I think it's very important to remember there's 300+ teams and they don't all play each other (off season what if idea?). I agree with this, and I like the way NET has been implemented. My understanding is that the men's basketball committee doesn't take your own NET into consideration. The purpose of the NET is to figure out how strong your opponents were and how much credit or blame you should take for wins and losses. In other words, it's used as a way to rank each team's resume. I think that women's volleyball would greatly benefit from something similar. I think the "switch" is greatly overestimated as well. RPI still exists and is still a tool in basketball. NET is just another tool. How people perceive things will always be different. In a sport where not everyone plays each other, there will always be controversy. |
|
bluepenquin
Hall of Fame
4-Time VolleyTalk Poster of the Year (2019, 2018, 2017, 2016), All-VolleyTalk 1st Team (2021, 2020, 2019, 2018, 2017, 2016) All-VolleyTalk 2nd Team 2023
Posts: 13,262 
|
Post by bluepenquin on Nov 29, 2022 12:03:49 GMT -5
I agree with this, and I like the way NET has been implemented. My understanding is that the men's basketball committee doesn't take your own NET into consideration. The purpose of the NET is to figure out how strong your opponents were and how much credit or blame you should take for wins and losses. In other words, it's used as a way to rank each team's resume. I think that women's volleyball would greatly benefit from something similar. I think the "switch" is greatly overestimated as well. RPI still exists and is still a tool in basketball. NET is just another tool. How people perceive things will always be different. In a sport where not everyone plays each other, there will always be controversy. Back when I followed MBB closely and tried to predict the tournament field and seeding - before ESPN was doing this kind of stuff and long before NET - the committee's reliance on RPI wasn't nearly what it is for WVB. |
|
|
|
Post by bbg95 on Nov 29, 2022 12:07:32 GMT -5
I agree with this, and I like the way NET has been implemented. My understanding is that the men's basketball committee doesn't take your own NET into consideration. The purpose of the NET is to figure out how strong your opponents were and how much credit or blame you should take for wins and losses. In other words, it's used as a way to rank each team's resume. I think that women's volleyball would greatly benefit from something similar. I think the "switch" is greatly overestimated as well. RPI still exists and is still a tool in basketball. NET is just another tool. How people perceive things will always be different. In a sport where not everyone plays each other, there will always be controversy. RPI still exists for men's basketball, but I don't think the committee actually uses it much anymore. |
|
|
|
Post by mikegarrison on Nov 29, 2022 12:35:47 GMT -5
I'm going to take the a mixed position here.
I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament.
But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is.
|
|
|
|
Post by bbg95 on Nov 29, 2022 12:38:13 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is.That's never been how committees have treated at-large bids or at least not as long as I've been following sports. If we look at football, it's pretty likely that Alabama is one of the four teams most likely to win the national championship if they were selected for the CFP. But with two losses, they are probably out. |
|
|
|
Post by mikegarrison on Nov 29, 2022 12:42:20 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is.That's never been how committees have treated at-large bids or at least not as long as I've been following sports. If we look at football, it's pretty likely that Alabama is one of the four teams most likely to win the national championship if they were selected for the CFP. But with two losses, they are probably out. I don't like the committee and I don't like many of their decisions. I'm arguing what I think the purpose should be, not trying to interpret what the committee thinks. I recognize that I am not the king and don't get to decide this on my own; that's not what I am saying. But I do think the NCAA would be served by having this discussion in public. And if not the NCAA, then at least I can advocate for my position here on Volleytalk, just like anyone else can advocate for their positions. |
|
|
|
Post by bbg95 on Nov 29, 2022 12:46:18 GMT -5
That's never been how committees have treated at-large bids or at least not as long as I've been following sports. If we look at football, it's pretty likely that Alabama is one of the four teams most likely to win the national championship if they were selected for the CFP. But with two losses, they are probably out. I don't like the committee and I don't like many of their decisions. I'm arguing what I think the purpose should be, not trying to interpret what the committee thinks. I recognize that I am not the king and don't get to decide this on my own; that's not what I am saying. But I do think the NCAA would be served by having this discussion in public. And if not the NCAA, then at least I can advocate for my position here on Volleytalk, just like anyone else can advocate for their positions. Well, you said you think the at-larges are in the tournament because they're the most likely to win the title. If that's the way you think it should be, then that's fine. I don't really agree, but it's a reasonable opinion. My only point was that the committees have never really viewed it that way. |
|
|
|
Post by donut on Nov 29, 2022 13:18:22 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is. I don't know how you rationalize this with the desire for more objectivity. If the committee adopts this view, holy subjectivity batman.
|
|
|
|
Post by stevehorn on Nov 29, 2022 13:19:47 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is. I believe the AQs are a continuation of how NCAA tournaments started (or at least how teams were selected in my youth). At that time, only conference champions got a NCAA bid. IIRC the prevailing thought at the time was that only the conference champion "earned" the right to play for a national championship. The at-larges started because it was recognized that other teams in some conferences were better than many of the conference champions and it was "unfair" that they couldn't play for the national championship. The first step was to allow the second place teams in a conference to compete for the initial at-large spots. Ironically in the first year of this in the MBB tournament, Indiana and Michigan (2nd place Big 10) played for the championship. As the size of the tournament field increased, they removed the limits on the number of teams per conference. I think the definition of at-larges is the 32 at-larges are the best 32 out of that at-large pool. Realistically, most people would say that's also the next 32 that most likely could win the championship. |
|
|
|
Post by bbg95 on Nov 29, 2022 13:25:40 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is. I believe the AQs are a continuation of how NCAA tournaments started (or at least how teams were selected in my youth). At that time, only conference champions got a NCAA bid. IIRC the prevailing thought at the time was that only the conference champion "earned" the right to play for a national championship. The at-larges started because it was recognized that other teams in some conferences were better than many of the conference champions and it was "unfair" that they couldn't play for the national championship. The first step was to allow the second place teams in a conference to compete for the initial at-large spots. Ironically in the first year of this in the MBB tournament, Indiana and Michigan (2nd place Big 10) played for the championship. As the size of the tournament field increased, they removed the limits on the number of teams per conference. I think the definition of at-larges is the 32 at-larges are the best 32 out of that at-large pool. Realistically, most people would say that's also the next 32 that most likely could win the championship. The history is interesting. I think most people probably see "best" and "most deserving" as somewhat interchangeable. I believe the committees usually go with most deserving, but they can be swayed by brand bias wherein a blue blood gets more benefit of the doubt than a lesser name. |
|
|
|
Post by mikegarrison on Nov 29, 2022 13:33:55 GMT -5
I'm going to take the a mixed position here. I think that the AQs are in the tournament as a reward for the season they had. This is entirely appropriate -- you win your conference season or tournament and you win the right to get into the national championship tournament. But I think the at-larges are in the tournament for a different reason. I think they are supposed to be in because they are the 32 other teams most likely to be able to win the tournament. I think this improves the competitiveness of the tournament and really makes it the exciting event that it is. I don't know how you rationalize this with the desire for more objectivity. If the committee adopts this view, holy subjectivity batman. It's quite simple. I would use Pablo, or Massey, or an average of the two, or a replacement for them that was tested to work at least as well. (It's not like the season results don't matter. These systems are entirely based on the season results.) However, let me add, I would want any such system that was used to be entirely transparent. That means not just the algorithm but also the code. |
|
|
|
Post by bbg95 on Nov 29, 2022 13:43:59 GMT -5
I don't know how you rationalize this with the desire for more objectivity. If the committee adopts this view, holy subjectivity batman. It's quite simple. I would use Pablo, or Massey, or an average of the two, or a replacement for them that was tested to work at least as well. (It's not like the season results don't matter. These systems are entirely based on the season results.)Not exactly. At least for Pablo (I'm less familiar with Massey), it's based on points, right? Not on who actually wins the matches? I think the NCAA has the right idea with NET where your own NET ranking is far more important to your opponent than it is to you and vice versa. |
|
|
|
Post by mikegarrison on Nov 29, 2022 14:00:04 GMT -5
It's quite simple. I would use Pablo, or Massey, or an average of the two, or a replacement for them that was tested to work at least as well. (It's not like the season results don't matter. These systems are entirely based on the season results.)Not exactly. At least for Pablo (I'm less familiar with Massey), it's based on points, right? Not on who actually wins the matches? I think the NCAA has the right idea with NET where your own NET ranking is far more important to your opponent than it is to you and vice versa. Points are results. I believe Pablo is no longer based on the "pure points" methodology, but that's not clear. One weakness of Pablo is that The Bofa on the Sofa keeps his algorithm proprietary. He describes it, but he doesn't give us the exact details. This is his right (it is his IP), but it means that it is easy for people to mythologize what it actually is doing. What NET is doing is a step in the right direction. (I will note that the NCAA didn't develop NET -- they hired Google to do it.) At least NET is judging wins by the quality of the opponent, while RPI only judges *schedules* by the quality of the opponents -- a huge flaw. Wins are made up of points, so there is a good reason why points might be more predictive than wins. There are only about 30 wins available per season as raw data. There are about 120ish sets available each season as raw data. But there are more like 5000 points available per season as raw data. Size matters. The potential weakness of a pure points model is that not all points matter the same. Teams in a 20-10 set won't be playing with quite the same intensity as teams in a 20-20 set. But still, given the nature of volleyball where you can't just stall and run out the clock, I think even blowout sets are not the same thing as "garbage time" in basketball. |
|
|
|
Post by bbg95 on Nov 29, 2022 14:03:25 GMT -5
Not exactly. At least for Pablo (I'm less familiar with Massey), it's based on points, right? Not on who actually wins the matches? I think the NCAA has the right idea with NET where your own NET ranking is far more important to your opponent than it is to you and vice versa. Points are results. I believe Pablo is no longer based on the "pure points" methodology, but that's not clear. One weakness of Pablo is that The Bofa on the Sofa keeps his algorithm proprietary. He describes it, but he doesn't give us the exact details. This is his right (it is his IP), but it means that it is easy for people to mythologize what it actually is doing. What NET is doing is a step in the right direction. (I will note that the NCAA didn't develop NET -- they hired Google to do it.) At least NET is judging wins by the quality of the opponent, while RPI only judges *schedules* by the quality of the opponents -- a huge flaw. Wins are made up of points, so there is a good reason why points might be more predictive than wins. There are only about 30 wins available per season as raw data. There are about 120ish sets available each season as raw data. But there are more like 5000 points available per season as raw data. Size matters. The potential weakness of a pure points model is that not all points matter the same. Teams in a 20-10 set won't be playing with quite the same intensity as teams in a 20-20 set. But still, given the nature of volleyball where you can't just stall and run out the clock, I think even blowout sets are not the same thing as "garbage time" in basketball. Points are results, but they're not the results that matter. I'd be fine with using Pablo the way NET is used, but not the way RPI is currently used. I think Pablo has a lot of value for some purposes like predicting who you think will win. But I don't think it should be used as a de facto ranking system. |
|
|
|
Post by mikegarrison on Nov 29, 2022 14:15:29 GMT -5
Points are results. I believe Pablo is no longer based on the "pure points" methodology, but that's not clear. One weakness of Pablo is that The Bofa on the Sofa keeps his algorithm proprietary. He describes it, but he doesn't give us the exact details. This is his right (it is his IP), but it means that it is easy for people to mythologize what it actually is doing. What NET is doing is a step in the right direction. (I will note that the NCAA didn't develop NET -- they hired Google to do it.) At least NET is judging wins by the quality of the opponent, while RPI only judges *schedules* by the quality of the opponents -- a huge flaw. Wins are made up of points, so there is a good reason why points might be more predictive than wins. There are only about 30 wins available per season as raw data. There are about 120ish sets available each season as raw data. But there are more like 5000 points available per season as raw data. Size matters. The potential weakness of a pure points model is that not all points matter the same. Teams in a 20-10 set won't be playing with quite the same intensity as teams in a 20-20 set. But still, given the nature of volleyball where you can't just stall and run out the clock, I think even blowout sets are not the same thing as "garbage time" in basketball. Points are results, but they're not the results that matter. I'd be fine with using Pablo the way NET is used, but not the way RPI is currently used. I think Pablo has a lot of value for some purposes like predicting who you think will win. But I don't think it should be used as a de facto ranking system. Well, again, this comes down to basic philosophy. I think a de facto ranking system *should* be based on "who is more likely to win". But I understand that you disagree. As an engineer, I am more focused on a "root cause" worldview, and points won are definitely the root cause of matches won. There are even lower-level root causes available (serving efficiency, etc.) that are root causes for points won, but you also have to balance how easy they are to measure. Points won are very easy to measure and have a completely objective definition. |
|
