|
Post by surfh2ovolleyjock on Nov 27, 2015 15:58:21 GMT -5
Creighton beats Marquette 3-2.(-19,-16,21-25,-20-25,-10)
|
|
|
Post by The Bofa on the Sofa on Nov 27, 2015 16:00:06 GMT -5
If you want to put more weight on late season matches, you're severely crippling the chances of teams outside of the Big 5 conferences to put together a good resume. Conference standings weight all matches the same, whether they are played in September or on Thanksgiving; I see no reason why a full regular season resume should be any different. The argument would be because who you were yesterday correlates more strongly with who you are today as opposed to who you were 2 months ago correlating with who you are today. Have you actually investigated this? I have. A lot. If you look at teams who play early in the season and then play again, there is very little difference in the outcome of the second match if it is played 1 week after the first vs 2 months later. There is a couple percent dropoff, but that's about it. So in the end, yeah, it's ok to weight later season matches slightly more than early season matches, but only very slightly, and far less than most people are wont to do. And you most certainly should not discount early matches completely. They need to be considered. On the whole, it's far better to just treat them equally than to try to weight them, because most people are going to weight them wrong. Alternatively, we could ask the question, if teams change a lot, then why do Pablo rankings from the third week of the season predict the rest of the season matches nearly as well as it predicts the upcoming week? This is true. Beginning of October Pablo rankings might predict 80.5% of the matches in the next week. And for the rest of the season, the prediction rate drops to maybe 80.2%. If teams changed a lot (relative to others) during the season, you wouldn't see this. However, that result is robust, and it is persistent throughout the season. Then again, it's not really surprising. Shoot, teams don't change relatively much over YEARS much less within a season. You can take Pablo rankings from 3 years ago and get a non-embarrassing result for predicting matches this year. On the whole, team changes are typically very, very slow. Arizona St is a good counter-example, but that is a situation with a known cause. And even there, you can't say for sure they would have maintained the same level. Variation happens.
|
|
|
Post by volleyfan24 on Nov 27, 2015 16:07:29 GMT -5
The argument would be because who you were yesterday correlates more strongly with who you are today as opposed to who you were 2 months ago correlating with who you are today. Have you actually investigated this? I have. A lot. If you look at teams who play early in the season and then play again, there is very little difference in the outcome of the second match if it is played 1 week after the first vs 2 months later. There is a couple percent dropoff, but that's about it. So in the end, yeah, it's ok to weight later season matches slightly more than early season matches, but only very slightly, and far less than most people are wont to do. And you most certainly should not discount early matches completely. They need to be considered. On the whole, it's far better to just treat them equally than to try to weight them, because most people are going to weight them wrong. Alternatively, we could ask the question, if teams change a lot, then why do Pablo rankings from the third week of the season predict the rest of the season matches nearly as well as it predicts the upcoming week? This is true. Beginning of October Pablo rankings might predict 80.5% of the matches in the next week. And for the rest of the season, the prediction rate drops to maybe 80.2%. If teams changed a lot (relative to others) during the season, you wouldn't see this. However, that result is robust, and it is persistent throughout the season. Then again, it's not really surprising. Shoot, teams don't change relatively much over YEARS much less within a season. You can take Pablo rankings from 3 years ago and get a non-embarrassing result for predicting matches this year. On the whole, team changes are typically very, very slow. Arizona St is a good counter-example, but that is a situation with a known cause. And even there, you can't say for sure they would have maintained the same level. Variation happens. The voice of reason has spoken.
|
|
|
Post by gators12 on Nov 27, 2015 16:08:35 GMT -5
So will ASU with a loss today would finish out the season 4-12 after the Gardner injury how will they look at this?
|
|
|
Post by n00b on Nov 27, 2015 16:12:11 GMT -5
So will ASU with a loss today would finish out the season 4-12 after the Gardner injury how will they look at this? Nobody is quite sure. There is very little precedence for a resume like theirs. One of the criteria the committee is allowed to consider is availability of athletes for the tournament, so it possible that they punish ASU for their poor results without Gardner. Personally, I think their overall RPI is too good for it to come to that. If their RPI would've fallen to the 40s, they would've been looked at more closely.
|
|
|
Post by volleyfan24 on Nov 27, 2015 16:14:46 GMT -5
How does the Syracuse win affect things?
|
|
|
Post by The Bofa on the Sofa on Nov 27, 2015 16:17:58 GMT -5
The argument would be because who you were yesterday correlates more strongly with who you are today as opposed to who you were 2 months ago correlating with who you are today. It's a Markov model for prediction about what will happen tomorrow. Teams suffer injuries, make positional switches, strategic and tactical switches, etc. as a season progresses. Some teams get better and some get worse. E.g. Nebraska's win against Oregon meant more when they played than it does now presumably because Nebraska has improved while Oregon has regressed (or maybe they weren't that good to begin with and enjoyed an inflated reputation based on previous years). I agree that's how you predict who will win a match. Should we give the SEC title to Florida because Pablo says they're the best team? In my opinion, tournament selection should be based on what you've accomplished and not necessarily on who would win if they played tomorrow. Point of Order: Pablo is 100% based on what teams have "accomplished." The only question is "which accomplishments are you taking into account"? The "SEC Title" is determined by a set of criteria. It starts with wins and then there is a tie-breaking process of some sort. The team that satisfies those criteria is, by definition, the champion. Nothing more, nothing less. Other conferences have different criteria. For example, in the MVC you can have the 6th best record, but if you win the end of the season tournament, you have won the conference title. So in that conference, you don't have to have the best record to win the championship. Now, no conference would actually award a title based on Pablo, but largely due to lack of transparency. But you know, given the unbalanced schedules in large conferences, it wouldn't be the worst idea. John Cook has pointed out the problem in the B1G, and, although I disagree with him that there is no "real champion" (the B1G has decreed that the team with the best record is the "champion", unbalanced schedules be damned), the point that in such situations there is no inherently best answer is legit. In the end, it comes down to the fact that "champion" is a matter of definition. You win the NCAA tournament, you are the national champion. No more, no less. Adding on baggage like "best team" and such completely begs the question of what that all means, and there is no single correct answer. However, it is indisputable that the team is the champion, because it is, by NCAA definition, true. If the NCAA (or the SEC) used a different approach to determine the champion, then the team that satisfied that criteria would be the champion just the same.
|
|
|
Post by n00b on Nov 27, 2015 16:24:34 GMT -5
I agree that's how you predict who will win a match. Should we give the SEC title to Florida because Pablo says they're the best team? In my opinion, tournament selection should be based on what you've accomplished and not necessarily on who would win if they played tomorrow. Point of Order: Pablo is 100% based on what teams have "accomplished." The only question is "which accomplishments are you taking into account"? The "SEC Title" is determined by a set of criteria. It starts with wins and then there is a tie-breaking process of some sort. The team that satisfies those criteria is, by definition, the champion. Nothing more, nothing less. Other conferences have different criteria. For example, in the MVC you can have the 6th best record, but if you win the end of the season tournament, you have won the conference title. So in that conference, you don't have to have the best record to win the championship. Now, no conference would actually award a title based on Pablo, but largely due to lack of transparency. But you know, given the unbalanced schedules in large conferences, it wouldn't be the worst idea. John Cook has pointed out the problem in the B1G, and, although I disagree with him that there is no "real champion" (the B1G has decreed that the team with the best record is the "champion", unbalanced schedules be damned), the point that in such situations there is no inherently best answer is legit. In the end, it comes down to the fact that "champion" is a matter of definition. You win the NCAA tournament, you are the national champion. No more, no less. Adding on baggage like "best team" and such completely begs the question of what that all means, and there is no single correct answer. However, it is indisputable that the team is the champion, because it is, by NCAA definition, true. If the NCAA (or the SEC) used a different approach to determine the champion, then the team that satisfied that criteria would be the champion just the same. Correct. I should've specified that I believe tournament selection should be based on what a team has accomplished in terms of match W-L. And that is essentially what the NCAA has listed as their criteria, so it's one of the things I think they have right. I know I've posted this in the past, but if two teams have identical resumes going into the last match of the season and Team A beats Team B but gets outscored, I strongly believe that Team A should go to the tournament over Team B (even though if they played again I'd bet on Team B).
|
|
|
Post by n00b on Nov 27, 2015 16:26:23 GMT -5
TCU completes the sweep of West Virginia. Their RPI puts them right on the bubble line, but with a win over a top 3 Texas team, I have a hard time seeing the committee leaving them out.
|
|
|
Post by spikerthemovie on Nov 27, 2015 16:42:44 GMT -5
So will ASU with a loss today would finish out the season 4-12 after the Gardner injury how will they look at this? Nobody is quite sure. There is very little precedence for a resume like theirs. One of the criteria the committee is allowed to consider is availability of athletes for the tournament, so it possible that they punish ASU for their poor results without Gardner. Personally, I think their overall RPI is too good for it to come to that. If their RPI would've fallen to the 40s, they would've been looked at more closely. The thing that most intrigues me, having not seen ASU play but having read a ton about them here, is that posters keep saying that ASU is a different team without Gardner and it seems as if the committee could take that approach: That team that was led by Gardner would have easily made the tourney; the team she's not on won't.
|
|
|
Post by The Bofa on the Sofa on Nov 27, 2015 16:42:46 GMT -5
Point of Order: Pablo is 100% based on what teams have "accomplished." The only question is "which accomplishments are you taking into account"? The "SEC Title" is determined by a set of criteria. It starts with wins and then there is a tie-breaking process of some sort. The team that satisfies those criteria is, by definition, the champion. Nothing more, nothing less. Other conferences have different criteria. For example, in the MVC you can have the 6th best record, but if you win the end of the season tournament, you have won the conference title. So in that conference, you don't have to have the best record to win the championship. Now, no conference would actually award a title based on Pablo, but largely due to lack of transparency. But you know, given the unbalanced schedules in large conferences, it wouldn't be the worst idea. John Cook has pointed out the problem in the B1G, and, although I disagree with him that there is no "real champion" (the B1G has decreed that the team with the best record is the "champion", unbalanced schedules be damned), the point that in such situations there is no inherently best answer is legit. In the end, it comes down to the fact that "champion" is a matter of definition. You win the NCAA tournament, you are the national champion. No more, no less. Adding on baggage like "best team" and such completely begs the question of what that all means, and there is no single correct answer. However, it is indisputable that the team is the champion, because it is, by NCAA definition, true. If the NCAA (or the SEC) used a different approach to determine the champion, then the team that satisfied that criteria would be the champion just the same. Correct. I should've specified that I believe tournament selection should be based on what a team has accomplished in terms of match W-L. And that is essentially what the NCAA has listed as their criteria, so it's one of the things I think they have right. I know I've posted this in the past, but if two teams have identical resumes going into the last match of the season and Team A beats Team B but gets outscored, I strongly believe that Team A should go to the tournament over Team B (even though if they played again I'd bet on Team B). You may believe that, but in the end, it is not incorrect to do it the other way. And others will contend it should go the other way. If you want the tournament to have the best teams, then choose the criteria that reflect that. If you don't want to do that, then whatever, but it is also a concession that you don't necessarily want the best teams. The committee uses head-to-head outcomes as a criteria. In that case, it would reflect your preference. Then again, the committee has shown that head-to-head is not big enough to overcome differences in other things. In the end, I think you have built a pretty big strawman. Where are these teams where everything is perfectly equal except for head-to-head, with only a single match between them? Ultimately, these discussions about your preferred selection criteria are about as interesting as talking about your favorite pizza toppings. If nothing else, they belong in a different thread
|
|
|
Post by rtael on Nov 27, 2015 17:08:29 GMT -5
Correct. I should've specified that I believe tournament selection should be based on what a team has accomplished in terms of match W-L. And that is essentially what the NCAA has listed as their criteria, so it's one of the things I think they have right. I know I've posted this in the past, but if two teams have identical resumes going into the last match of the season and Team A beats Team B but gets outscored, I strongly believe that Team A should go to the tournament over Team B (even though if they played again I'd bet on Team B). You may believe that, but in the end, it is not incorrect to do it the other way. And others will contend it should go the other way. If you want the tournament to have the best teams, then choose the criteria that reflect that. If you don't want to do that, then whatever, but it is also a concession that you don't necessarily want the best teams. The committee uses head-to-head outcomes as a criteria. In that case, it would reflect your preference. Then again, the committee has shown that head-to-head is not big enough to overcome differences in other things. In the end, I think you have built a pretty big strawman. Where are these teams where everything is perfectly equal except for head-to-head, with only a single match between them? Ultimately, these discussions about your preferred selection criteria are about as interesting as talking about your favorite pizza toppings. If nothing else, they belong in a different thread Pepperoni.
|
|
|
Post by jayj79 on Nov 27, 2015 17:17:43 GMT -5
oops, I thought this was the "Bubbles Watch" thread PS, I like pineapple pizza, I know that's somewhat controversial
|
|
|
Post by n00b on Nov 27, 2015 17:18:02 GMT -5
Pitt leads Virginia 2-0 and 10-8
Villanova leads Xavier 1-0 and 20-16.
|
|
|
Post by n00b on Nov 27, 2015 17:42:41 GMT -5
Pitt def. Virginia 3-0
Villanova leads Xavier 2-0 and 12-7.
Arizona leads Arizona State 1-0.
|
|