(05-02-2012 12:27 PM)Wedge Wrote:  That's the real problem with the polls, isn't it? Whenever you start a coaches' poll or a Harris Poll, you have a lot of voters who start their ballot by looking at the AP poll, which has been running for several weeks, and using the AP poll as the basis for their ballot, probably tweaking a few teams up or down based on their personal views or biases. Then in the following weeks, they use last week's polls and/or their own previous ballot as their starting point.

What is really needed for any polls used to pick teams in a playoff -- but probably can't be found -- is a large group of voters who are knowledgeable, and watch as many games as compulsive gamblers do, and are independent thinkers capable of starting their own ballot from scratch based on the football games they've watched.

You just described a selection committee.

(05-02-2012 12:27 PM)Wedge Wrote:  That's the real problem with the polls, isn't it? Whenever you start a coaches' poll or a Harris Poll, you have a lot of voters who start their ballot by looking at the AP poll, which has been running for several weeks, and using the AP poll as the basis for their ballot, probably tweaking a few teams up or down based on their personal views or biases. Then in the following weeks, they use last week's polls and/or their own previous ballot as their starting point.

What is really needed for any polls used to pick teams in a playoff -- but probably can't be found -- is a large group of voters who are knowledgeable, and watch as many games as compulsive gamblers do, and are independent thinkers capable of starting their own ballot from scratch based on the football games they've watched.

Wedge - You've found the answer to our dilemma. The only unfortunate thing is that the answer can never be used.

The accurateness of a group of predictions skyrockets when people have their own money on the line. This applies to markets where people can bet things ranging from presidential elections to the Oscars. These markets have shown to be much more accurate than opinion polls because when you have your own money on the line, it naturally makes you more objective and you go with your brain instead of your emotion. Your brain goes from thinking about what you *want* to happen to focusing on what you truly believe *will* happen and that starts providing more clarity.

In this case, there is absolutely no greater forecast of who the best 4 teams are in the country than the lines in Vegas. This group, as a whole, has only one single purpose: predict who is going to WIN in head-to-head matchups. They don't care about conferences. They don't care about strength of schedules. They don't care if it's fair whether a non-AQ team is included or not. This group is more financially motivated to pick the 4 best teams in an objective fashion than any other group of people out there.

If the university presidents weren't so spooked about connections to gambling (because, by golly, everyone watches the NCAA Tournament for pure amateurism as opposed to all of the money that they've wagered in office pools), this would be the most objective way to determine the top 4 teams from a group of subjective humans. You're not getting any BS, politics or emotion from that group as a whole.

(05-02-2012 11:47 AM)Frank the Tank Wrote:  Here's why I don't think a selection committee will fly: people and teams WANT to follow rankings each week. Teams want to know where they stand and who needs to win or lose to move up or down at an given time. This isn't like the NCAA Tournament where you know every reasonable contender for the national championship is going to get an at-large bid even if it doesn't win its conference tournament. People also don't want to see a selection committee come back with 4 completely different teams than what they had assumed to be the top 4 teams up to that point. If the goal is transparency, then it needs to be a ranking system that people see and can follow from week to week (and not be surprised out of the blue in December).

Honestly, I don't envy the powers that be on the ranking issue because I think it's damned if they do and damned if they don't. When the rankings use computers and strength of schedule components more heavily, you get results like the 2001 Nebraska team getting ranked #1 despite having lost its last game by 26 points and not winning its division. ALL of the things that we hear from people complaining about how Alabama ended up at #2 and wanting "objective" measures belie the fact that such objective measures actually produced results that were even more infuriating to people in the past. The public has a short memory, so they've largely forgotten years like 2001 and 2003 when the complaints were all about how strength of schedule and the computer rankings were used TOO much.

At the same time, to be fair to the current BCS rankings, they really haven't been that crazy since the formula was overhauled in 2005. People might quibble over a spot or two (which, to be clear, can matter a lot in a top 4 playoff), but it wasn't as if though we saw any teams at the top of the rankings where there was a belief that a team was 5 or 10 spots off of where it should have been. I've been researching past rankings a lot lately with my blog posts and the BCS rankings since 2005 at least made reasonable sense and were justifiable. Once again, the pre-2005 rankings that used more computer formulas and strength of schedule components were the ones that produced the truly wacky results. That's the dilemma.

Basically, the public really wants something that's probably impossible to achieve: they want an objective computer formula that will generate the exact same outcome as a subjective poll. That's really the only thing that will make them happy. People don't like the *thought* of using a subjective poll (because it seems unfair and subject to bias), yet they still want the *outcome* of the subjective poll (because it reflects who they actually believe to be the top teams in reality using their own eyes). My take is that we'll ultimately end up with something that produces the desired output (meaning more reliance on a poll) than focusing on the inputs.

Frank, people en masse are stupid, and it's usually loud mouthed jerks who lead things, by preying upon whatever issue that sways the masses. That's a big reason why I hate popularity contests. The popular vote often sucks, when you look back on it with a historical perspective...

(05-02-2012 12:46 PM)Frank the Tank Wrote:  

(05-02-2012 12:27 PM)Wedge Wrote:  That's the real problem with the polls, isn't it? Whenever you start a coaches' poll or a Harris Poll, you have a lot of voters who start their ballot by looking at the AP poll, which has been running for several weeks, and using the AP poll as the basis for their ballot, probably tweaking a few teams up or down based on their personal views or biases. Then in the following weeks, they use last week's polls and/or their own previous ballot as their starting point.

What is really needed for any polls used to pick teams in a playoff -- but probably can't be found -- is a large group of voters who are knowledgeable, and watch as many games as compulsive gamblers do, and are independent thinkers capable of starting their own ballot from scratch based on the football games they've watched.

Wedge - You've found the answer to our dilemma. The only unfortunate thing is that the answer can never be used.

The accurateness of a group of predictions skyrockets when people have their own money on the line. This applies to markets where people can bet things ranging from presidential elections to the Oscars. These markets have shown to be much more accurate than opinion polls because when you have your own money on the line, it naturally makes you more objective and you go with your brain instead of your emotion. Your brain goes from thinking about what you *want* to happen to focusing on what you truly believe *will* happen and that starts providing more clarity.

In this case, there is absolutely no greater forecast of who the best 4 teams are in the country than the lines in Vegas. This group, as a whole, has only one single purpose: predict who is going to WIN in head-to-head matchups. They don't care about conferences. They don't care about strength of schedules. They don't care if it's fair whether a non-AQ team is included or not. This group is more financially motivated to pick the 4 best teams in an objective fashion than any other group of people out there.

If the university presidents weren't so spooked about connections to gambling (because, by golly, everyone watches the NCAA Tournament for pure amateurism as opposed to all of the money that they've wagered in office pools), this would be the most objective way to determine the top 4 teams from a group of subjective humans. You're not getting any BS, politics or emotion from that group as a whole.

You're right that Vegas doesn't care who wins, but they care about who's going to bet on who wins. The oddsmakers set lines to make profits for their books based on who they think will bet on a team. They don't set the line on 50-50, because that doesn't make money. They set the line on 60-40 and move from there based on how bets are being taken. They might have a feeling that a team will win, but more often than not, they know that they can play off of a team's popularity to get more bets for them if they move the line a certain way.

Great points Dan! Would fully support this plan!

(05-02-2012 11:03 AM)CommuterBob Wrote:  

(05-01-2012 06:25 PM)GiveEmTheAxe Wrote:  Maybe if we ever see starters being rested for Michigan-tOSU or the SECCG, we'll reconsider the worthiness of a champions-only rule.

There's no way that would happen. Not a chance. The risk of getting dropped from the top 4 (or even losing the ability to host a game) or losing a rivalry game would be too great.

I think that it would be rare, but not impossible. 2006 saw a 1 v 2 battle in the Game, right? Ok, those teams hate each other enough to never rest a starter, but what about last year's SECCG? Is there really any doubt that LSU would have remained a top 2 team had it lost to UGA?

(05-02-2012 01:48 PM)GiveEmTheAxe Wrote:  

(05-02-2012 11:03 AM)CommuterBob Wrote:  

(05-01-2012 06:25 PM)GiveEmTheAxe Wrote:  Maybe if we ever see starters being rested for Michigan-tOSU or the SECCG, we'll reconsider the worthiness of a champions-only rule.

There's no way that would happen. Not a chance. The risk of getting dropped from the top 4 (or even losing the ability to host a game) or losing a rivalry game would be too great.

I think that it would be rare, but not impossible. 2006 saw a 1 v 2 battle in the Game, right? Ok, those teams hate each other enough to never rest a starter, but what about last year's SECCG? Is there really any doubt that LSU would have remained a top 2 team had it lost to UGA?

LSU would have probably stayed in the top 3 merely because it was a 1-loss team, but this isn't like the NFL where you can be assured of a spot, so I don't think you'd see teams rest their starters. You're playing with an awful lot of fire there (and to the extent subjective opinion comes into play in a poll, conceding a game NEVER looks good). There's no off week in college football (and that's part of the beauty of it).

(05-02-2012 01:48 PM)GiveEmTheAxe Wrote:  

(05-02-2012 11:03 AM)CommuterBob Wrote:  

(05-01-2012 06:25 PM)GiveEmTheAxe Wrote:  Maybe if we ever see starters being rested for Michigan-tOSU or the SECCG, we'll reconsider the worthiness of a champions-only rule.

There's no way that would happen. Not a chance. The risk of getting dropped from the top 4 (or even losing the ability to host a game) or losing a rivalry game would be too great.

I think that it would be rare, but not impossible. 2006 saw a 1 v 2 battle in the Game, right? Ok, those teams hate each other enough to never rest a starter, but what about last year's SECCG? Is there really any doubt that LSU would have remained a top 2 team had it lost to UGA?

No, but a loss could have put them at #2 and a host of a tougher team in the semis, or even #3 where they could have to go on the road. I'm assuming on-campus semis and to some extent BCS games. The NFL only rests their players when their seeds are set. That wouldn't happen in CFB.

(05-02-2012 12:46 PM)Frank the Tank Wrote:  

(05-02-2012 12:27 PM)Wedge Wrote:  That's the real problem with the polls, isn't it? Whenever you start a coaches' poll or a Harris Poll, you have a lot of voters who start their ballot by looking at the AP poll, which has been running for several weeks, and using the AP poll as the basis for their ballot, probably tweaking a few teams up or down based on their personal views or biases. Then in the following weeks, they use last week's polls and/or their own previous ballot as their starting point.

What is really needed for any polls used to pick teams in a playoff -- but probably can't be found -- is a large group of voters who are knowledgeable, and watch as many games as compulsive gamblers do, and are independent thinkers capable of starting their own ballot from scratch based on the football games they've watched.

In this case, there is absolutely no greater forecast of who the best 4 teams are in the country than the lines in Vegas. This group, as a whole, has only one single purpose: predict who is going to WIN in head-to-head matchups. They don't care about conferences. They don't care about strength of schedules. They don't care if it's fair whether a non-AQ team is included or not. This group is more financially motivated to pick the 4 best teams in an objective fashion than any other group of people out there.

The goal of setting lines is to generate an even amount of money on each team. If both sides of a game have a perfectly even amount of action, a bookmaker or legalized gambling establishment will automatically win, regardless of the result. That is the perfect scenario for setting a line.

(05-02-2012 01:01 PM)CommuterBob Wrote:  

(05-02-2012 12:46 PM)Frank the Tank Wrote:  

(05-02-2012 12:27 PM)Wedge Wrote:  That's the real problem with the polls, isn't it? Whenever you start a coaches' poll or a Harris Poll, you have a lot of voters who start their ballot by looking at the AP poll, which has been running for several weeks, and using the AP poll as the basis for their ballot, probably tweaking a few teams up or down based on their personal views or biases. Then in the following weeks, they use last week's polls and/or their own previous ballot as their starting point.

What is really needed for any polls used to pick teams in a playoff -- but probably can't be found -- is a large group of voters who are knowledgeable, and watch as many games as compulsive gamblers do, and are independent thinkers capable of starting their own ballot from scratch based on the football games they've watched.

Wedge - You've found the answer to our dilemma. The only unfortunate thing is that the answer can never be used.

The accurateness of a group of predictions skyrockets when people have their own money on the line. This applies to markets where people can bet things ranging from presidential elections to the Oscars. These markets have shown to be much more accurate than opinion polls because when you have your own money on the line, it naturally makes you more objective and you go with your brain instead of your emotion. Your brain goes from thinking about what you *want* to happen to focusing on what you truly believe *will* happen and that starts providing more clarity.

In this case, there is absolutely no greater forecast of who the best 4 teams are in the country than the lines in Vegas. This group, as a whole, has only one single purpose: predict who is going to WIN in head-to-head matchups. They don't care about conferences. They don't care about strength of schedules. They don't care if it's fair whether a non-AQ team is included or not. This group is more financially motivated to pick the 4 best teams in an objective fashion than any other group of people out there.

If the university presidents weren't so spooked about connections to gambling (because, by golly, everyone watches the NCAA Tournament for pure amateurism as opposed to all of the money that they've wagered in office pools), this would be the most objective way to determine the top 4 teams from a group of subjective humans. You're not getting any BS, politics or emotion from that group as a whole.

You're right that Vegas doesn't care who wins, but they care about who's going to bet on who wins. The oddsmakers set lines to make profits for their books based on who they think will bet on a team. They don't set the line on 50-50, because that doesn't make money. They set the line on 60-40 and move from there based on how bets are being taken. They might have a feeling that a team will win, but more often than not, they know that they can play off of a team's popularity to get more bets for them if they move the line a certain way.

That's more accurate. I knew someone who said the surest way to make money was to bet against Kentucky in basketball--because the Kentucky basketball fans would be betting with their emotion, and there were a lot of them. The betting magazines have history of how schools have done against the spread. That doesn't have much to do with the team-that has to do with the attitude of the bettors.

Its like an economics professor told us, "Don't ask me about the stock market. That's a topic for a psychology class."