Open Forum

Regular season's over, best and only meaningful regular season in all of major American sport.

Chime away on all the BCS weirdness etc. I'll stick my nose in from time to time, actually on the road right now so it's difficult to get more than a few moments in front of the computer to put something up. 

Comments

[GatorDave]

What's the point of having the computers in the system if the pollsters completely overwhelm them?

[CFR]

I'd rather not have the computers in the system, period.

I'd take out the coaches poll as well.

And then I'd look to sharpen the roll of the Harris poll and then invite the AP back and sharpen its voter rolls as well.

[michael]

Where? O'where is CFR? We need your special brand of humor to enliven us.

Please. Please! Post more on the "defense" on the current post-season format.

We need you, CFR, and your outstanding humor when we are watching 6-6 Who Cares U vs. 6-6 Whatever College in the Mediocre Bowl (please note that two 6-6 teams, which is the exact definition of average, will face each other this year).

Please. Please! CFR tells how wonderful this really is and we can all join in the joke.

[mmortal03]

As usual, I have the opposite view of CFR, and would almost like ONLY the computers to rule the world, if they can be tweaked properly.

Regarding gatorDave's post, the voters don't always overwhelm them, but yeah. basically, if you want the top 2 in the polls to be the top 2 in the end, then the computers lose all influence if you are only selecting the top 2.

I'd rather see the computers take up 50% and the Harris take up 50%. Get rid of the Coaches, since they are horrible at it. The computers are usually MUCH better than the humans at being unbiased.

[MECU]

The problem with polls is that most people think there is a pre-defined "correct" poll that the polling should produce. That the computers should be "tweaked" to produce the "correct" results. But what is correct? Isn't what the poll produces correct? You may disagree with it, but that's your opinion. The formula(s) that are used can't be "wrong," only your interpretation of them makes them wrong.

But in the end, why shouldn't everyone's (sane) opinion count? Why do we only let the coaches, random hand-selected football-connected people (aka Harris Poll) and 6 formula computers (of which 1/3 is disregarded even before adding into the BCS calculation!) get to decide? Why can't everyone decide? Why can't the fans? They aren't anymore biased than the coaches or well-connected & selected people. Checkout the BCS Fan Poll at http://www.bcsfanpoll.com where you can submit your own poll. Bad voters are kept out and all polls are reviewed and constantly evaluated. The goal of the site is to get a fan vote as part of the BCS calculation.

[Dawgy]

CFR is still rehabing. When he completes his course he will be back proclaiming the power of the SEC and pushing for a playoff.

[mmortal03]

MECU, you hit the nail on the head. I haven't heard anyone talking about the computers picking Virginia Tech as #1 and them not being in the championship being wrong. I think that is just as wrong as USC being ranked #1 in both human polls and being left out of the championship. If the computers don't represent what most people think, then are the people just biased, or are the computers "wrong". Yes, the computers are mathematically accurate, but that doesn't mean that their formulas are best designed to pick the teams how the people want them to be picked. That is what I was talking about in terms of "tweaking". Basically, the results show that people don't really care as much about Strength of Schedule as they say they do, because the main reason Virginia Tech was ranked number 1 in the computers was because they had a 9th ranked SOS, versus LSU's 10th and Ohio State's 50th.

Now, we must hope that the computers already compensate for the fact that VT's schedule strength was boosted by playing LSU, but they lost to them, so they shouldn't get as much credit for that game as if they had won it. You would assume that the computers grant teams some points for playing a hard schedule, but more points for beating hard teams. But we don't know, because they don't release their formulas.

[jabba]

C.F.R.,

What do you think about the Rose Bowl trying to stick with tradition, even when it means taking a lesser team? I could understand it if the Big 10 had a down year but the Rose Bowl wanted to take the champ anyway, but to take the runner up? What's traditional about the Pac-10 champ playing the Big 10 runner up?

I'm all for Pac-10 Big 10 in the Rose Bowl when it involves the champs, but not this nonsense. Now, instead of a compelling SC-UGA matchup, viewers will be "treated" to SC vs. Illinois. I think Mario Manningham put it best during his chat on espn.com today:

Elliott (Chicago): Hey Mario, you played against Illinois...do you think they have a shot against USC?

Mario Manningham: No.

[phil]

I thought the original concept for the computers was to break ties if the two human (originally AP and coaches) disagreed. That made sense to me.

You can't give the computers too much power in the BCS formula because the computers have no common sense. If we used the computer choices this year they would have taken VA Tech over LSU even though LSU slaughterd them when they played head to head (and since margin of victory was taken out that makes it even worse in this case.

The computer formulas are not all public knowedge anyway, so why should they be trusted? Based on the results this year it is clear that the the key factor is who a team lost to. Thus VA Tech losing to LSU ends up hurting them far less than USC losing to Stanford or Oklahoma losing to Colorado. Is that right? Maybe partially, but most fans look more towards who a team has beaten in order to determine their worthiness for a championship game.

[mmortal03]

The thing is, the human polls don't disagree in most cases, but they DO introduce massive cumulative bias in most cases. Yes, the computer formulas should all be made public, and they should be designed for a specifically defined goal, where every voter is instructed as to what the goal is, and the voters should have to start with the averaged computer poll as their starting point, and not their previous week's ranking. Then, they should rank them based on that. If these voters had an idea of how the computers worked, instead of some subjective view of the game that they have in their head, then they would give more respect to the computer's output.

Then they could move the teams around based on correcting whatever "lack of common sense" that they think the computer results have. That way, you don't get such the b.s. where voters drop teams for a loss further than they really should, some nonsense number of spots that they pulled out of their asses. Furthermore, they get a feel for ranking teams on more objective merits.

If you start with the computers and have the humans correct it, it would have less error than if you give the humans free reign and then the computers somehow have to correct for such blatant stupidity.

[CFR]

Jabba,

Others can correct me if I'm wrong, but I think the Rose Bowl didn't have a chance to grab Georgia as the Sugar Bowl used some exemption to protect from a remaining SEC team getting grabbed.

The Rose may have used that little loophole in other years as well to retain either a Big Ten or Pac-10 team.

Thus, I believe the Rose had anyone outside of the BCS title game locks and USC and Georgia to choose from and at that point if it couldn't get Georgia went with the traditional matchup instead of a Hawaii or whatever. I could be wrong but I think that's how it went so I'm not sure how to judge what happened.

I like watching Illinois play for what it's worth, more than USC particularly on offense. USC's defense may just clobber them but if Illinois has a good day on offense we might actually get a decent game because we know USC's offense doesn't really care to do much beyond 24 points/game of late.

[CFR]

"Yes, the computer formulas should all be made public, and they should be designed for a specifically defined goal, where every voter is instructed as to what the goal is"

WTF?

What could possibly be the goal? It would obviously have to be something mathematical and just as arbitrary as whatever voters reason with, which is just weird as hell and doesn't really mesh to the game of football.

At the end of the day observation matters and I'm not even close to comfortable with having numbers and formulas have any kind of say over final rankings without any real human observation added to the mix.

The polls are imperfection, but at least they're imperfection with some soul and a combination of numbers awareness and actually seeing the product on the field.

[Dawgy]

CFR, it is my understanding that the Rose Bowl could have chosen Georgia. Once Ohio State was unavailable the Rose could have chosen any BCS team that was not lock to another BCS bowl by a conference championship i.e. Oklahoma, Viginia Tech.

What's also disturbing is that after the Rose Bowl dumped on us fans the Sugar Bowl did the same thing as they could have chosen West Virginia or Kansas or Mizzou.

I'm not positive on this but, I believe that's the way it went down.

It's my belief that after the BCS NC teams were selected that the BCS bowls were "encouraged" not to upstage the MNC? We couldn't have USC/Ga because a lot of fans and media might perceive that as the real NC game?

This is why we need a playoff. Any negatives to a playoff couldn't possibly offset all the negatives of this pitiful system we have. Bad, very bad.

[CFR]

Well right now you and I both have different, uncertain sides of the story. I'll see if I can learn more, but I don't think the sometimes lousy politics of this exactly poisons things.

Postseason play is the cherry on top for college football's wonderful regular season. Not every bowl is going to be to our liking and I'd like to see something like USC/UGA, but at the same time we can work towards that, it's not like it's set in stone that in the future something like that will happen.

A playoff is a guarantee towards what I think is a lousy way to end the season, without any real way to amend what happens the way we can tweak BCS and bowl selection rules and criteria.

[mmortal03]

"Yes, the computer formulas should all be made public, and they should be designed for a specifically defined goal, where every voter is instructed as to what the goal is"

WTF?

What could possibly be the goal? It would obviously have to be something mathematical and just as arbitrary as whatever voters reason with, which is just weird as hell and doesn't really mesh to the game of football."

No different a goal than, say, what you set out for with your poll, CFR, except it wouldn't be allowed to break its reasoning a few spots down. You say that reason and logic doesn't mesh with the game of football? What do you want to use to rank them? Fairydust, lollipops, and magical emotions?

It could be something consistent, agreed upon, and understood, like the RPI, except tweaked for football, and the voters would use it as a baseline. You just can't tell me that the voters are currently consistent and methodical with their entire top 25.

If the voters are going to be trusted they should be trying to use a computer to apply their methodologies consistently, across all teams. If they can't implement their own basic method mathematically, then it just isn't worth anything.

Voters today just throw names up there based on some inconsistent method using their fingers to count on and their imagination, mixed in with a bit of hype.

I actually trust the computer polls to get it just about right as a starting point, and then the human voters could go from there if they think something is radically out of line. The human voters need to go back to math class and understand what the computers are doing; otherwise they are totally pathetic.

[CJH]

Not surprisingly, those opposed to "computers" do not understand their purpose. Computers are actually just different sets of objective rules. Like all objective rules, whether they be for a game or competition, their purpose is simply determine the the winners of said game or competition. Furthermore, it is unfair to dismiss any objective rules outright based are their results when you don't know why the results are what they are on top of the fact no teams are actually influenced by those rules. Based on the system I devised, if schools continue the cupcake scheduling, they would hurt their teams would be punished for it much more than by the BCS rules.

BTW, another thing some fans do not seem to understand, is that there is no such thing as a universally correct result. I wrote about this on my own blog, but when people ask if the BCS got it right, they only show that the BCS is inherently wrong as a result of the question being asked. Whether or not anyone agrees with a particular set of rules, those rules must produce one internally undisputed result. Otherwise, they don't qualify as rules. If a set of rules can produce more than one answer given the exact same results, than who wins is completely arbirtrary.

http://www.objectiverules.blogspot.com

[Mikem]

Throw out the BCS computer and just sit down and set up a playoff brackett system. I can't understand how lower divisions can have a playoff system and so do the pro's,and college basketball, but D1 has to have a computer determine it's national champ. I realize there is big bucks invloved and all those bowl backers will loss lots a money,they should use that money in creating a playoff system that is fair and will allow teams from any conference to be considered.College football is big money and you don't want to upset the status quo.. College football fans need to rebel against the BCS and lets create a system that works for everyone.