re: The great polling debate....

quote:

---

The problem is that it is proof of nothing. For it to be proof, it would have to actually explain what we observed. For example, if preseason ranking controls, how was AU able to tie OK in the AP one week? How does preseason ranking explain why AU was ranked behind USC and OK in practically every computer poll known to man?

---

Whoa feller... Re-read your post below and it does prove they have "power" that "they are claimed to have".

quote:

---

I wish someone would actually provide some evidence that preseason polls have the power that they are claimed to have.

---

Don't go flip flopping on me like that by changing it to a question of legitimacy.

quote:

---

quote:


The problem is that it is proof of nothing. For it to be proof, it would have to actually explain what we observed. For example, if preseason ranking controls, how was AU able to tie OK in the AP one week? How does preseason ranking explain why AU was ranked behind USC and OK in practically every computer poll known to man?


Whoa feller... Re-read your post below and it does prove they have "power" that "they are claimed to have".

quote:


I wish someone would actually provide some evidence that preseason polls have the power that they are claimed to have.

---

The claim is that the preseason ranking order causes the final ranking order (i.e. the final for BCS selection). 2004 is said to be proof of this cause and effect. I'm saying that the 2004 preseason and final rankings are merely coincidently because there is a more rational explanation for the voter behavior than preseason polls.

First, after an impressive win over GA, AU actually tied OK for #2 in the AP poll. The following week they dropped to #3. Which explanation for this makes more sense:

1. Due to a case of temporary amnesia, voters suddenly remember that AU was the lower ranked team in preseason and decided to correct the grave injustice. Or,

2. The tie put AU's schedule under the spot light. Remember the Citadel became a battle cry to represent AU's weak OOC schedule. The increased scrutiny combined with a lethargic win over a mediocre Bama team caused a drop in the polls.

I submit explanation 2 makes more sense to a rational person.

Second, if 2004 was proof of preseason rankings being the sole cause of the final rankings, one would expect that rankings known not to be dependent on preseason rankings would get contradictory results. As I pointed out, essentially every computer poll known to man had AU 3rd among the 3 teams. So what makes more sense:

1. The power of preseason human polls is so great that the final rankings of inanimate objective models were bent to its will, or

2. Other factors, like SOS, put AU at a decided disadvantage and this was the primary influence in the final human rankings.

Again, I pick 2.

Maybe preseason rankings effect final rankings as you say it does, but 2004 is not proof that it does. The final rankings were mere coincidence.

Polls create interest, that creates audiences, that crete advertising interests, that sells ads, that crete revenues.
They are here to stay.

They are also like opinions and you know what's said about opinions...
they are like A@#H&^*, everyone has one.

However, if they are going to exist.....
they should not play a part in the ongoing rankings...as it is today, a team ranked #1 does not loose its place unless it looses....they get 100% of the doubt.
An opinion should not carry that much weight without "proof"
I think a pre-season poll is OK, but after about 3-5 weeks "RANKINGS" based on "numbers" should be generated and the pre-season poll thrown away.

quote:

---

Neither, they should be ranked after week three according to how good they look at that point.

---

+1

Juging teams before seeing what they can do is a bad idea. Politics, getting good ratings (playing up to teams in bigger markets)and such get into the mix.

quote:

---

think a pre-season poll is OK, but after about 3-5 weeks "RANKINGS" based on "numbers" should be generated and the pre-season poll thrown away.

You do realize how utterly silly that is, right?

---

That's not silly at all. Have your pre-season rankings, sell your paper, your ads, cause debate. However, after 3-5 weeks of empirical data then start the rankings, just like the computers. The computers need data to come to a conclusion, so should the pollsters.