CollegeFootballPoll.com Message Board

http://www.collegefootballpoll.com/analysis_sec.html

Congrove, interestingly, has Tennessee and South Carolina trailing Vandy in the East....Vandy no longer has Cutler, I wonder if they will do that well?

Tennessee has the potential to finish better than Vandy in the SEC if Ainge plays well. And I also think that South Carolina has better athletes than Vandy and may finish higher this year.

In the West...I have no argument whatsoever...I see it as Auburn, LSU...and everybody else. Bama may be lucky to win ten without Brody.

I personally think that the one thing computer rankings have the HARDEST time with...

is returning starters (or lack thereof).

You cannot possibly give each lost starter any more impact on the team than another.

If you lose 10 starters, you lose (random number) 10 power points. Even if one was a left tackle and the other was a Heisman winning QB.

Oh well.

If Vanderbilt couldn't beat Middle Tennessee with Cutler.....

They're going to have a really hard time in 2006. I'm thinking a 3-9 season should be pretty good. Maybe 4 is the maximum:

9-2-06 at Michigan
9-9-06 at Alabama
9-16-06 ARKANSAS
9-23-06 TENNESSEE STATE (tougher than expected)
9-30-06 TEMPLE
10-7-06 at Mississippi
10-14-06 at Georgia
10-21-06 SOUTH CAROLINA
10-28-06 at Duke
11-4-06 FLORIDA
11-11-06 at Kentucky
11-18-06 TENNESSEE

I see them at 3-9.

I agree with Open Skies...

Computers also can't balance loss of "starters" to the quality of the returning back-ups...in many power schools, the "starters" are in name only and the backups play as much time.

Last year's Nole starters and first rounders Bunkley and Wimberly were backups the year before. The computers can not score that.

openSkies wrote:I personally think that the one thing computer rankings have the HARDEST time with...

is returning starters (or lack thereof).

You cannot possibly give each lost starter any more impact on the team than another.

If you lose 10 starters, you lose (random number) 10 power points. Even if one was a left tackle and the other was a Heisman winning QB.

Oh well.

Uh....OpenSkies....This is a completely incorrect post.
A) The exact methodology isn't revelead
B) The use of "experience" (returning starters/lettermen, etc.) as a factor is minor
C) The mathematical process of its use is nothing close to what you suggest.

Eric wrote:If Vanderbilt couldn't beat Middle Tennessee with Cutler.....

They're going to have a really hard time in 2006. I'm thinking a 3-9 season should be pretty good. Maybe 4 is the maximum:

9-2-06 at Michigan
9-9-06 at Alabama
9-16-06 ARKANSAS
9-23-06 TENNESSEE STATE (tougher than expected)
9-30-06 TEMPLE
10-7-06 at Mississippi
10-14-06 at Georgia
10-21-06 SOUTH CAROLINA
10-28-06 at Duke
11-4-06 FLORIDA
11-11-06 at Kentucky
11-18-06 TENNESSEE

I see them at 3-9.

One player does not a championship team make. Case in point: John Elway. As good as he was, he never played in a bowl game (thanks in no small part to the Cal vs. Stanford 'miracle').

That being said, one good player can 'inspire' a team to do better. Just because Vanderbilt lost their 'star' QB doesn't mean they won't be good.

I think the CFP ratings, by-and-large, are pretty close to being right. Any pre-season 'assessment' is going to miss some things. For one thing, nobody knows for sure what will happen, all are simply 'guesses' no matter how precise.

Tell me, did the computer 'predict' that Vince Young would win the Rose bowl? In other words, too many variables to adequately predict.

I think the only way to really know how a team will do, is to play the game, and let the results speak for themselves. One reason whY I support a BCS that allows for 'equal' representation. The BCS doesn't really select a 'unanimous' NC, even last year. Penn St never played Texas. Neither, for that matter, did W. Virginia. I"ts an inexact science.

Look, Vanderbilt was a pathetic team with an NFL-caliber quarterback; he's what made them competitive. Having a transfer from Arizona that wasn't great to begin with is too big of a downgrade.

Actually...one or two players at skill positions can make a huge impact...When a Duante Culpepper plays for UCF, they are much better, or Phillip Rivers at NC State, Randy Moss at Marshall, etc.

VT will not be as good this year without Vick at QB...and Texas will miss Vince Young...

Take Brohm and Bush away from Louisville, and they look like a very different team.

I can see CLF's point. Look at Tennessee. Peyton Manning didn't win the National Championship - they won it with Tee Martin one year after Manning left.

Still, Vanderbilt may not win 6 games and maybe it's ridiculous to even consider the notion of them doing so. But, what if....

billybud wrote:VT will not be as good this year without Vick at QB...and Texas will miss Vince Young...

The Hokies are better off WITHOUT Marcus Vick. He was a cancer on the team. We didn't win any games because of him. We did lose to Miami becasue of him. VT will be as good, if not better, without him.

Tee Martin was a very, very good QB at times...he did labor in the shadow of Manning...but:

T Martin holds the NCAA record for:

Most Consecutive Completions

Highest Completion Percentage in a Game

Lowest Career Interception Rate

Tennessee was a different offense under Tee Martin than Peyton...in four years, Peyton attempted 1354, in four years, Tee attempted 588...

In a way, Peyton was the Dan Marino of Tennessee....

I believe when you look at polls, computer or human, it isn't how well they start it is how well they adjust. If a poll makes a mistake, does it hold back the team/teams it made a mistake on or does it fix the error.

I know the congrove poll wants to have it all and has a good record picking, but how well the computer adjusts to the way the season plays out is the most important thing to me.

CFP Admin wrote:Uh....OpenSkies....This is a completely incorrect post.
A) The exact methodology isn't revelead
B) The use of "experience" (returning starters/lettermen, etc.) as a factor is minor
C) The mathematical process of its use is nothing close to what you suggest.

CFP Admin, like I said, I threw up a random number. And that's just my own opinion on the matter. I just think that (IMO), returning or leaving star players would be one of the most impossible things to get a solid number on. Also, I said "computer rankings"... not CCR. As a whole. May be incorrect for your formula, but not all.

Am I wrong in assuming that one lost starter is considered just as important in the formula as another, even though one is a Left Tackle and one is a Heisman winning QB? If so, guess I was wrong =] It was at 9am after being up all night, though, haha.

Eric wrote:If Vanderbilt couldn't beat Middle Tennessee with Cutler.....

They're going to have a really hard time in 2006. I'm thinking a 3-9 season should be pretty good. Maybe 4 is the maximum:

9-2-06 at Michigan
9-9-06 at Alabama
9-16-06 ARKANSAS
9-23-06 TENNESSEE STATE (tougher than expected)
9-30-06 TEMPLE
10-7-06 at Mississippi
10-14-06 at Georgia
10-21-06 SOUTH CAROLINA
10-28-06 at Duke
11-4-06 FLORIDA
11-11-06 at Kentucky
11-18-06 TENNESSEE

I see them at 3-9.

That is quite teh difficult scheulde...Sucks to be them. I'll agree 3-9.

Spence wrote:I believe when you look at polls, computer or human, it isn't how well they start it is how well they adjust. If a poll makes a mistake, does it hold back the team/teams it made a mistake on or does it fix the error.

That is one of the problems with the whole polling system. Certain teams start at or near the top every year, based on history, while other teams have to make the long climb. Iowa is a great example of a team that seldom gets any early respect. (OK, I am a homer)