How does Congrove do this?

[billybud]

Congrove has filled in spots on his power rankings...but how can he do that without completing all the teams?

http://www.collegefootballpoll.com/pre_ ... kings.html

I am curious about the methodology...on the surface, it loks odd to have a team pluncked down at, say #51, with blank spaces left all around...

[mountainman]

Hey, BlackpowerEER, as far as the Mountaineers are concerned, I'm sure you will agree that they have been hard to pin down as far as predictions are concerned.

We've won when we were not suppose to and lost when we were not supposed to .

It has been quite a ride for both them and us die-hard fans (not to mention the predictors).

[billybud]

LOL...more of "it's a dark conspiracy against Wext Virginia", etc.

Yep...writers, pollsters, rankers, all line up to conspire to drag down WVU....mention WVU's weak schedule and it's an attack on WVU...rank WVU based on perceived power in the national rank, rather than on predicteded number of wins, and it's a conspiracy.

I do not think that ranking algorithimns are designed to "assist" some schools and to "devalue" others...although the basic philosphy behind the logic sequences may incidentally do so (such as some systems put more value on the total number of predicted wins others value more highly wins vs value of opponents).

The seasons have a way of sorting themselves out...

[mountainman]

Don't know where you came up with the 1st half of your post, billybud ... I sure didn't read anything like that.

But, I believe the 2nd half of your post is correct and right on point.

[billybud]

Phil Steele is a Virginia resident and a HUGE VT fan. So much so that he has been known to call into radio shows unannounced to extol the virtues of VT. He is a well known critic of the Big East....and all things Big East. Go do a quick look at his polls for the last 10 years. I think you will see what I'm talking about. His poll artifically boosts the stock for VT and his poll has been wrong about them more than it was right. Is this on purpose??

1. Phil Steele's polls artificially boost the stock of VT...huh?

The lead guy in this endeavor is a Louisville grad. I don't think it is strange then, that UofL is artifically boosted in the ratins of this game. And once again I don't think it's some dark conspiracy against WVU and the Big East....I just think it's the fault of the programmers due to thier own tendencies and the computers inability to discern what is homer information and what is not. He is also a programmer for ATS Consultants. Notice that ATS's ranking is one of 2 that have Louisville ranked in the Top 5. Is this on purpose?? I don't think so.....but it shows (in my mind) that it's possible to change the outcome of the predictions by creative input by the programmers.

2..And because a guy at EA Sports is a Louisville grad da dah...da dah.. Blackpowder makes an allegation, than says, "not that I think this is happening"...(By the way, ever wonder why FSU has been on more covers of EA than any other team...yep, an FSU grad must work there).


Math itself, has no favorite and can not discern the name of the team (you could label them as team #1 through #119)...sequences of logic progression, what a algorithmn values, may have a logical bias...again, valuing a win in different ways, valuing opponent strength in differing ways and amounts, etc.

It would be really interesting to have Congrove himself address whether having a prior team affiliation has any impact on algorithmns...'d like to hear what a professional coder/ranker has to say.

[mountainman]

I'm sure that Congrove tries to be fair and exact....just like the rest of the computer poll operators.
**
Is this on purpose?? I don't think so, I think it's just human nature to have your favorites and you have a tendancy to overlook some minor things to put your favorite team in the best light possible.
**
And once again I don't think it's some dark conspiracy against WVU and the Big East....
**
And no...I don't think there is a conspiracy...I just think it's human nature.

Here's what the man wrote .... like I said, don't know where you came up with the 1st half of your post.

[billybud]

Well than read what I quoted...he makes allegations against EA and Steele...and than "hedges"....if he had just written what you quote, I would not have responded.

[billybud]

"Artificially boosted"...means just what it means...that someone went out of their way to "artificially boost" a team...You like to ignore that part...that's my objection to his post. I do not believe it..

To artificially boost is to cheat...he is accusing rankers of cheating...

[colorado_loves_football]

Congrove has filled in spots on his power rankings...but how can he do that without completing all the teams?

http://www.collegefootballpoll.com/pre_ ... kings.html

I am curious about the methodology...on the surface, it loks odd to have a team pluncked down at, say #51, with blank spaces left all around...

I'm pretty sure all those 'blanks' will be filled in shortly, its' a work in progress, Billybud, it's not like Las Vegas, where you put your money down, and 'let it ride' this is something a lot more scientific (and hopefully a lot more reliable, too).
Interesting, TCU basically 'winds-up' where they did, last year. I don't entirely disagree, I think they will likely 'stumble' along the way, but I don't think it will be to Texas Tech. Still, that puts them somewhere at, or near, #11 overall, a 'fair' rating if there ever was one.
Also interesting (to me, anyway) is Louisville's #4 ranking. I think that may come to fruition, assuming Miami, doesn't stop them, as they did, 2004. Still, 12-0 and #4? Guess that means 3 other teams 'run the table'. I'll be interesting in seeing which ones.
There aren't any 'holes' this rating has been proven time & time again, as one of the 'most reliable'. Guess we'll find out where FSU stands.

[billybud]

CLF...once again, you failed to comprehend the meaning of the post...sigh. I was wondering about how one can rank some teams without ranking all 119 teams..it's a math question that I guess that Ishouldn't expect you to pick up on.

Mountainman...I see Coaches have Louisville at 13...Steele has Louisville at #3...he is obviously "biased" towards the Big East...maybe he knows that guy at EA Sports who is a Louisville grad..LOL

[openSkies]

Congrove has filled in spots on his power rankings...but how can he do that without completing all the teams?

The computer formula is 100% completely filled in and ready to go. Every year CFP releases the information over the course of a week for a few different reasons (in order of lowest ranked conference to highest ranked conference). Two of the main reasons are listed below:

(1) Marketing. Why spit everything out at once and get a huge number of visitors on one day only, when you can bring them back every day for a whole week?

(2) Most importantly... Time. It takes a while to go through 150+ pages and update everything. It also takes a while to write the articles for each conference preview. Dave has been working on all of this for about a month now. If we waited until everything was ready to go online at once, you guys wouldn't see a season preview until around August 20th.

[billybud]

Thanks...that answers my math question..actually, I knew it had to be completely done...having worked with statistics and ranking in my career...

[mountainman]

Everyone puts so much emphasis on computer rankings, but what everyone forgets is the computers only spit out what the programmers put in.
**
Everyone seems to overlook the programmers tendencies are human and they, like all of us have thier favorites.
**
Phil Steele is a respected handicapper in College Football. His poll artifically boosts the stock for VT and his poll has been wrong about them more than it was right. Is this on purpose?? I don't think so, I think it's just human nature to have your favorites and you have a tendancy to overlook some minor things to put your favorite team in the best light possible.
**
I just think it's the fault of the programmers due to thier own tendencies and the computers inability to discern what is homer information and what is not.
**
...but it shows (in my mind) that it's possible to change the outcome of the predictions by creative input by the programmers.
**
...I just think it's human nature.

Again, here's what the man wrote .... no mention of cheating .... cheating to me is doing something to swindle or defraud or mislead or delude ... who would these guys be cheating? I know of no consequence to anybody if they are wrong or right. It's anyone's choice whether these polls are viewed as creditable or not creditable.

Like I said earlier, I thought the 2nd half of that post was correct and right on point.

[Derek]

THIS is why there are several programs, 6 I think total...They take the top ranking and the bottom ranking and throw them out.

There is a process to take care of any inconsistency's in the programs.

They DONT just blindly take a comptuers word for it. They know what's going in and it's not Garbage.

[billybud]

Again...my last post on the matter...since you only want to post half the post...

He is saying that certain programmers design their algorithimn to favor a team that they like....that is the jist of his argument...(and if I know that you read well enough to figure out his point)...I cry Bullfeathers...

A human poll is much more likely to be subject to such a bias than a series of mathmatical equations. Now, I am not privy to Steele's seven power equations, nor am I privy to Sagarin, Congrove, etc. But, I think it is absurd to think that the math is designed to favor a specific team.

I, for instance, do not like Congrove as well as some because I think that his math does not count for SOS as stringently as some..(and I freely admit that this is a casual observation based on his rankings, not an analysis), but I do not believe that Congrove has coded his system solely to let him rank Fresno State #12 as he has. I think that the Coaches don't rank Fresno State because they know Fresno is not a Top 25 team...Congrove ranks Fresno #12 because his computer algorithmn is designed that way...I don't think that Congrove is a fan of Fresno, I just think that his algorithmn ranks Fresno much higher than most will. It's just the math.