The CSN Way; FCS Top 25 Quick Who, Where, When; The CAA Today; CSN West <---CSN

[CSN-info]

The CSN Way: The Number One Question
Charles Burton and Ralph Wallace, CSN Columnists
For two straight weeks, pollsters and computers have been digesting the data. People have looked over game results. Game data about Appalachian State, Montana, and North Dakota State have been plugged into a variety of different mathematical formulas. All in an effort to find out who is the real No. 1 team. The results, for two straight weeks, have been inconclusive.
READ MORE... http://www.collegesportingnews.com/article.asp?articleid=87966

FCS Top 25 Quick Who, Where, When
College Sporting News
There are two games between ranked FCS teams and three ranked teams are idle.
READ MORE... http://www.collegesportingnews.com/article.asp?articleid=87962

The CAA Today: Weeks Five & Six – Where is the Rankings Logic?
Bruce Dowd, CSN columnist
I am so grateful that we have the playoffs to make that final determination. However, one unfortunate part of the playoffs is that the rankings still play a key part in the outcome starting with the obvious one of who gets in and who stays home. Almost as important, is the final seeding which determines who plays who on the road. So, please, can somebody give the polls a shot of logic! It is definitely needed.
READ MORE... http://www.collegesportingnews.com/article.asp?articleid=87968

CSN West: Where’s the Big Sky Been This Year?
Kent Schmidt, CSN Columnist
It is not too often that a conference has a number one ranked team in the country and that conference sees little press coverage. That, however, is the case for the Big Sky Conference this year.
READ MORE... http://www.collegesportingnews.com/article.asp?articleid=87949

[blukeys]

Good articles. Bruce Dowd is doing yeoman's work promoting the CAA. He makes a very persuasive case that CAA teams are underrated.

Of course he picked on the the big bear himself and is sure to get some emails for this quote

"To be fair, I thought okay, perhaps it was just a quirk in the schedule and the rest of the season gets real difficult with some FBS games or something. Wrong! The eleven teams on Montana’s schedule are now a collective 14-34. Seven of those wins came against each other. I decided to dig even a little deeper. For all the teams collectively on Montana’s schedule, only seven victories have come against teams other then themselves. Of those seven wins, only two can even be considered a semi difficult win. Those two wins were Albany’s win against Fordham and Eastern Washington’s win against UC Davis. Who were the other five against you ask? Well, even if you didn’t ask here they are. They were against Montana Western, South Oregon, Western College, Dixie State, and Western New Mexico presented by NAU Army ROTC. Yes you read that last one correctly. There are no games on Montana’s schedule against any FBS schools. As mentioned they already played a Div II school and the only FCS ranked team on their schedule is their final game of the year against No. 13 Montana St."

Stay tuned for the "Montana Reaction" (as opposed to the old song Psychotic Reaction)

[Cap'n Cat]

Regarding Montana getting #1 votes in the preseason, no one had bothered to peruse their schedule yet.


xwhistlex

[BobbyMo]

The CSN Way: The Number One Question
Charles Burton and Ralph Wallace, CSN Columnists
For two straight weeks, pollsters and computers have been digesting the data. People have looked over game results. Game data about Appalachian State, Montana, and North Dakota State have been plugged into a variety of different mathematical formulas. All in an effort to find out who is the real No. 1 team. The results, for two straight weeks, have been inconclusive.
READ MORE... http://www.collegesportingnews.com/article.asp?articleid=87966

Always enjoy The CSN Way. Just wanted to say thanks for including an NEC game. I think it will be much tighter than the predicited score but always enjoy reading other peoples thoughts on the NEC.

[McNeese75]

Texas State Over No. 6 McNeese State
We all know that McNeese State is a very good team: quarterback Derrick Forroux and defensive end Bryan Smith are great athletes. But they will face a test in Texas State despite their bad loss to South Dakota last week 38-3, and having lost their last three games after a season-opening win against nationally-ranked Cal Poly. The Cowboys are banged up in major offensive positions (runningback and wide receiver), while the Bobcats will be playing for their season this year as runningback Stan Zwinggi seems to finally be getting healthy. Add to the fact that Texas State has won three straight in this series, the Southland always seems to have a few home surprises in the league, and a huge heaping helping of “gut feel”, and guess what, Virginia. You have an upset.
Believable Bobcats 30, Cow-Towed Cowboys 25

Ralph and Chuck need to do a little better on their homework xnodx

The two deepest positions on the McNeese Team are wide receiver and runningback. Based on what I have read there is a good chance Zwinggi is not going to play this week.

The Cat defense has not been exactly stellar so far this year so I am not sure how they are going to hold the Poke offense to its lowest offensive output of the year. But, who knows, AGS and unless McNeese takes care of the ball better than they have the last three years, who knows xcoolx

[Gil Dobie]

EWU over Montana Hmmmmmmmmm xeyebrowx

[Dane96]

"Critics of computer-based ranking systems are fond of saying that the programmers’ inherent bias is present in the data that is chosen to crunch in order to make the rankings. This is not a fair criticism. Every person has an opinion as to what is most important in a computer poll and what weight it has, but it is applied evenly and does not discriminate."

TWO WORDS: CIRCULAR REASONING. Clearly, logic goes out the door when writing this. By virtue of bias, how does bias even bias out...not sure I understand that...at all!

[McNeese72]

Texas State Over No. 6 McNeese State
We all know that McNeese State is a very good team: quarterback Derrick Forroux and defensive end Bryan Smith are great athletes. But they will face a test in Texas State despite their bad loss to South Dakota last week 38-3, and having lost their last three games after a season-opening win against nationally-ranked Cal Poly. The Cowboys are banged up in major offensive positions (runningback and wide receiver), while the Bobcats will be playing for their season this year as runningback Stan Zwinggi seems to finally be getting healthy. Add to the fact that Texas State has won three straight in this series, the Southland always seems to have a few home surprises in the league, and a huge heaping helping of “gut feel”, and guess what, Virginia. You have an upset.
Believable Bobcats 30, Cow-Towed Cowboys 25

Ralph and Chuck need to do a little better on their homework xnodx

The two deepest positions on the McNeese Team are wide receiver and runningback. Based on what I have read there is a good chance Zwinggi is not going to play this week.

The Cat defense has not been exactly stellar so far this year so I am not sure how they are going to hold the Poke offense to its lowest offensive output of the year. But, who knows, AGS and unless McNeese takes care of the ball better than they have the last three years, who knows xcoolx

If we had lost a Offensive or Defensive linemen, I'd have really been worried. But we have so many good WR's and RB's that have and will step up, we really haven't missed Steven Whitehead (for almost three complete games now) and Jamie Leonard for the whole game last weekend. If we don't make all the turnovers we made last season against them, we should win. And I don't think Derrick wil throw the interceptions he threw last season against them. He is a lot better QB now than he was as a redshirt freshman at the beginning of the conference season last year.

Doc

[slycat]

Texas State Over No. 6 McNeese State
We all know that McNeese State is a very good team: quarterback Derrick Forroux and defensive end Bryan Smith are great athletes. But they will face a test in Texas State despite their bad loss to South Dakota last week 38-3, and having lost their last three games after a season-opening win against nationally-ranked Cal Poly. The Cowboys are banged up in major offensive positions (runningback and wide receiver), while the Bobcats will be playing for their season this year as runningback Stan Zwinggi seems to finally be getting healthy. Add to the fact that Texas State has won three straight in this series, the Southland always seems to have a few home surprises in the league, and a huge heaping helping of “gut feel”, and guess what, Virginia. You have an upset.
Believable Bobcats 30, Cow-Towed Cowboys 25

Ralph and Chuck need to do a little better on their homework xnodx

The two deepest positions on the McNeese Team are wide receiver and runningback. Based on what I have read there is a good chance Zwinggi is not going to play this week.

The Cat defense has not been exactly stellar so far this year so I am not sure how they are going to hold the Poke offense to its lowest offensive output of the year. But, who knows, AGS and unless McNeese takes care of the ball better than they have the last three years, who knows xcoolx

hey who knows. i laughed when they picked acu to beat us and look waht happened. i fully expect yall to give us a tough time but you never know.

[Proud Griz Man]

EWU over Montana Hmmmmmmmmm xeyebrowx


Montana +13 versus Eastern Washington? I'll take that bet Mr. Charles Burton. xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx xnodx

[Lehigh Football Nation]

hey who knows. i laughed when they picked acu to beat us and look waht happened. i fully expect yall to give us a tough time but you never know.

I just wanted to note that I am 2-0 in picking Texas State games this year xnodx

[Lehigh Football Nation]

"Critics of computer-based ranking systems are fond of saying that the programmers’ inherent bias is present in the data that is chosen to crunch in order to make the rankings. This is not a fair criticism. Every person has an opinion as to what is most important in a computer poll and what weight it has, but it is applied evenly and does not discriminate."

TWO WORDS: CIRCULAR REASONING. Clearly, logic goes out the door when writing this. By virtue of bias, how does bias even bias out...not sure I understand that...at all!

A person who comes up with a computer model has a philosophy as to what makes a team good. Some have a lot of factors, some only have a few, but no individual team bias enters into play. For example, Sagarin has Northern Iowa at #1 in FCS this week. Is this because Sagarin is a closet Panther fan? NO! It's because he plugged in the numbers of all the FCS teams, had his criteria, applied them evenly, and what was spit out at the other end was: Northern Iowa #1. There is no "bias" towards one team or another. The criteria are applied evenly.

You can argue that human polls have biases in them, since human polls are inherently flawed if they lean too far towards one region, not all humans have "perfect knowledge" of all the games, etc. But computer polls do not. They are a set of rules created by the designers, applied to statistics.

As for throwing out the top score and the lowest score, that has been a way to get rid of some bias in all sorts of rating systems. For example, in Olympic figure skating they do just that. So if one rating system is almost a pure "margin of victory" formula and has (for example) San Diego as #1, that one is thrown out in favor of other polls that have that criteria not carry as much weight, if any. Unlike the Olympics, nobody is paying CSN to "weight" their system over one another. The systems are chosen on their merits.

No system, not less the GPI, proports to be a "perfect" system. The GPI is a hybrid index that takes a bunch of polls (I think the number is 10) and tries to reconcile all those ratings to come up with an index. Like the computer polls, it does not favor one team or conference over another: data gets grabbed, the numbers get crunched, and the "GPI Ranking" comes out the other end.

The idea of the GPI or any computer model having "bias" does not enter into play, while the biases of the human voters are tempered by throwing out the highest and lowest ranking and weighing them against computer models.

[DetroitFlyer]

Of course computer models have "bias".... They all have the bias of the folks that created them.... Once you get past that bias, the programs themselves are 100% objective assuming that the programs do not change over time. That is why the idea behind the GPI makes sense. Every computer model has the bias of its designer.... But, when you take a number of these polls or programs, and roll them up into yet another objective program, you tend to minimize the biases that exist. I have never heard anyone say that the GPI is 100% perfect, but it is no doubt one of the best reference tools available for measuring teams during the season....

The problem I have is that at large playoff bids are 100% subjective. The "committee" is certainly influenced by various polls and the biases that each member has from their own experience. There is simply no way for any one person to ever know enough about FCS teams scattered across a vast country to be able to make the 100% "right" decision for at large bids. At least the AQ's are based on fact. Fact: This team won its conference. For at large bids, a judgement is made as to whether the third place team in this conference is better than the first place team in this conference, ( as in Gateway and PFL for example ).

[dungeonjoe]

I just wanted to note that I am 2-0 in picking Texas State games this year xnodx
you know what they say about blind squirrels:)

[Dane96]

A person who comes up with a computer model has a philosophy as to what makes a team good. Some have a lot of factors, some only have a few, but no individual team bias enters into play. For example, Sagarin has Northern Iowa at #1 in FCS this week. Is this because Sagarin is a closet Panther fan? NO! It's because he plugged in the numbers of all the FCS teams, had his criteria, applied them evenly, and what was spit out at the other end was: Northern Iowa #1. There is no "bias" towards one team or another. The criteria are applied evenly.

You can argue that human polls have biases in them, since human polls are inherently flawed if they lean too far towards one region, not all humans have "perfect knowledge" of all the games, etc. But computer polls do not. They are a set of rules created by the designers, applied to statistics.

As for throwing out the top score and the lowest score, that has been a way to get rid of some bias in all sorts of rating systems. For example, in Olympic figure skating they do just that. So if one rating system is almost a pure "margin of victory" formula and has (for example) San Diego as #1, that one is thrown out in favor of other polls that have that criteria not carry as much weight, if any. Unlike the Olympics, nobody is paying CSN to "weight" their system over one another. The systems are chosen on their merits.

No system, not less the GPI, proports to be a "perfect" system. The GPI is a hybrid index that takes a bunch of polls (I think the number is 10) and tries to reconcile all those ratings to come up with an index. Like the computer polls, it does not favor one team or conference over another: data gets grabbed, the numbers get crunched, and the "GPI Ranking" comes out the other end.

The idea of the GPI or any computer model having "bias" does not enter into play, while the biases of the human voters are tempered by throwing out the highest and lowest ranking and weighing them against computer models.

Though I did complete it, I didnt have to read past your first line-- BIAS...IS BIAS...team oriented, model oriented, or functionality oriented.

Hence, by logic...all COMPUTER MODELS ARE BIASED. This has nothing to do with the teams themselves, rather it deals with the WEIGHTING of the data used in each computer model. If I say away game scores are important in my model, that is a bias.

That, was my point and something proponents (not saying you are one) of the GPI fail at the very least to admit, though they may understand it.

A compilation of BIAS...does not flesh out the most NON-BIAS "index." The only way this could be done is if all the polls and indexes met and said "HEY...I am bias this way...you are that way...let's have them balance." It would have to be a 50-50 balance of let's say away scores are important in my poll and home scores are impt in yours; SOS is impt in my poll, while SOS is not in yours.

Nope...never gonna happen.

Bias is bias.

[HIU 93]

They picked Norfolk State over South Carolina State. Are they drunk?

[Dane96]

Clearly

[Lehigh Football Nation]

http://esoriano.files.wordpress.com/2007/05/drunk.jpg

What The Hell.... Norfolk State this week!

[slycat]

I just wanted to note that I am 2-0 in picking Texas State games this year xnodx

im 1-3 in picking texas st games. i picked mcneese st this week. hopefully im wrong again.

[skinny_uncle]

EWU over Montana Hmmmmmmmmm xeyebrowx
It's the closest thing the Griz have had to a real game this year.