Coaches Poll Top 25 (ND #7)

[Cackalacky]

I'm not sure what that has to do with anything. Navy cut blocks left and right and we still hung 50 on them in 2012. Margin of victory has more to do with quality than style.

It has very much do with the whole game. Citadel injured two of their starters and a back up. FSU paid Citadel to play them. Fisher shut it down so no more key players got hurt. Why risk it on top of paying them $900,000. Fisher said as much.

 

[T Town Tommy]

Oregon is by far the best team in the county right now. Hell Alabama at 2 is even a question mark for me. They struggled against WV and have a rotational QB situation.

Disagree on Oregon. Same old Oregon. Too light on the lines. Mich St will look at the film of that game and get sick to their stomach. I like the Sparties coach, but once they took the lead they missed opportunity after opportunity to close the Ducks out. The QB from State simply missed on two or three big plays to knock the Ducks out. Kudos to the Ducks for hanging on and eventually getting the plays they needed. Good teams do that and Oregon is good. But they are going to struggle when they get to the teams that play ball control, physical running game, and really good d line play. There is a reason why they lost to Stanford in the past. Maybe not this year but to me, same old Oregon. I do think right now they win the Pac-12 and move to the playoff however. Then they have as even a chance as the other three teams.

Alabama at #2 is too early to call. I have to see where they go at QB and how that works before I am sold on them. Fla and Ole Miss will go a long way in telling me if they are going to be there when it matters.

 

[irishog77]

K State - 55-16 over Stephen F. Austin, 32-28 over Iowa State
Blowout cupcake, scrape by mediocre school.

North Carolina - 56-29 over Liberty, 31-27 over SDSU
Blowout cupcake, scrape by mediocre school.

UCLA - 28-20 over Virginia, 42-35 over Memphis
Scrape by two mediocre schools.

LSU - 28-24 over Wisconsin, 56-0 over Sam Houston
Scrape by mediocre school, blowout cupcake.

Yes, I'd say each of those is about as good as 37-31 over Oklahoma State and 37-12 over Citadel.

Did you actually watch any of these games? UNC faced a 1st and goal from SDSU with a minute to play. SDSU's QB threw a terrible pick to seal the loss. Otherwise, it's a SDSU win. Then there was the Liberty game.

Since when is Iowa State mediocre? They just lost by 20...to a I-AA school. KSU had to rally late in the 4th to win that game...after being down double digits.

UCLA's wins??? Haha! Virginia is MAYBE mediocre and Memphis is dog sh!t. UCLA was EXTREMELY lucky to win either game.

LSU lost a ton of players from last year's team and looked absolutely gross for 3+ quarters against a team that had a QB that made Rick Ankiel look accurate during his mental breakdown saga pitching for the Cardinals.

FSU beat OSU in Dallas (definitely more of a home game for OSU). OSU is a quality football program. It wasn't the toughest team in the world, but there's no shame in not being able to blow out OSU. Did FSU look great against Liberty? No. But they've looked pretty good for 16-18 straight games now. And EVERY team on FSU's schedule is gunning for them. They'll get every opponent's absolute best shot as defending champions and the number one team in the land.

Saying those other teams, particularly the first 3 have looked as, if not more, impressive than FSU is laughable.

 

[irishog77]

Disagree on Oregon. Same old Oregon. Too light on the lines. Mich St will look at the film of that game and get sick to their stomach. I like the Sparties coach, but once they took the lead they missed opportunity after opportunity to close the Ducks out. The QB from State simply missed on two or three big plays to knock the Ducks out. Kudos to the Ducks for hanging on and eventually getting the plays they needed. Good teams do that and Oregon is good. But they are going to struggle when they get to the teams that play ball control, physical running game, and really good d line play. There is a reason why they lost to Stanford in the past. Maybe not this year but to me, same old Oregon. I do think right now they win the Pac-12 and move to the playoff however. Then they have as even a chance as the other three teams.

Alabama at #2 is too early to call. I have to see where they go at QB and how that works before I am sold on them. Fla and Ole Miss will go a long way in telling me if they are going to be there when it matters.

And besides, don't all the pundits, for the last 5 years, say Oregon is the best looking team in the country through week 6-8? Then they lose.

 

[T Town Tommy]

And besides, don't all the pundits, for the last 5 years, say Oregon is the best looking team in the country through week 6-8? Then they lose.

Yep. But LSU, Auburn, and Stanford has showed everyone what the Duck weakness is. Even the Ducks know that and wanted to use the MSU game as a measuring stick as to how well they have developed in the physical play department. They won the game, but didn't pass the physical toughness test in my eyes.

I don't know yet if the Pac-12 is as good as the early season conversations have been. Oregon may be the best they have this year, but I look at a team like Va Tech and B. Foster's defense - who share the same char. as past LSU, Auburn, Stanford teams - and would take the Hokies at almost even money against the Ducks on a neutral site.

 

[greyhammer90]

Can someone explain to me why USC is ranked so high?

 

[wizards8507]

Can someone explain to me why USC is ranked so high?

"Statement win" against Stanford lulz. Because nobody apparently watched that steaming pile of a game.

 

[ACamp1900]

Even SC fans i know that are usually blow hards are pumping the breaks... they all know they got out played.

 

[Whiskeyjack]

SBNation has created a new composite weekly ranking for CFB teams. It combines the AP poll, the Coaches' Poll, F/+ and Sagarin's ELO/ Chess.

We broke just broke the top-10.

 

[wizards8507]

SBNation has created a new composite weekly ranking for CFB teams. It combines the AP poll, the Coaches' Poll, F/+ and Sagarin's ELO/ Chess.

We broke just broke the top-10.

What the hell is the F/+ poll and how is Stanford #5?

 

[Irish#1]

Disagree on Oregon. Same old Oregon. Too light on the lines. Mich St will look at the film of that game and get sick to their stomach. I like the Sparties coach, but once they took the lead they missed opportunity after opportunity to close the Ducks out. The QB from State simply missed on two or three big plays to knock the Ducks out. Kudos to the Ducks for hanging on and eventually getting the plays they needed. Good teams do that and Oregon is good. But they are going to struggle when they get to the teams that play ball control, physical running game, and really good d line play. There is a reason why they lost to Stanford in the past. Maybe not this year but to me, same old Oregon. I do think right now they win the Pac-12 and move to the playoff however. Then they have as even a chance as the other three teams.

Alabama at #2 is too early to call. I have to see where they go at QB and how that works before I am sold on them. Fla and Ole Miss will go a long way in telling me if they are going to be there when it matters.

Agree, if MSU had a heady QB they would have won. I didn't watch a lot of the game, but what I did he was inconsistent and looked a little hesitant.

 

[Whiskeyjack]

What the hell is the F/+ poll and how is Stanford #5?

It's a combination of two separate models: (1) FEI, a drive-based metric; and (2) S&P+, a play-based metric. Those are my go-to sources for advanced opponent-adjusted stats.

 

[IrishLax]

What the hell is the F/+ poll and how is Stanford #5?

Because Stanford is really, really good... as reflected by their efficiency metrics.

 

[wizards8507]

Because Stanford is really, really good... as reflected by their efficiency metrics.

Good for their efficiency metrics. I'm more concerned with wins and losses. Standings aren't based on Defense-Adjusted Value Over Average or Batting Average on Balls in Play. Nonsense focus on "metrics" is why the computer polls were always ridiculous in the BCS and continue to be ridiculous now.

 

[irishog77]

Yes, but coaches (who don't watch games and are biased) and media members (who don't watch games and are biased) are a good method for ranking teams.

 

[GATTACA!]

Can someone explain to me why USC is ranked so high?

Thank you, ive been saying this all year. What justified their preseason ranking either? How where they ranked above us, a team that beat them the last two meetings? Just another scoop of media bias, it's USC they must be great. Ignore the fact that they have a brand new coaching staff.

 

[GATTACA!]

Good for their efficiency metrics. I'm more concerned with wins and losses. Standings aren't based on Defense-Adjusted Value Over Average or Batting Average on Balls in Play. Nonsense focus on "metrics" is why the computer polls were always ridiculous in the BCS and continue to be ridiculous now.

Right lets focus on wins. Like UCLA beating Memphis? Id much rather look at a computer metric than listen to someone try and defend that as a quality win.

 

[ACamp1900]

They're USC guys... the kings of college football.. they are so good in fact that they can actually win National Championship games without even playing in them...

 

[koonja]

Fuck USC.

 

[Whiskeyjack]

Good for their efficiency metrics. I'm more concerned with wins and losses.

Many polls and models are very high on both USC and Stanford, who basically played each other to a stand still. But apparently Stanford must now drop like a rock, because "Ws and Ls!"

I look forward to reading your case for BYU getting a play-off berth when they're undefeated at the end of the season.

Standings aren't based on Defense-Adjusted Value Over Average or Batting Average on Balls in Play.

God forbid there be any objective data involved in this process.

Nonsense focus on "metrics" is why the computer polls were always ridiculous in the BCS and continue to be ridiculous now.

There are bad models, just like there are bad polls. F/+ and Sagarin's ELO/ Chess are two of the very best, and I'd trust those sources far more than a self-interested CFB coach or an AP writer who only watched a couple marquee games over the weekend.

 

[MNIrishman]

I've never understood the disdain for computer rankings. The basic formula goes like this:

1) We come up with standards for what constitutes a good team.
2) We use statistics to determine which teams best meet those standards

This has the advantages of being 1) unbiased 2) really good at measuring network effects (i.e., not all teams play each other) and 3) as accurate as it is possible to be.

Can you imagine a world where the SEC no longer has a free shoo-in to the post-season?

 

[Irish Insanity]

Wait, who's that at #1. Ah, Oregon.

 

[Irish YJ]

I've never understood the disdain for computer rankings. The basic formula goes like this:

1) We come up with standards for what constitutes a good team.
2) We use statistics to determine which teams best meet those standards

This has the advantages of being 1) unbiased 2) really good at measuring network effects (i.e., not all teams play each other) and 3) as accurate as it is possible to be.

Can you imagine a world where the SEC no longer has a free shoo-in to the post-season?

computers are almost as imperfect as humans. some computer rankings are better than others, etc.. don't get me wrong, if the computer output was consistently good, i'd buy in totally. take for instance the current sagarin ratings. FSU at #17... come on. maybe it trues up at the end of the year, maybe not. Last year ND was like 28 to close the season while ASU was like 18 or 19, both ended the year with 4 losses with ND winning the head to head match up. Both of us lost to Stanford, ND a close loss, both ASU losses huge margin.

 

[MNIrishman]

computers are almost as imperfect as humans. some computer rankings are better than others, etc.. don't get me wrong, if the computer output was consistently good, i'd buy in totally. take for instance the current sagarin ratings. FSU at #17... come on. maybe it trues up at the end of the year, maybe not. Last year ND was like 28 to close the season while ASU was like 18 or 19, both ended the year with 4 losses with ND winning the head to head match up. Both of us lost to Stanford, ND a close loss, both ASU losses huge margin.

I think there's a misunderstanding here between what a computer is and what an algorithm is. The computer isn't imperfect, though the algorithm could have flaws. I don't think it would be too tough to come up with a meaningful consensus from the coaches in Div 1 on what constitutes a good team and what constitutes a bad team--this would lead to good algorithm design. At the end of the season, the criteria are applied perfectly via this algorithm, as opposed to humans, who physically cannot watch all of the games and have different qualifications for evaluating what they DO see.

It's no different from any other evaluation method. You agree on what constitutes good performance, then you play the games.

 

[Whiskeyjack]

take for instance the current sagarin ratings. FSU at #17... come on.

Computer models usually become much more accurate toward the end of the season as they gather more data, but even the best ones have to start with something; and that usually ends up being very unscientific pre-season projections, which get filtered out by mid-season.

maybe it trues up at the end of the year, maybe not. Last year ND was like 28 to close the season while ASU was like 18 or 19, both ended the year with 4 losses with ND winning the head to head match up. Both of us lost to Stanford, ND a close loss, both ASU losses huge margin.

ASU made a strong case for being ranked higher than ND last season, even though they lost the head-to-head. FEI, which rated the Sun Devils' SoS as the most difficult in the country, put them at 10th and ND at 23rd. I'm still amazed we won that game.

 

[Irish YJ]

Computer models usually become much more accurate toward the end of the season as they gather more data, but even the best ones have to start with something; and that usually ends up being very unscientific pre-season projections, which get filtered out by mid-season.



ASU made a strong case for being ranked higher than ND last season, even though they lost the head-to-head. FEI, which rated the Sun Devils' SoS as the most difficult in the country, put them at 10th and ND at 23rd. I'm still amazed we won that game.

key word usually on the top half.


bottom half... this is where I think the human brain plays a part. i'd even be OK with one or two spots ahead of ND, but head to head has to weigh heavily, especially when both teams are carrying the same number of losses. also factor in ASU got jail sexed by a 7-5 TTU team that finished the season by losing 5 straight.

 

[Whiskeyjack]

key word usually on the top half.

bottom half... this is where I think the human brain plays a part.

No argument there. I don't think anyone has seriously suggested that we should defer to a computer model entirely. But they're very useful analytical tools because they allow us to objectively compare teams who have played very different schedules, which is invaluable in CFB.

i'd even be OK with one or two spots ahead of ND, but head to head has to weigh heavily, especially when both teams are carrying the same number of losses. also factor in ASU got jail sexed by a 7-5 TTU team that finished the season by losing 5 straight.

This is a great example for why computers are important. What you've just described is probably what goes through the head of most AP voters (or at least those few who take their ballots seriously). But it's placing way too much emphasis on a couple data points, and completely ignoring others.

As I mentioned in my previous post, FEI ranked ASU's 2013 SoS #1, and it's not hard to see why. They had wins against #s 12, 16, 19 and 20. Their only losses came against #2 (Stanford, twice), #23 (us) and #44 (TT). Conversely, FEI ranked our 2013 SoS 14th, with notable wins over #8 and #10, but losses to #2, #14, #36 and #52.

So I don't really have a problem with ASU being ranked higher than us last year, despite our head-to-head win. Their offense was crazy explosive, and they hung a lot of points on some very good defenses.

 

[MNIrishman]

No argument there. I don't think anyone has seriously suggested that we should defer to a computer model entirely. But they're very useful analytical tools because they allow us to objectively compare teams who have played very different schedules, which is invaluable in CFB.



This is a great example for why computer are important. What you've just described is probably what goes through the head of most AP voters (or at least those few who take their ballots seriously). But it's placing way too much emphasis on a couple data points, and completely ignoring others.

As I mentioned in my previous post, FEI ranked ASU's 2013 SoS #1, and it's not hard to see why. They had wins against #s 12, 16, 19 and 20. Their only losses came against #2 (Stanford, twice), #23 (us) and #44 (TT). Conversely, FEI ranked our 2013 SoS 14th, with notable wins over #8 and #10, but losses to #2, #14, #36 and #52.

So I don't really have a problem with ASU being ranked higher than us last year, despite our head-to-head win. Their offense was crazy explosive, and they hung a lot of points on some very good defenses.

I would suggest this. Objectivity and consensus are key. Every voter is biased and underinformed. This is an undeniable fact of the human condition.

I'd be willing to entertain debate on whether we could agree on the algorithm though. The process of producing it could generate the same bias and flaws we aim to avoid. Lots of pressure might happen to introduce an 'SEC bonus' on top of an objective SoS measure, for example.

 

[Irish YJ]

No argument there. I don't think anyone has seriously suggested that we should defer to a computer model entirely. But they're very useful analytical tools because they allow us to objectively compare teams who have played very different schedules, which is invaluable in CFB.



This is a great example for why computers are important. What you've just described is probably what goes through the head of most AP voters (or at least those few who take their ballots seriously). But it's placing way too much emphasis on a couple data points, and completely ignoring others.

As I mentioned in my previous post, FEI ranked ASU's 2013 SoS #1, and it's not hard to see why. They had wins against #s 12, 16, 19 and 20. Their only losses came against #2 (Stanford, twice), #23 (us) and #44 (TT). Conversely, FEI ranked our 2013 SoS 14th, with notable wins over #8 and #10, but losses to #2, #14, #36 and #52.

So I don't really have a problem with ASU being ranked higher than us last year, despite our head-to-head win. Their offense was crazy explosive, and they hung a lot of points on some very good defenses.

I honestly prefer fact and data but I also know algs have flaws as do people. It is my personal belief that head to head results should be heavily weighted as well as outcomes from common opponents. As I said I'm ok with being a few spots behind but certainly not the final spread.

Question - what would the Computers say in 1993 and did you support FSU

 

[MNIrishman]

I honestly prefer fact and data but I also know algs have flaws as do people. It is my personal belief that head to head results should be heavily weighted as well as outcomes from common opponents. As I said I'm ok with being a few spots behind but certainly not the final spread.

Question - what would the Computers say in 1993 and did you support FSU

Depends on the algorithm. I would say a subroutine which places head to head matchups as paramount would be very important, and I think most people would agree with me. In the 1993 case, we won the head to head with the same record. That's pretty unequivocal.