In Defense of The Committee (Or A Reminder of What Could Be Happening Right Now)

B1GTide · Dec 10, 2017

bamacon said:
I’ll just say computers are great for some things. They Cannot however watch tOSU Buckeyes and say “Man, that is one sloppy, overrated team!” The cannot watch Bama/LSU and say, despite, the final score, “those are CLEARLY the two best teams this year and it ain’t close.”

There is something to be said about the human eye. JMHO.

I agree, but a result of the move away from computers is the score escalations that we have seen in the last 4 years. It isn't good enough to win by 14 anymore - you have to kill your opponent. The committee pays attention to margin of victory - something that was largely ignored during the BCS era.

The P5 conferences have not really gotten weaker at the bottom over the last 4 years, but it sure looks that way. Why? Because the teams at the top of those conferences know that they have to win those games by huge margins to get noticed, and getting noticed is a requirement in a committee era.

Still, I am not opposed to the committee. They are 16 for 16 in the first 4 years of the CFP. Really hard to criticize those numbers.

RT27 · Dec 10, 2017

81usaf92 said:
not true.

2002 Rose Bowl

Nebraska was #4 in both the coaches and AP poll, but #2 in the computers.

2004

Oklahoma was #3 in both of the coaches and AP poll but #1 in the BCS

2011

Bama was #3 in the computers, but #2 in the Harris and Coaches poll

MY point is computers are not totally 100% logic, as all computers run on programs written by people. You can write a program to do exactly as you like. MY point is computers are 100% a reflection of the human written program and what it puts as priorities. Just like any human, except it will always be consistent with it's outputs. But depending on who writes the program and the parameters they use, it is not a sure lock for unBIASED answers. If a programmer wants it can factor in pretty much anything they want it to. Just like people do, only difference is it does not change its mind like people do or sway to others views. BUT it by no means assures any better a choice than human polls. And with your above list I wonder who is right or wrong? Just cause one poll say this and one says that how did you determine which one was more correct. When a computer can watch a game and decide between 2 games which one looked BETTER I am all in on a computer. But 2 separate games played to almost exact same stats and final score to a computer is equal. Computers cannot factor some things in that DO make a difference in HUMAN sport. Sorry but computers can not judge some things, only humans can and this is a human game not a computer game.

teamplayer · Dec 10, 2017

BamaInBham said:
Was Wisc pathetic ? They beat 2 mediocre teams: Fla Atlantic, Purdue; 3 decent to good teams solidly (all at home): Northwestern, Iowa, Mich. They lost to OSU by 6 on a neutral field and won the rest of their games. They were kept back in most of the rankings til the end. But it is hard to justify leaving an undefeated P5 champ out. In general I wouldn't mind even if they weren't literally one of the 4 best. Nothing is certain anyway.

They say "best", but IMO, it's a combination of best and most deserving. Sometimes because of an injury return or QB change, etc. a team improves dramatically, though they lost early in the year. But they could be a top 4 "best" team at the end of the year and I wouldn't want them in.

I agree. Wisconsin has a solid team this year. I think they are probably better than or at least equal to the Buckeyes last year and Spartans the year before when those two teams couldn't even score in the playoffs. It seems that OU and FSU have also taken whippings in the playoffs, so I'm not sure based on my eye test that this year's Wisconsin team wouldn't have been equal to those whipping post teams. I think we deserve to be there this year, but if Wisconsin had been 13-0 and Big Ten champs, they would have deserved to be there.

teamplayer · Dec 10, 2017

RT27 said:
MY point is computers are not totally 100% logic, as all computers run on programs written by people. You can write a program to do exactly as you like. MY point is computers are 100% a reflection of the human written program and what it puts as priorities. Just like any human, except it will always be consistent with it's outputs. But depending on who writes the program and the parameters they use, it is not a sure lock for unBIASED answers. If a programmer wants it can factor in pretty much anything they want it to. Just like people do, only difference is it does not change its mind like people do or sway to others views. BUT it by no means assures any better a choice than human polls. And with your above list I wonder who is right or wrong? Just cause one poll say this and one says that how did you determine which one was more correct. When a computer can watch a game and decide between 2 games which one looked BETTER I am all in on a computer. But 2 separate games played to almost exact same stats and final score to a computer is equal. Computers cannot factor some things in that DO make a difference in HUMAN sport. Sorry but computers can not judge some things, only humans can and this is a human game not a computer game.

The computers are just for crunching numbers. If the data being used to crunch the numbers is based on human polls, a computer ranking is just a compilation of human rankings. The computer crunches numbers without bias, which is what many people, including myself, fear about the committee. I will admit that the committee has done fine so far; I just don't think they are needed.
Edit- Perhaps I also say this because I am envious of their position and wish I were on the committee. Who knows?

selmaborntidefan · Dec 10, 2017

However.....THIS is the far more likelier scenario we would have just endured (grab your socks and hold on).....

AP RANKING (probably would have been PRIOR to last week)
1) Wisconsin
2) who knows? Most likely Clemson or Oklahoma
3) Oklahoma or Clemson
4) Auburn
5) Georgia
6) Miami
7) Alabama

And if Wisky had beaten Ohio State........in the old pre-BCS era......we would have had this insanity:

1) Wisconsin - in the Rose Bowl vs USC
2) Clemson or OU -
3) Clemson or OU -

And depending on the time frame in discussion, either: a) they'd have played separate bowls or b) played each other, and the hype would be "if Wisconsin loses to USC, the winner of this game should win the championship."

4) Georgia - who would immediately start pouting about how their loss to Auburn was better than the two team's above them's losses

And Georgia would PROBABLY get Ohio St in the Sugar Bowl, too.

5) Alabama -

Now here is where it gets interesting. If Clemson and OU did not play each other then there would be pressure for which ever team was ranked #2 to play us......and then we'd be in a somewhat similar situation to 1965 or 1977.

And btw - under BOTH the old polls system AND the BCS system, Ohio St would not even merit a mention.

There would be a TON of screaming about:
a) how we need a playoff of the top 8 teams (sound familiar)
b) how Alabama is overrated because they finished third in the SEC and it wasn't even that good this year (sound familiar)
c) Wisconsin had not played anybody (sound familiar)
d) why these bowl games are awful when we could just match up the 'right teams' (sound familiar)

What I'm saying is this - the VERY SAME complaints we are hearing NOW would get prominence because I lived through those and it happened back then. The difference now is that if you win your games or lose just once to a decent foe, you still have a shot.

Look at the 1985 final regular season AP poll:

1) Penn State - indy, unbeaten
2) Miami - one loss to Florida, indy
3) OU - one loss to Miami, Big 8
4) Iowa - one loss to Ohio St, Big 10 champion

In the UPI (now coaches) poll:
1) Penn State - indy
2) OU - who lost to Miami, wth...??
3) Miami
4) Iowa

Penn St got to pick what they wanted. Because OU was ranked higher in BOTH polls when bowl invitations were made (in those pre-www days), Penn St accepted the Orange Bowl to get a 1 vs 2. Miami took #8 Tennessee in the Sugar Bowl because:
a) #4 Iowa was obligated to the Rose
b) #6 Florida was bowl ineligible
c) #5 Michigan was selected by the Fiesta Bowl (no conference tie-in at all) to play....
d) #7 Nebraska (back then if you could get Mich-Nebraska, by gawd, you took it)

So Miami got an 8th ranked Vols team that proceeded to beat holy be-jeezus out of them.

Back in 1985, there was a HUGE cry for "a four-team playoff." In fact, I received "The Sporting News" back in the day, and there was a huge write up about how controversial ANY outcome other than Penn State beating OU (they didn't) was going to be.

There is - LITERALLY - nothing new about the complaints last week except one thing......we now DO have a remedy for the situation. Btw - nobody would dare have argued in 1985 that "you have to win your conference." Look at the national champions from a brief period of time:

1982 - Penn State (Ind)
1983 - Miami (Ind)
1984 - BYU (WAC)
1985 - OU (Big 8)
1986 - Penn St (Ind)
1987 - Miami (Ind)
1988 - Notre Dame (Ind)
1989 - Miami (Ind)

Notice anything???????

In fact, except for 1984 (which I'll write about over the holidays), EVERY SINGLE YEAR at least ONE Independent of conference team played for the title.

So........what has changed is that we no longer have four top teams where "you know, we really can't tell them apart"......but only two get the hype.

RT27 · Dec 10, 2017

teamplayer said:
The computers are just for crunching numbers. If the data being used to crunch the numbers is based on human polls, a computer ranking is just a compilation of human rankings. The computer crunches numbers without bias, which is what many people, including myself, fear about the committee. I will admit that the committee has done fine so far; I just don't think they are needed.
Edit- Perhaps I also say this because I am envious of their position and wish I were on the committee. Who knows?

LOL I get that, the cool thing is it allows for years of discussions, as there is no one correct answer to this stuff. Just as I get to thinking one way some one posts a view I have not considered and I do a 180, it is maddening. I say take top 10 from polls then pick numbers out of a hat for top 4 bahahahahaha No matter how you do it 1-2 teams are always going to say they were left out for some reason.

selmaborntidefan · Dec 10, 2017

teamplayer said:
I agree. The committee is just a joke due to the fact that it isn't needed. They act like it is such a tough job and try to explain all types of information each week. Give me a break. The BCS also got it right most of the years, and only a few times did people think the third ranked team had an argument. If that's the case, and I think it is, why didn't they just use the same formula and extend it to four teams? The BCS really counted losses more than this eye test thing people try to do now. Well, I think they started counting losses more after OU got absolutely hammered in their conference championship game and still made the national title game. Anyway, I'm not sure why I keep talking about it because it is clear they plan to continue using the committee. I guess one of the reasons I love Tidefans is so I can put my two cents in with everyone else.

I can explain this one very easily. Just look at the BCS poll from November 21, 2010:

1) Auburn
2) Oregon
3) TCU
4) Boise St

THAT'S why they do it.

If Colin Kaepernick had not led Nevada to an upset win over Boise, guess what? If we had a four-team playoff with the BCS formula then TWO of the four teams would have been Boise St and TCU.

There. Is. NO. WAY. They. Were. EVER. Letting. THAT. Happen.

That's the REAL reason why. It's not - as 1/2 this board kept whining in 2012 - "to keep 2 SEC teams out." It's to keep two teams that have no business being there out.

81usaf92 · Dec 10, 2017

RT27 said:
MY point is computers are not totally 100% logic, as all computers run on programs written by people. You can write a program to do exactly as you like. MY point is computers are 100% a reflection of the human written program and what it puts as priorities. Just like any human, except it will always be consistent with it's outputs. But depending on who writes the program and the parameters they use, it is not a sure lock for unBIASED answers. If a programmer wants it can factor in pretty much anything they want it to. Just like people do, only difference is it does not change its mind like people do or sway to others views. BUT it by no means assures any better a choice than human polls. And with your above list I wonder who is right or wrong? Just cause one poll say this and one says that how did you determine which one was more correct. When a computer can watch a game and decide between 2 games which one looked BETTER I am all in on a computer. But 2 separate games played to almost exact same stats and final score to a computer is equal. Computers cannot factor some things in that DO make a difference in HUMAN sport. Sorry but computers can not judge some things, only humans can and this is a human game not a computer game.

Okay but what is the committee judging? Is it the best teams or the most deserving teams, and where is the criteria that they are judging who is in and who is not? The truth is that we have had two different committees and one favored championships over best teams and the second favors best teams over champs. So there is inconsistency. If you take the BCS computers and use that method there is not any inconsistency because it’s multiple computers averaging out the rankings. Humans pick and choose what they want while the computers try to find the average.

So the best way to do this is 2/3 human and 1/3 computers.

selmaborntidefan · Dec 10, 2017

I watched Colin Cowherd (I know, I know) DEFEND our 2011 selection for the LSU rematch by saying that we would win the game and he made a valid point - we've ALWAYS used the "eye test" to separate teams and then the record is just used to vindicate our findings.

Does ANYONE here - anyone - actually think that Tulane was robbed in 1998? Not even their own coach (Tommy Bowden) thought that. We took one look at K-State, Tennessee, and UCLA and we sorta set a pecking order. We agreed FSU was up there with them. The hunch was that K-State or Tennessee was probably the best team in the nation. And, in fact, they would have played except K-State blew a lead in the Big 12 title game.

We can usually tell who is the best or who it's between.

teamplayer · Dec 10, 2017

selmaborntidefan said:
I can explain this one very easily. Just look at the BCS poll from November 21, 2010:

1) Auburn
2) Oregon
3) TCU
4) Boise St

THAT'S why they do it.

If Colin Kaepernick had not led Nevada to an upset win over Boise, guess what? If we had a four-team playoff with the BCS formula then TWO of the four teams would have been Boise St and TCU.

There. Is. NO. WAY. They. Were. EVER. Letting. THAT. Happen.

That's the REAL reason why. It's not - as 1/2 this board kept whining in 2012 - "to keep 2 SEC teams out." It's to keep two teams that have no business being there out.

You make a very real point here, but most of the time when those types of teams feel the pressure, they crack. If they knew they were close to making the playoff, those types of teams would probably choke. It seems to happen all of the time. I would still say the end of the year BCS top four would be a fair playoff without all of the fake drama of the committee selection.

BamaInBham · Dec 10, 2017

Here is a good example as to why we don't want computers as an official component of any ranking system. Again, yes, they can be consulted but not used as a fixed part of any ranking. The following computer rankings are from the 6 that were used in the last, 2013, BCS ranking. Here are a couple of notes about their final rankings for this year.

UCF was #3 in 2 of the rankings and #1 in another. I.e., 3 of the 6 had them in the playoff; more than had Bama or OSU in the playoff and as many as had OU in.
Wisconsin was included in 2 of the 6 rankings, as many as Bama and OSU.
Bama was #1 in one of the only 2 polls that included them in the playoff - Sagarin.
OU made it in only 3 polls
OSU made it in 2 polls
Bama was ranked from 1-7, OU from 3-7, UCF from 1-16, Clemson and UGA had the smallest range of the participants 1-3. So, you can see that the computers have different criteria, demonstrating that they are not some magical source of impartial wisdom, but, repeating myself, they are nothing but the inflexible expression of a particular individual's or group's thinking.

Here is the link: http://www.colleyrankings.com/foot2017/bcsLike/bcsLike14.html

teamplayer · Dec 10, 2017

Those are also very fair points.

81usaf92 · Dec 10, 2017

BamaInBham said:
Here is a good example as to why we don't want computers as an official component of any ranking system. Again, yes, they can be consulted but not used as a fixed part of any ranking. The following computer rankings are from the 6 that were used in the last, 2013, BCS ranking. Here are a couple of notes about their final rankings for this year.

UCF was #3 in 2 of the rankings and #1 in another. I.e., 3 of the 6 had them in the playoff; more than had Bama or OSU in the playoff and as many as had OU in.

Wisconsin was included in 2 of the 6 rankings, as many as Bama and OSU.

Bama was #1 in one of the only 2 polls that included them in the playoff - Sagarin.

OU made it in only 3 polls

OSU made it in 2 polls

Bama was ranked from 1-7, OU from 3-7, UCF from 1-16, Clemson and UGA had the smallest range of the participants 1-3. So, you can see that the computers have different criteria, demonstrating that they are not some magical source of impartial wisdom, but, repeating myself, they are nothing but the inflexible expression of a particular individual's or group's thinking.

Here is the link: http://www.colleyrankings.com/foot2017/bcsLike/bcsLike14.html

You do realize the BCS threw out your highest and lowest rankings and averaged the other 4. So in all likelihood we still get the same result just different seeds.

The point is there is an unbiased failsafe in case the human element turns biased. We never know what a committee is going to say, and before you say that we do. Think about the week after the iron bowl and how Hocutt compared Bama to tosu and how he did on selection day. How do you go from “ razor thin” to “unequivocally better” in a matter of 5 days? The truth is ESPN has taken the Eric Biscoff theory of entertainment that states “ controversy =cash”. The problem is that there is no real criteria from committee to committee.

selmaborntidefan · Dec 10, 2017

teamplayer said:
You make a very real point here, but most of the time when those types of teams feel the pressure, they crack. If they knew they were close to making the playoff, those types of teams would probably choke. It seems to happen all of the time. I would still say the end of the year BCS top four would be a fair playoff without all of the fake drama of the committee selection.

The Power 5 isn't going to take the chance that you're wrong, though. There's too much money involved.

BamaInBham · Dec 10, 2017

81usaf92 said:
You do realize the BCS threw out your highest and lowest rankings and averaged the other 4. So in all likelihood we still get the same result just different seeds.

The point is there is an unbiased failsafe in case the human element turns biased. We never know what a committee is going to say, and before you say that we do. Think about the week after the iron bowl and how Hocutt compared Bama to tosu and how he did on selection day. How do you go from “ razor thin” to “unequivocally better” in a matter of 5 days? The truth is ESPN has taken the Eric Biscoff theory of entertainment that states “ controversy =cash”. The problem is that there is no real criteria from committee to committee.

My point was not that this particular group of computer polls would yield the right or wrong result, but to show that computers are not some magical source of impartial logic. We see the tremendous difference in just 6 sets of computer logic.

In regard to your speculation, even if correct, another set of 6 polls may be worse than this. What if we whittled it down to 4 of these 6, or expanded to 8, we could get a perverted result. Just think now, 3 of 6 chose UCF, with one ranking them #1, only 2 of the 6 chose Bama and OU. I did calculate the total number of points, throwing out the high and low and OSU would have been in and Bama out. But Bama, OSU, Wisc and UCF were within 3 or 4 points of one another - a razor thin margin. I.e., it would take almost nothing for that result to be dramatically changed, so it was just luck that it turned out almost right.

But again, my primary point was simply to show how different the polls are, as well as demonstrating how far off they can be. Again, anyone who knows much about software development knows that we don't want computers as anything more than a consultant. In reality, it's the human polls that keep the computers in line, not vice versa.

I agree that the SC's comments are often contradictory and sometimes just wrong. But that is not necessarily a reflection of the judgment used in the selection process, just unfortunate communication.

In this enterprise, give me human logic every time. I even prefer the media poll to computers and much prefer the Selection Committee with all of their subconscious biases and imperfect knowledge and logic. Even if those who contribute to the computers' algorithm are unbiased, it does not make up for their stupidity or better stated their inflexibility, their incapacity for nuance. Computers are not what many think.

BamaInBham · Dec 10, 2017

selmaborntidefan said:
The Power 5 isn't going to take the chance that you're wrong, though. There's too much money involved.

True. And it just would not be right to include UCF this year, like half of the computer polls would have yielded, a team that played two P5 opponents. The country would have been in an uproar - it would have been maddening - Bama played 9 P5s and OU 10. Also, only 2 of 6 computers included Bama and OU, the same as Wisc.

81usaf92 · Dec 10, 2017

BamaInBham said:
Here is a good example as to why we don't want computers as an official component of any ranking system. Again, yes, they can be consulted but not used as a fixed part of any ranking. The following computer rankings are from the 6 that were used in the last, 2013, BCS ranking. Here are a couple of notes about their final rankings for this year.

UCF was #3 in 2 of the rankings and #1 in another. I.e., 3 of the 6 had them in the playoff; more than had Bama or OSU in the playoff and as many as had OU in.

Wisconsin was included in 2 of the 6 rankings, as many as Bama and OSU.

Bama was #1 in one of the only 2 polls that included them in the playoff - Sagarin.

OU made it in only 3 polls

OSU made it in 2 polls

Bama was ranked from 1-7, OU from 3-7, UCF from 1-16, Clemson and UGA had the smallest range of the participants 1-3. So, you can see that the computers have different criteria, demonstrating that they are not some magical source of impartial wisdom, but, repeating myself, they are nothing but the inflexible expression of a particular individual's or group's thinking.

Here is the link: http://www.colleyrankings.com/foot2017/bcsLike/bcsLike14.html

That also isn't entirely correct because its using the AP. The AP hasn't been a part of the process for sometime. Lets use what would really be used in today's football. Lets use the CFP, USA, and computers.

First lets use the computers alone. Take into account they factor SOS, SOR, and different methods per computer. Also keep in mind your highest and lowest ranking is dropped and the other four are averaged out.

Computer rankings

1. Clemson 1.5
2. UGA 2
3.tOSU 4
4.OU 4.25
5. UA 5.25
6. Wisc 5.5
7. UCF 6
8. USC 8.25
9. Auburn 9.25
10. PSU 9.25
11. Miami 11.5
12. WU 13.75

Yes the computers don't favor us, and that's why they shouldn't be an end all be all ranking. But they are consistent in their rankings in November. They aren't switching favorites in order to produce a certain matchup of teams. If they are used as a failsafe to prevent a rogue coach (like Bob Stoops in 2011) and a potentially biased committee full of playoff hopeful ADs from doing so then it is a very effective tool in deciding who goes.

Here is how they would end up.

CFP USA C Divide by 3= answer
1. Clemson #1 + #1+ 1.5 / 3 = 1.2
2. UGA #3 + #3+ 2 /3 = 2.6
3. OU #2+ #2+ 4.25 /3 =2.75
4. Bama #4+ #4+ 5.25 /3 = 4.42
5. tOSU #5+ #5+ 4 /3 = 4.7
6. Wisconsin #6+ #6+ 5.5 / 3 = 5.8
7. USC #8+ #7+ 8.25 /3 = 7.75
8. Auburn #7 + #8+ 9.25 /3 = 8.1
9. PSU #9+ #9+ 9.25 /3 = 9.08
10. UCF #12 + #10+ 6 /3 = 9.3
11. Miami #10 + #11+ 11.5 /3 = 10.8
12. WU #11 + # 12+ 13.75 /3 =12.25

The point is that the computers show where there is bias in the human polls, and prevents big shifts week to week because it puts it on the human voters to either be unbiased or get exposed.

drwho · Dec 10, 2017

81usaf92 said:
Lets use the CFP, USA, and computers.

Which computers are using? Sagarin, Colley, ???

81usaf92 · Dec 10, 2017

BamaInBham said:
My point was not that this particular group of computer polls would yield the right or wrong result, but to show that computers are not some magical source of impartial logic. We see the tremendous difference in just 6 sets of computer logic.

.

read my above post

BamaInBham said:
In regard to your speculation, even if correct, another set of 6 polls may be worse than this. What if we whittled it down to 4 of these 6, or expanded to 8, we could get a perverted result. Just think now, 3 of 6 chose UCF, with one ranking them #1, only 2 of the 6 chose Bama and OU. I did calculate the total number of points, throwing out the high and low and OSU would have been in and Bama out. But Bama, OSU, Wisc and UCF were within 3 or 4 points of one another - a razor thin margin. I.e., it would take almost nothing for that result to be dramatically changed, so it was just luck that it turned out almost right.
.

1) not speculation because the BCS operated under the idea of taking away the Highest and lowest rankings then averaging out the rest.
2) who cares if it keeps the human voters in check because I know for a fact 90% of Bama fans were sweating bullets after tOSU won because Hocutt said "razor thin" difference 4 nights before when Bama was #5 and TOSU was #8. In the BCS 4 place jumps were about impossible on championship weekends

BamaInBham said:
But again, my primary point was simply to show how different the polls are, as well as demonstrating how far off they can be. Again, anyone who knows much about software development knows that we don't want computers as anything more than a consultant. In reality, it's the human polls that keep the computers in line, not vice versa.
.

How about 1983, 1984, and 1990.

Since then the only issue with the true champ was in 03

BamaInBham said:
In this enterprise, give me human logic every time. I even prefer the media poll to computers and much prefer the Selection Committee with all of their subconscious biases and imperfect knowledge and logic. Even if those who contribute to the computers' algorithm are unbiased, it does not make up for their stupidity or better stated their inflexibility, their incapacity for nuance. Computers are not what many think.

So you would fully go with a committee that changes the rules every two years? What if the committee said "UCF" just because they couldn't decide between Bama, USC, and tOSU? The computers keep the human pollsters in check plain and simple because otherwise you have so many sham champs.

81usaf92 · Dec 10, 2017

drwho said:
Which computers are using? Sagarin, Colley, ???

A&H = Anderson & HesterBill = Billingsley College Football Report
Colley = Colley Matrix Rankings
Massey = Massey Ratings
Sagarin = Jeff Sagarin Rankings
Wolfe = Peter Wolfe Rankings

Take the Highest and lowest out and then average them out

In Defense of The Committee (Or A Reminder of What Could Be Happening Right Now)

TideFans Legend

All-American

Hall of Fame

Hall of Fame

TideFans Legend

All-American

TideFans Legend

TideFans Legend

TideFans Legend

Hall of Fame

All-American

Hall of Fame

TideFans Legend

TideFans Legend

All-American

All-American

TideFans Legend

Suspended

TideFans Legend

TideFans Legend

TideFans.shop - NEW Stuff!