"Spend a few minutes reading College Football Resource" - Whit Watson, Sun Sports

"Maybe you should start your own blog" - Bruce Feldman, ESPN

"[An] Excellent resource for all things college football. It’s blog index is the definitive listing of the CFB blogosphere ... [A] must-read for fans." - Sports Illustrated (On Campus)

"The big daddy of them all, the nerve center of this twisted college football blogsphere" - The House Rock Built

"Unsurprisingly, College Football Resource has generated some discussion" -Dawg Sports

Top Teams 2008

After Week Seven

  1. Alabama
  2. Penn State
  3. Texas
  4. Oklahoma
  5. Florida
  6. USC
  7. Georgia
  8. LSU
  9. BYU
  10. Missouri
  11. Ohio State
  12. Oklahoma State
  13. Texas Tech
  14. Utah
  15. Kansas
  16. USF
  17. North Carolina
  18. Miami
  19. Boise State
  20. Georgia Tech
Display
RSS
Search CFR
Submission Corner
« Here's a start | Main | This week's games, week nine »
Wednesday
Oct262005

The myth of records, and being undefeated

Is being undefeated the standard by which we judge championship teams now in college football?  If it is, I think we're being a bit strict with our rules.  In my eyes, undefeated seasons don't happen in a vacuum.  Teams have to be very good, but they also have to be very lucky.

Look at USC last year.  They survived a last-minute type scares against Stanford and California.  Or look no further than the 2002 Ohio State Buckeyes, who must have set some kind of record for close escapes.  Rarely is a schedule so light that a good team can skate right through without several scares (Texas 2005 aside).  Sometimes schedules are set up in such a way that a very good team often can't finish undefeated, yet some other team with a little lighter slate and a little better luck earns the spotless record.

But if that one or two loss team appears to be better, why are they not treated as better.  There's some kind of reflex in all of us to say "but, team X is undefeated!" in response.  I understand that, but sometimes that reflex is wrong.

This argument doesn't even have to be about undefeateds, but it can be about how we regard all teams.  So often the polls just slot teams by loss.  Is that accurate?  I'm not always so sure.  This week I have Notre Dame, they of two losses, at number four in my rankings, ahead of undefeateds Georgia and UCLA (as well as Alabama and one loss Miami).  I think Notre Dame's simply better than those teams, but a rough early slate that saw them barely lose to a hot Michigan State offense and a juggernaut USC team by the narrowest of Margins has them with two losses.  But I see beyond that.  I think their team found itself and as they currently stand, I think they'd beat nearly everyone ranked behind them.

In fact, the Irish remind me a bit of the 2002 USC squad.  Those Trojans lost early to Kansas State and Washington State before they really became the team they are now, and finished the season with eight straight wins.  At the end of the year there was very little doubt in my mind they wouldn't have beaten both Miami and Ohio State.  But in fact, they finished just #4 in the polls and BCS.  That situation made me realize that how we have been going about this ranking process is egregiously flawed.

At the bottom of it, I think its because there are conflicting demands placed on us as to how we should view success in college football.  On the one hand, we want to be as reasonably accurate as possible, and earnestly take into account all the factors that go into a team's season, from player health, coaching, schedule, rivalries and a host of other things.  But another demand is to consider the season as the be-all, end-all, and knock teams for losses, because we don't have a playoff.  That is to say, using the NFL as comparison, that whereas when an NFL team loses, its season isn't over, with a college team, they can forget about the big prize at the end of the season.

I guess I'm arguing that we should maybe slow things down and recognize teams for their inherent worth, and judge them as such.  Maybe college football's Patriots lost a game somewhere this year.  Shouldn't we still recognize such a hypothetical team as being of that quality?  This is a hypothetical, of course, as I don't think this year's "Patriot" comparison (USC) has lost a game, but if that were to happen to USC, or someone down the road, I'd hope that we might get it right and still recognize such a team as a contender.

It did in fact happen in 2003 when USC lost to California, but only through a similar fate happening to Oklahoma and LSU in their seasons, did USC even have a chance at being considered championship worthy.  To me this is an egregious reflection on how badly we've gotten in determining a team's worth of late, to nearly leave that kind of team out in the cold because what appeared to be at least one, if not two inferior teams ahead of it had the fortune of surviving their entire schedule.

Taking this model, and applying it to the real world, I offer you this: right now, as I see it, and unless things change fundamentally for these teams, I'm pretty content with going USC 1, VT 2, Texas 3.  That's just how I view things, right or wrong.  If Texas had lost last weekend, I'm not sure I would have moved them down.  I'd obviously have had to observe how the game played out and if something fundamentally wrong with them had occurred that had not been there before.  If not, I'd very likely have kept my ranking as such.  The same goes for USC and Virginia Tech.  Obviously between now and then, if that were to happen, I can change my mind, but I do my best to view these teams relative to each other, not solely in the context of their individual records, which can be deceptive. 

This plays into other things that concern me, namely schedules.  I'll get into that later, but to be brief on here, I think its funny when people talk about Texas and their weak record.  Normally, I'm a big proponent of difficult out-of-conference schedules.  I think it shows a commitment by an athletic department by having a serious OOC slate to expose its team to different situations and not just duck teams so that their program has a shot at a sneaky undefeated schedule.  But, sometimes soft schedules will damn a team, as may happen to Texas this year.  I think out of fear people rank teams with similar records by who they have played.  This is a flawed approach, because it takes no interest in the fundamental strength of each team compared to the other.  Schedules allow us to see how teams react to various kinds of opponents, their styles of play, their talent, their coaching, and how a team plays on the road.  But judging a team solely on its schedule is particularly flawed.

I happen to think Texas is the third best team behind Virginia Tech right now, but its not because of their schedule.  I trust my eyes in observing their games, and the things I've learned about football over the years, to make that conclusion.  However, when the final BCS rankings come out and somehow USC, Texas and Virginia Tech are all undefeated, Texas may get the heave-ho because their schedule was viewed the lightest by the computers.  This would be a mistake.  The BCS will have gotten it right, but for the wrong reasons.  It's not to dissimilar to what happened in 2003, when USC was left out of the BCS championship game because the computers felt their schedule was lightest.  In reality, USC was a much better team than the two that finished ahead of them, but because the BCS could not take into account what USC was worth, compared to what Oklahoma and LSU were worth, it got things wrong.  Just the same, when at the end of the year, if we look at those three teams and just compare schedules, we'll have gotten it wrong.

PrintView Printer Friendly Version

EmailEmail Article to Friend

Reader Comments (23)

RA, mostly agree with you on this. I think one thing that could help a lot is that we have NO preseason polls. Let all the polls and rankings be established after say, (5) games are played. That way a team wouldn't be penalized because they started out ranked #16 etc. and, the strength of schedule numbers might be a little more accurate. The old adage of "it's better to lose early" might not be as much of a factor?

Also, if you trust your eyes, take another look, you might of had something in one of them when you looked at Notre Dame.
October 27, 2005 | Unregistered CommenterDawgy1
The problem is the word "think". I think team X is better than Y. I think Team X would beat these other teams. Thinking is opinion. Period. And thinking something doesn't make it so.

The thing that I love about true sport is that it's objective. A guy makes it to the finish line first. Team X score more than Team Y. There is a result, and it's unimpeachable. College Football doesn't need to be different. There is an objective task every team is asked to perform. Win every game you play. To me, that should be the first and foremost consideration.

I cannot see a possible situation where an undefeated team (limit it to the BCS conferences if you like) doesn't play in the title game behind a one-loss team. You might "think" that one-loss team is better, but they failed to accomplish the task before them. If another team has accomplished it, they deserve to play for the title. And at the end of the day, all it is is what you think, not what you know.

I've said it before, but when we get subjective in talking about who's better, college football becomes figure skating.

October 27, 2005 | Unregistered CommenterLD
LD, if we could have say a (4) or (8) team playoff it would resolve most of this issue.

There's a playoff system in every college sport except for D-1 football. As long as we have this system there's going to be a lot of things decided on what someone "thinks".

I mean with a play-off using the top (8) ranked teams in the BCS you would probably have 2-3 teams arguing that they should have been in the playoffs but, those teams probably wouldn't be claiming that they could have won it all.
October 27, 2005 | Unregistered CommenterDawgy1
I'm not one to beg for a playoff, since it isn't happening, for at least a decade. I'm for making the system we are stuck with better by removing as much of the opinion based decisionmaking as possible. As everyone knows, opinions are like... everyone's got one and they all stink.

My issue is with the title of the post. "The Myth of Records." A team's record is the only objective fact that anyone can lay claim to when it comes to who is better than anyone else. It's no myth. It's what's real. If you say a teams undefeated record isn't as good as another teams two loss record, your claim isn't rooted in fact, but only opinion. The burden has to be high to show otherwise.
October 27, 2005 | Unregistered CommenterLD
LD,

You HAVE to be subjective about deciding which team is better if the two teams in question have not played each other. There isn't any truly objective approach that is accurate in that case. Even common opponents can't be used as the exclusive measure (as demonstrated on a thread a month or so ago, it can easily create a paradox).

If team X doesn't have any tough games and wins out, and team Y plays the toughest schedule in the country and winds up dropping a close one on the road against a really good team, can we really conclusively say that team X is better than team Y?

It certainly doesn't work to just use losses as the standard. What if you have a 1 loss team whose only loss is to a team that played a much tougher schedule and happened to drop two games? Head to head, the two loss team won, but in order of losses, the one loss team obviously has fewer.

It's impossible to have a completely objective standard unless you have two teams with reasonably comparable schedules who play head to head.
October 27, 2005 | Unregistered CommenterGinn Fan
According to the Sagarin ratings, Texas' schedule strength so far is ranked 30th.

USC's is ranked 24th.

Virginia Tech's is ranked 54th.

October 27, 2005 | Unregistered CommenterCS
I think probably VT will have the best SOS at the end of the year. They've got BC, Miami, and probably FSU to go. If they run that table I'll be a believer!
October 27, 2005 | Unregistered CommenterDawgy1
I agree that VT's strength of schedule will be the strongest at the end of the year. That being said, Texas currently has played a harder schedule, so criticisms of the Longhorns skating by are unwarranted.

A quick thought: given two undefeated teams does the team with the harder schedule automatically warrant a better ranking? Can't the relative strength of the teams be gauged by perhaps watching some actual football and comparing the merits of the teams?
October 27, 2005 | Unregistered CommenterCS
Ginn Fan: Of course you have to be subjective if you're going to make statements like that. I subscribe to a different school. I think it was Bill Parcells who said a while back about how "you're usually about as good as your record." I believe if a team doesn't lose, you can't impute the idea that they would've. I prefer to rely on facts.

"You HAVE to be subjective about deciding which team is better if the two teams in question have not played each other."

No, you don't. You can say that the teams are as good as each other, since there is no evidence to say otherwise. If two teams haven't played each other, and both are undefeated, I don't think you have to decide that one is definitely better than the other. In fact, I don't think anyone CAN know, unless they play each other. I don't know if Utah would've beaten SC last year, and I'm not going to assume they definitely would not have.

If we do not settle it on the field, I don't see why we have to use style points or anything else.

And your example of the one loss team who loses to a two loss team: who is to say that the two loss team hasn't lost to two teams the one loss team has beaten? All I know is that at the beginning of each season, every team looks at their schedule and sees 11 or 12 games and they know that their goal is to win each one. Period. If one team does a better job at winning every game than another, I think that proves that it's a better team.

What I really don't like is the way this seems like an argument for keeping SC or Texas in one of the top two slots ahead of an undefeated BCS-conference team beacuse "you're just SURE that they'd still win." Coffee is for closers. Championships are for winners. Win all your games and you can claim a title. Lose one and try to rely what some dude "thinks"? Weak.
October 27, 2005 | Unregistered CommenterLD
Also, I've looked at all the methodologies of SOS, and I think Colleyratings does the best job. They run the program continually with shifting data until it no longer changes (better explanation at their site).

Texas: 26
Alabama: 54
Georgia: 64
VT: 70
USC: 72
UCLA: 85

October 27, 2005 | Unregistered CommenterLD
Alabama, Georgia, and VT are going to take a jump in the next few weeks.
October 27, 2005 | Unregistered CommenterDawgy1
LD,

I guess if you're limiting it strictly to people analyzing games (i.e. limiting this to media members without any votes in any of the polls), then saying two teams are equal is fine. But, if you're voting in the polls, don't you have to vote somebody #1 and somebody #2? And isn't what we're really talking about here the way we think the rankings should be?

As far as the "Win out if you want to win a championship," you're right. Championship teams should win out, but we know that doesn't always happen (i.e. 2003-04 LSU and USC).

I guess what it boils down to is two different schools of thought: You'd like to see a national championship game with the two teams with the best records, regardless of anything else; whereas I would prefer to see the two teams that I think are truly the best teams, even if one of them has a loss.

Is that the correct way of stating what you would prefer? That's my impression of your take, but I could be wrong.

Limting it just to BCS conferences, would you have been comfortable with West Virginia playing in the National Championship Game if they had managed to run the table in this year's Big East and, say, USC had been stopped on Leinart's QB sneak on the last play against Notre Dame and lost the game, and Texas ended up being the only other undefeated team?

To me, the Big East is TERRIBLE, and SC, the two-time defending national champs, would clearly give Texas a better game. But, one 3 point loss (that was inches from being a victory) in a hostile environment to a top-10 team would keep them out. Personally, I'd want to see Texas-SC.
October 27, 2005 | Unregistered CommenterGinn Fan
Ginn Fan: I think your summary of our viewpoints is about right. I guess my view is that a record is the objective statement of how good a team is, so matching the two best records, is in my view, matching the two best teams.

About ranking teams in polls, I guess I don't care all that much about them. They're a burden on college football - the intrustion of the subjective. I'd like championships to be decided on the most objective and known-to-the-teams-beforehand manner. Where teams know if we do X, we win. As it is today with polls, there is too much subjectivity and a team can do everything it's supposed to do and still be left out in the cold.

Now, when dealing with multiple one-loss teams, I have more trouble determining champions. But I honestly don't have a problem with a team that hasn't been beaten claiming a title. Up until the 70's there were multiple titles awarded each year. It didn't make college football all that much worse. If we're not going to settle it on the field, give out multiple trophies.

As for your WVA-USC hypo, I'd say, yeah, why not let WVA play for the title? If nobody had beaten them (and that would include the Va. Tech OOC game), who is to say that anybody would've. Again, just because you think something is bad doesn't make it so. And for USC losing a close game, I'd say it sucks if they're really great, but it's why you don't lose. Great teams don't (and USC is a prime example of that). The argument here is that I DON'T know if anyone would beat an undefeated WVA team. I DO know that someone has beaten a one-loss USC team. I won't presume to know things that haven't happened, and I won't assume that a team which hasn't been beaten would be. I'll wait for the evidence.

To me, it's a philosophical approach. I believe in objectivity in sport. Yes, I have my own opinions about who's better than who, but in the end, it doesn't mean much - and someone else may just have a diametrically opposed opinion.
October 27, 2005 | Unregistered CommenterLD
LD,

Sounds like it is just a case of different schools of thought. I wouldn't say that having WVA in my hypo in the NC game was bad or wrong, just that I'd prefer to see the SC-Texas matchup, and I think it would be more relevant in determining who the champ is. WVA would certainly be an easy choice to be in the championship game though, I acknowledge.
October 27, 2005 | Unregistered CommenterGinn Fan
The original question was, "Is being undefeated the standard by which we judge championship teams now in college football?" Sure it is, when has it ever been different? LD is on to something. This is a thinly veiled attempt to begin an argument for USC playing in the BCS title game if they lose a game. By the way Ginn Fan, it's 2 AP titles and 1 BCS title for USC.
October 27, 2005 | Unregistered CommenterAndy
I think we have two different kinds of comparisons going on in this thread: the comparison between two teams in retrospect (i.e. comparing Ohio State, Miami, and USC in '02) and comparing two teams in an unfinished season (USC, Texas and VT this year)

To me, looking at teams in retrospect is about looking at who had the best season. In '02 USC may have been playing better football than anyone else in the country at the end of the year, but they had lost two games. Perhaps they would've beaten tOSU and Miami, but they didn't earn that opportunity.

Looking at teams in real-time is a whole different story. You can base some of your ranking on the season a team is having (i.e. their record) but the emphasis should be on how impressive a team is and who they've beat. Texas and USC each have marquee wins against teams (tOSU and ND respecitvely) who are the best two loss teams in the country. VT lacks that marquee win, and that is why VT is currently behind.
October 27, 2005 | Unregistered CommenterJJ
Andy,

We judge championship teams by much more than just being undefeated. Even if Boise State had gone undefeated last year, they likely would not have played for the National Championship (even if two of Oklahoma, Auburn, and SC lost).

Now, that example can be easily explained: They aren't in a BCS conference. But none the less, we now have at least two standards: Unbeaten and from a BCS conference.

Additionally, Miami and LSU both have one loss right now, but are both ahead of undefeated UCLA in both the AP and coaches poll, and Miami is ahead of UCLA in the Harris poll. It is uncertain whether, if UCLA beats SC and UGA, VT, and 'Bama lose, UCLA would end up playing for the national championship. It has been suggested that beating SC could be enough to vault them up in the polls, but others have speculated that LSU or Miami (if they win out) would be in the NC game ahead of UCLA. If being unbeaten was the only factor, certainly UCLA would be in the NC game in that case.

This certainly suggests that some other criteria is a factor. It can't be placed squarely on being unbeaten. Whether is a good thing or a bad thing, that's up for debate. But certainly, unbeaten does not necessarily equal National Champ.

And, just for the record, if SC loses while Texas and UGA, VT, UCLA, or Alabama is also undefeated, I don't think SC should be in the title game. My example was WVA, a Big East team (the worst BCS conference).
October 27, 2005 | Unregistered CommenterGinn Fan
Just a note on strength of schedule -

Sure, VT, Bama, and Georgia will all improve. So will USC and UCLA, when the two play each other. Just noting that no team except probably Texas has a full slate of easy games left (which explains, probably, why they currently have the best SOS). Also, Texas's good SOS is very dependant on Texas Tech and OSU playing well.
October 27, 2005 | Unregistered CommenterUnderbruin
Q: How can a team be less defeated than “undefeated”?
A: A duck.

Face it, there is no “National Champion” and there never has been. What we’ve had are popularity contests. (And, with the BCS, popularity contests that are modified by obtusely calculated “strength of schedule” ratings or whatever--as much a farce as the polls.)

I’m a USC alum and lifelong fan, but I have to agree with LD on this. To say that a team that wins all their games isn’t a contender for the national title no matter what any other team is doing (or how they are regarded) is just wrong. Of course they are. There’s no way to prove otherwise. You think Auburn couldn’t have beat USC last year? Really? Prove it.

That’s right, I think Auburn is as much a “National Champion” as USC last year. That is to say that NEITHER of them really are because they both won out. The polls and computers were wrong to pit Oklahoma against USC. Were they wrong to keep Auburn out? We’ll never know.

We’re essentially telling teams from “lesser” conferences (the qualification being something determined by who? The BCS? We all know that’s all about $$) that there’s nothing they can do to win a championship unless the supposed “best” teams (USC, Texas, etc) happen to lose, and even then, probably not.

What’s the point of playing games at all in that case? (Heck, teams like Boise State should be dropped from Div. 1-A if they’re not allowed to contend for a title.) The coaches could submit their schemes to a committee, the players could all go through strength and speed tests, and then we could all agree on who’s the best and award the trophy. (I know, I know, it’s a gross exaggeration to make a point, but still, I say it.)

Seems obvious to me that we should either have a playoff in which the winners of each conference play each other to determine a champ, or we go back to the old bowl system where the winners of each bowl can argue over who’s the best and all have a legitimate claim.

(And anyway, this obsession with knowing who the “best” is just seems ridiculous to me. It doesn’t matter who’s the “best”, what matters is who wins. If you think your team is better than mine, then lets get them to play, otherwise, have another beer.)

Bottom line: Losing games should count just as much as winning them. Otherwise, neither has value.
October 27, 2005 | Unregistered CommenterMonroe
My argument wasn't about inserting USC into a title game if they lose (or VT or Texas or UGA or UCLA or Bama for that matter).

It's about how we evaluate teams.

The majority of you, perhaps reflecting the notion that you're unsure what you're watching and how to distinguish very good from just good from average, etc. and to be able to distinguish among the very good teams.

To me, I think the whole idea of "deserving" to play in a title game is absolute bunk. I'm not happy with the way some teams make their scedules, but just because one has a better schedule than another is a flawed way to reward who goes and who doesn't.

To me, the whole point of polls, and of ranking teams, is to take in all the information possible about a team (stats, competition, observation, etc.) and do our best to rank them.

I think the problem, looking at CFB history, is that we've unevenly applied different measures of determination. Sometimes it looks like teams have been ranked with less a regard to their record, and other times, it looks like all that matters is the record.

I view records as inaccurate measures of a team's aptitude. Schedules vary so much across the game and across the conferences, and each team's experience is different. Maybe one team got a trap game on the road that another similar-record foe just never had, and lost a close one.

Because so much is up in the air, even judging by records, the only real way to do this right is to do our best to rank teams by WHO IS BEST.

Someone said that my suggestion boils down to a popularity contest. Yes and no. The current way it's done, with the polls, yes. I think that's because over the years pressure has gotten so heavy on the voters to not appear out of line, they vote MOSTLY the way many of you argue---by record, and then along lines of conference bias and judgment of schedule faced. The problem is, this has been incomplete and not all that accurate.

Additionally, I think a lot of the voters just don't know a whole lot about what's going on out there. It may not be their fault, but they don't have the most trained eyes. Its not out of lack of effort, this isn't a criticism, but I think its easier to get things much more accurately if we have more trained eyes doing the voting. Do this, and we eliminate the "popularity contest" aspect of it because there are plenty of smart football people out there (Lord knows I've bumped into a few) who don't get all mixed up with that, and can separate the wheat from the chaff, so to speak.

If that were to happen, I hunch we'd see the polls start to look a little different, not perhaps exactly as I've done my rankings, but certainly not this weird slotting system we have now.

Anyway, I trust my eyes and schooling in football over the years more than I trust simply going by record---and I've provided plenty of examples in the post already where judgments of record were wrong.
October 27, 2005 | Registered CommenterCFR

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Post:
 
All HTML will be escaped. Hyperlinks will be created for URLs automatically.