Writers of Pro Football Prospectus 2008

Most Recent FO Features

FolesNic12.jpg

» 2014 KUBIAK vs. ADP: The Overrated

Where does KUBIAK differ most strongly from public opinion, and which players are most likely to disappoint their owners in the upcoming fantasy football season?

01 Dec 2008

Confessions of a Football Junkie: Big 12 Fail

by Russell Levine

In the wake of Oklahoma passing Texas, a team with the same record and to which Oklahoma lost by 10 points on a neutral field, it is tempting to blame the BCS and proclaim this just the latest disaster of the system.

Such arguments are low-hanging fruit. If the BCS had an approval rating, it would make the Bush administration and Congress seem like smashing successes by comparisons.

But such arguments are also misplaced.

The BCS was created to perform a single task: Match the Nos. 1 and 2 teams at the end of the season in a championship game. Nothing more, nothing less. It was not intended to create a single bowl matchup beyond the title game. It was not even intended to provide midseason rankings, as the computer rankings on which it relies for one-third of its formula weight need an entire season's worth of data to approach accuracy. And it certainly was not designed to break conference ties.

If there is a villain in the Texas-Oklahoma-Texas Tech (and make no mistake, this is a three-team argument; more on that in a bit) kerfluffle, it is the Big 12 conference for allowing the BCS formula to be used to settle its standings conundrum. Granted, it is the fifth tie-breaker and the conference organizers probably never felt it would come into play.

The reasons against using the BCS for such a task are numerous, but ABC's Brent Musburger gave the best argument during Saturday night's Oklahoma-Texas Tech telecast: Why would the Big 12 willingly allow coaches from other conferences (who vote in the coaches poll) and writers from other areas of the country who never cover the Big 12 (who vote in the Harris poll) to determine the outcome of its divisional race and potentially its representative in the national championship game?

Musburger's proposed solution was to use a formula of point differential against common opponents to break the deadlock, and he's exactly right. Coaches get nervous any time point differential comes into play, because we have all seen many a 21-point game that was far closer than other 10-point games, plus it brings into to play the potential for running up the score, a label that no coach wants to wear.

Still, as a fifth tiebreaker, it's innocuous enough to use point differential, and only when applied to common opponents. Don't like point differential? Fine -- use points allowed as a standard. Either way, the key is to look at how a group of teams, in this case three, played against the same group of teams to determine which one is better.

Doing so would remove the ugly specter of politicking that we saw from Mack Brown, who was interviewed during Saturday's broadcast -- and who seems to find himself in this situation a lot. It also would remove the stench of voter shenanigans, some of which came into play in this week's polls, as expertly detailed by ESPN.com's Brad Edwards.

Some voters, having no doubt heard the week's worth of propaganda from the Texas delegation about the head-to-head result with Oklahoma, tried to alter the outcome. These are the same voters who willfully, and without hesitation, put Oklahoma in front of Texas a week ago after the Sooners destroyed Texas Tech (a team that had beaten Texas in a nail-biter). How else to explain Oklahoma losing ground to Texas in both polls on the heels of a 61-41 road win at Oklahoma State -- a team that Texas had narrowly beaten in Austin? There is no other explanation other than the obvious: Having decided a week ago to forget about the head-to-head result between Texas and Oklahoma, the voters did a John Kerry flip-flop (have to get in a dig at each party in the interests of remaining politically neutral here at FO) once they realized the ramifications of their move a week ago. Only it was too late -- the die was cast and Oklahoma remained ahead in the BCS.

I feel for Texas, and I don't blame Brown for lobbying for his team's cause. I continue to rate Texas ahead of Oklahoma in the BlogPoll (although I consider Texas, Oklahoma, and Florida to basically be equal) based primarily on that head-to-head result. But I didn't fill out a BlogPoll ballot in order to break a three-way tie, and I'm also fine with the results of the BCS standings and think that sending Oklahoma to the conference title game is the fairest result.

The fact that this is a three-way tie and not a head-to-head argument removes all the injustice from Texas's cause. The Texas contingent would simply like to dismiss Texas Tech from the argument because the Red Raiders are buried in the BCS thanks to the blowout loss at Oklahoma. But this three-way tie is strictly a Big 12 matter, and has nothing to do with national standing. Texas can no more dismiss Texas Tech from the discussion than Oklahoma could dismiss Texas. Doing so would make Texas, not Oklahoma, the primary beneficiary of the Sooners' impressive performance against the Red Raiders.

The fact remains that if head-to-head is to be the deciding factor, Texas loses the argument to Texas Tech just as it wins the argument against Oklahoma, leaving us right back where we started -- in a three-way tie.

We've seen similar situations in college football before. In 2000, Miami felt it had been robbed of a national championship game berth that went instead to Florida State, a team Miami had beaten and which had the same record. But the argument had no legs. Miami had suffered its only loss that season against Washington, another team with an identical record and which had beaten Miami head-to-head. If head-to-head was to be the deciding factor between the Hurricanes and Seminoles, why wouldn't the same standard apply to the Hurricanes and Huskies?

Getting back to the Big 12 South, a deeper analysis such as the one performed by BCS Guru Sam Chi suggests that Oklahoma was the most worthy of the three candidates. The entire piece is worthy of a read, especially for anyone who feels Texas got the short end of the stick, but the talking points boil down to this: Oklahoma performed better against the rest of the Big 12 South and had more impressive non-conference results (two wins vs. BCS top 15 teams in Cincinnati and TCU). Oklahoma also had a more impressive resume in the entire OU-Tech-UT round robin, with bigger point- and yardage-differentials in their favor (hat tip: Saurian Sagacity).

No matter what method had been used to break the three-way tie, someone would have been unhappy. That's just the nature of the beast. Texas has its arguments and the win over Oklahoma is a strong emotional factor in its favor. But it's not the only factor.

Regardless, expect the Big 12 to change its tie-breaking procedures in the wake of this result. The Texas backers will demand, and get, at least that pound of flesh. In the meantime, do not blame the BCS. It has enough issues of its own without being saddled with a misplaced responsibility as it has been here.

John L. Smith Trophy

There were many worthy candidates for this week's JLS. There was Nebraska's Bo Pelini, whose fake field goal try went just a bit awry (apparently, the Raiders didn't see the highlight). There was also LSU's Les Miles, who mismanaged the clock after Arkansas took the lead late in an eventual win. But the trophy goes instead to West Virginia's Bill Stewart for another clock mishap. The Mountaineers led Pittsburgh, 15-13, in the final minutes when the Panthers drove into potential game-winning field-goal territory. The conditions were bad, and Heinz Field is a notoriously difficult stadium in which to kick, so the proper play was to try and force the field goal. But once Pitt moved into extra point-like range, Stewart needed to think about the clock. He called his final timeout with 2:21 remaining and Pitt having a first-and-goal at the West Virginia 5-yard line. At that point, Stewart should have instructed his defense to allow Pitt to score a touchdown. Since a team can burn roughly 45 seconds per play, even a heroic goal-line stand would have burned all but the final few ticks off the clock and left Pitt with a field goal attempt of less than 25 yards. Instead, West Virginia played tough defense for two downs before appearing to allow Pittsburgh to score from the one with 52 seconds left. Even with that little time remaining, West Virginia drove to the Pitt 18 before an incomplete pass in the end zone sealed the game. Had they allowed Pitt to score on first down, they could have had more than two minutes left -- invaluable given that their best offensive weapon is a scrambling quarterback.

BlogPoll Ballot

This season, I am again voting in the BlogPoll, hosted by mgoblog, and now available on CBS Sportsline. I'll post my ballot in Junkie each week. Feel free to comment, and I may make changes based on comments for a revised ballot later in the week.

Rank Team Delta
1 Texas --
2 Oklahoma --
3 Florida --
4 Alabama --
5 Penn State --
6 Southern Cal --
7 Texas Tech --
8 Utah --
9 Ohio State 2
10 Boise State 2
11 Oklahoma State 2
12 Oregon 9
13 TCU 1
14 Georgia Tech 3
15 Cincinnati 1
16 Missouri 6
17 Brigham Young 1
18 Boston College 1
19 Georgia 4
20 Oregon State 7
21 Ball State 1
22 Mississippi 2
23 Michigan State 1
24 Northwestern 1
25 Virginia Tech 1



Dropped out: Florida State (No. 25).

Rankings that may require further explanation: As mentioned above, I'm keeping Texas ahead of Oklahoma because I'm not trying to break a three-way tie. I watched the Longhorns beat the Sooners once and I think they could do it again, but I'm not making any predictions about my top three. I do feel there is a dropoff to Alabama, but they may prove me wrong this week and a 'Bama win probably moves them back up to No. 1. Not a whole lot of change elsewhere in the poll, other than a major leap up for Georgia Tech.

Games I watched at least part of: Texas A&M-Texas, West Virginia-Pittsburgh, LSU-Arkansas, Colorado-Nebraska, Georgia Tech-Georgia, Florida-Florida State, Auburn-Alabama, Oregon-Oregon State, Oklahoma-Oklahoma State, Notre Dame-USC.

Posted by: Russell Levine on 01 Dec 2008

65 comments, Last at 02 Dec 2008, 11:13pm by Pat (filler)

Comments

1
by Dennis :: Mon, 12/01/2008 - 5:10pm

Given the current system, using the BCS rankings as the tiebreaker makes the most sense. The Big 12 wants to give its champion the best chance to play in the BCS championship, and the team that is ranked the highest in the BCS will have the best chance to do so, provided it wins the Big 12 championship game.

Suppose there was a tiebreaker in place that Texas Tech would win. If they went on to win the conference championship, there's still a good chance they wouldn't make it to #2 in the BCS rankings. (Or make it a bit more obvious and suppose the team that won the tiebreaker also had a non-conference loss, so it was clear it would not make it to #2.) The Big 12 is better served by having a team play in the BCS championship than not having a team in the BCS championship. So it is in their best interest interest to have the team with the best chance of getting there win the tiebreaker.

As for the question "Why would the Big 12 willingly allow coaches from other conferences (who vote in the coaches poll) and writers from other areas of the country who never cover the Big 12 (who vote in the Harris poll) to determine the outcome of its divisional race and potentially its representative in the national championship game?", the bigger question is "why doesn't anyone think it's a complete and total conflict of interest to have coaches determining who plays in the BCS championship to begin with?"

16
by Steve (not verified) :: Mon, 12/01/2008 - 8:20pm

You're assuming that all the out of conference Coaches and Harris Poll writers are being honest in their voting. Suppose they deliberately voted a weaker team in to sabotage a conferences BCS chances. How does that serve the interests of the conference?

19
by Eddo :: Mon, 12/01/2008 - 9:14pm

I don't know, I think he's implying that the voters are unreliable when he says:
"the bigger question is 'why doesn't anyone think it's a complete and total conflict of interest to have coaches determining who plays in the BCS championship to begin with?'"

22
by Steve (not verified) :: Mon, 12/01/2008 - 10:36pm

If the voters are unreliable, then the original point about the BCS being a good way to break the tie doesn't hold water.

26
by Dennis :: Mon, 12/01/2008 - 11:25pm

I'm saying that I think it's absurd that the coaches poll is used in the BCS to begin with because it is a huge conflict of interest. At the same time, given the current system, using the BCS rankings to break the tie makes sense for the Big 12 because it maximizes the chance to get the conference champion into the BCS championship.

39
by Eddo :: Tue, 12/02/2008 - 11:39am

His point was that using the system itself is a viable tiebreaker, given that the system decides the national champ. Ideally, the system wouldn't be flawed, but it is.

2
by Joseph :: Mon, 12/01/2008 - 5:11pm

This is the copy of a comment from the BCS guru site that Russell links:

Howard said...
Everyone is talking how to tweak the system and to me its pretty simple. Have a stipilation that if there are 2 teams in the same conference both ranked in the top 5 (BCS) then they play in the confernce championship game regardless of who wins the North or South. It should have happened last year in the SEC also with Georgia playing LSU instead of Tennessee, and it should happen this year also with Texas. Imagine Texas/Oklahoma and Florida/Alabama playing this weekend for the right to go to the BCS championship game? Missouri has no business in that game if there are other teams in their conference with National championship hopes.

Now, that is some great common sense--although I don't think that it would fly with the AD's or Presidents or whoever makes those decisions.

It appears I will have the first comment for this thread--what say ye, FO commenters?? Does this guy "Howard" have a good idea? And, USC/Penn State homers aside, are these 4 (UF, ALA, OU, UT) the four best teams in the country? Should UT play the SEC winner if OU loses to Missouri in the CC game?

6
by DrewTS (not verified) :: Mon, 12/01/2008 - 5:52pm

I guess you could just have the top 2 teams in the conference play for the title, but that would kind of defeat the whole purpose of having divisions. It's a similar argument that people make about the NFL -- that 10-6 team staying home while an 8-8 division champ gets into the playoffs is unfair. That may be so, but if you're going to have divisions, winning a division has to mean something. The problem that the Big 12 has (and to a lesser extent, the NFL) is that its divisions are unbalanced. You could try re-aligning, but that's a temporary fix because teams go in cycles. It wasn't so long ago that Nebraska, Colorado and Kansas State were powers, and Texas Tech was a cellar dweller.

As for what happens if OU loses in the Big 12 CG, I would think Texas would go to the BCS CG, with one caveat -- voters are flaky. This week, the voters tried to game the system by jumping Texas over Oklahoma. Next week, those same voters would be saying "you shouldn't be in the national championship game if you didn't win your own conference", and they'd find a way to drop Texas below (probably) USC. The moral of the story is, a system that uses opinions to determine a champion is stupid.

35
by Dennis :: Tue, 12/02/2008 - 10:28am

"The moral of the story is, a system that uses opinions to determine a champion is stupid."

Exactly.

59
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 4:08pm

All systems use opinions to determine a champion. The difference is that some are codified ("a division champion is a more worthy candidate champion than a non-division champion with a better record") and some change during the year (the BCS standings).

In the limit of a large number of games (and in a universe where attrition does not exist) the number of opinions needed goes to zero. But we're pretty darn far from that limit.

9
by hrudey (not verified) :: Mon, 12/01/2008 - 6:06pm

Except that the conference championship game, while it does in fact determine the conference's representative for a BCS bowl, is still primarily the championship game for that conference. Look at Georgia - Tennessee last season -- Tennessee went because they straight up beat Georgia. Rewarding Georgia for having a better BCS rank but not winning their division and excluding Tennessee would have been simply unacceptable.

3
by Harris :: Mon, 12/01/2008 - 5:25pm

Points allowed?! And ask a Big 12 team to actually play defense? Russ, you so crazy.

"A little celery is always nice after a good pee."

5
by Dennis :: Mon, 12/01/2008 - 5:36pm

Not to mention it still encourages teams to run up the score since teams benefit by their opponents giving up more points.

4
by Cosmos (not verified) :: Mon, 12/01/2008 - 5:30pm

Some might think me crazy but I think college football thrives on not having a playoff. It's work for lower bowls because nobody cares as strongly about the outcomes. Wrong you say? All of the BCS drama peaks the interest to anyone involved and interested in the college game and even a lot of people not overly interested. I'm not saying no one would be interested in college ball if the playoff system took over but the BCS is the prefect PR storm of bad publicity, in that there is no such thing in bad publicity. It's pre-made drama for the masses. Who get's in, who get's left out, people eat it up. Is it fair? No, but people eat it up because it flys directly in the face of what we "perceive" college athletics is about, fairness, sportsmanship. It would be like Maddie and David finally doing it on Moonlighting. Could college football suffer from "Jumping the Shark"?

7
by DangerGnat :: Mon, 12/01/2008 - 5:55pm

I'm disappointed that you consider Sam Chi's blog post a "deeper analysis" of the situation, when in reality he just looks at common opponents and the flawed metrics of point differential and total yards, with absolutely zero context for either. I expect better from FO. And just to qualify these comments - I am an OU fan.

8
by Tom Gower :: Mon, 12/01/2008 - 6:01pm

What bugs the ever-lovin' heck out of me is that Oklahoma beat Texas Tech so badly they're not even in the picture. If TTU is in the picture, we have a legitimate 3-team argument, and using things like performance against common opponents and non-conference results (each of which I think OU wins on) is more legitimate. Instead OU beat the Red Raiders worse than any #2 team had ever lost to a non-#1 team before, and we have effectively a 2-team race where many voters, Russell apparently included, believe head-to-head is the only result that matters. It's almost like saying "OU, you're too good, you shouldn't have beaten the team Texas lost to by 44." It just drives me batty.

I also don't think the use of the BCS standings as a tiebreaker is in any way wrong, silly, or unjustified. First, as noted above, the B12 as a whole is best off if it gives its top-ranked team the best chance possible to play in the BCS CG. Putting the top BCS-ranked team helps satisfy that goal. Second, the BCS standings are, as a whole, the most commonly accepted formula for measuring a team using both subjective (polls) and objective (computers) factors. It also rewards teams for playing a better non-conference slate, something I would argue is somewhat of a public good-the conference as a whole benefits when its good teams play (and win) good non-conference opponents. Overall, I think I still prefer use of performance against common opponents and in conference games as a tiebreaker over use of the BCS standings, but that doesn't mean using BCS standings is bad.

If I wanted to be a real contrarian, I'd argue for use of anti-head-to-head result as a tiebreaker, but (i) that'd raise too much of a storm to ever happen, and (ii) I'm not sure I believe in that argument myself.

10
by Travis :: Mon, 12/01/2008 - 6:16pm

Overall, I think I still prefer use of performance against common opponents and in conference games as a tiebreaker over use of the BCS standings, but that doesn't mean using BCS standings is bad.

By "performance", do you mean win-loss or net points? The former is already included; the BCS standings were the 5th tiebreaker, with various conference games tiebreakers making up the first four. Though I would have loved to see what Oklahoma could have done if net points were a tiebreaker - Oklahoma could have put up 80+ points in several games this year if given the chance.

15
by Tom Gower :: Mon, 12/01/2008 - 6:54pm

I was thinking more along the lines of net points, both among the tied teams and among all common opponents (with MOV capped at 21 here but maybe not among tied teams). I don't like unlimited net points as a tiebreaker, particularly within a conference setting. I assumed "record against common opponents" was one of the earlier (pre-BCS) tiebreakers; thanks for the link confirming that.

11
by Eddo :: Mon, 12/01/2008 - 6:22pm

"If I wanted to be a real contrarian, I'd argue for use of anti-head-to-head result as a tiebreaker, but (i) that'd raise too much of a storm to ever happen, and (ii) I'm not sure I believe in that argument myself."
Doug Drinen at pro-football-reference.com just asked that question himself, and his simulation suggested that, if anything, the head-to-head matchup was indeed indicative of the better team (link). This makes sense to me, as the better team should win a head-to-head matchup more often than not.

33
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 5:55am

The problem there is that in all likelihood, in the subset of situations where the head-to-head matchup was not indicative of the better team, the better team would likely perform better against common opponents. Which is what happened here.

Using the head-to-head as an end-all be-all tiebreaker is too simplistic. You have more information than one game.

47
by Eddo :: Tue, 12/02/2008 - 2:15pm

I don't know; if head-to-head indicates the best team more often than not, then it's probably the best. And yeah, in the cases where it doesn't, some other measure would be better, but how do you determine that? You can't make the tiebreaker to use head-to-head only when appropriate. Tiebreakers need to be used in all cases, no matter what.

57
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 4:02pm

if head-to-head indicates the best team more often than not, then it's probably the best.

But it doesn't. Oregon State beat USC. You wouldn't want to give them the PAC-10 championship, because they also lost to Oregon and Stanford.

Your statement is that "head to head indicates the best team IF their records are identical." You've already got one "IF" in there. You're already not using it in all cases. What's wrong with adding one more "IF"? Why are you so tied to a linear decision tree, ranking each qualifier in direct order?

61
by Eddo :: Tue, 12/02/2008 - 5:04pm

Well, obviously it only applies when they have identical records; it is a tiebreaker, after all.

But what do you propose, if not a linear decision tree? Why even have won-loss records if you're not going to rank the teams in the same manner every year (or at least knowing the tiebreakers going in)?

Anyways, won-loss record is just a tiebreaker. Every team starts on equal footing. First tiebreaker: won-loss record. What about those cases where won-loss record is not the best indicator of the better team?

64
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 7:16pm

But what do you propose, if not a linear decision tree?

You could easily combine head-to-head and common-opponent performance into a single ranking, weighting each one appropriately.

What about those cases where won-loss record is not the best indicator of the better team?

Much less important in intraconference play. The conference is heavily connected - won-loss records are a good first-order indicator.

18
by Crash (not verified) :: Mon, 12/01/2008 - 8:32pm

"It also rewards teams for playing a better non-conference slate, something I would argue is somewhat of a public good-the conference as a whole benefits when its good teams play (and win) good non-conference opponents."

Only if they win. A single OOC loss by any of these teams would have eliminated them from consideration, which is wrong when we are trying to decide a conference question. Suppose one of team had travled to Gainesville or USC (a pipe dream, I know) and lost a squeaker, while the other got fat off of I-AA and Sunbelt teams. The team with the weaker schedule would be rewarded here for not having a second loss.

12
by B :: Mon, 12/01/2008 - 6:27pm

My solution for using net points as a tiebreaker is to cap them at some number per game, something around 25-28 I think would be fair. That way when OU or whomever blows out cupcake state by 50 points, it only counts as +25 in their point differential, so the team has an incentive to beat the opponent soundly, but no benefit to running up the score past some arbitrary point. Of course what that point is is up for debate, but the stomps vs guts analisys seems to indicate that it is meaningful to win by 20+ vs win by 7.

14
by DangerGnat :: Mon, 12/01/2008 - 6:43pm

I agree, and would even go a little further, to say that there needs to be some kind of sportsmanship code in college football - like, automatically put your 2nd team in the game if you begin the 3rd quarter leading by 28+ points. If Oklahoma is going to schedule ridiculous opponents like Chattanooga, and doormats like Baylor and Iowa State are part of the major conferences, the BCS has to stop rewarding poor sportsmanship by giving voters leeway to consider "style points". How about a penalty for winning by too many points? I bet we would see those 4th quarter TD's against hopelessly overmatched opponents stop very quickly.

Of course a "code" would be nearly impossible to enforce, but at least it would be a written expression of what is good sportsmanship, spelled out in clear terms for all coaches to follow. Also, if the computers enforced a "blowout penalty", that would have a real, measurable effect on the bad sportsmanship factor inherent in the BCS.

13
by MC2 :: Mon, 12/01/2008 - 6:39pm

I agree that the BCS is not to blame for this mess, but it still seems a little unfair. Also, if Florida beats Alabama, it will seem a little unfair that Alabama will have one loss (to a great team), while Florida will have one loss (to a mediocre-at-best team), yet Florida will get to play for the National Championship. So much injustice...

If only there were a way to settle it on the field, where Florida, Alabama, Oklahoma, Texas, Texas Tech, USC and maybe even a couple of other teams (like Penn State and Utah, for example) could all have a fair shot at winning it all. Boy, that would be great...

Unfortunately, I can't think of any way to make that utopian fantasy into a reality. If only there were a way, though, that'd really be great...

29
by Kibbles :: Tue, 12/02/2008 - 12:38am

So what you're saying is that it's a bit unfair that if Florida beats Alabama, Florida will play for the title and Alabama won't. You'd rather have a system that settled the matter on the field, where Florida and Alabama played each other and the winner went on to play for the title and the loser didn't. Is that about right?

30
by MC2 :: Tue, 12/02/2008 - 3:26am

I'd rather have a system where if one team gets a mulligan, the other does as well, as opposed to a system where Florida's home loss to a mediocre Ole Miss team just disappears down the memory hole.

40
by Eddo :: Tue, 12/02/2008 - 11:50am

But how is a playoff system any different? Why didn't the Patriots get a mulligan for last year's Super Bowl when the Giants already had six of them?

46
by MC2 :: Tue, 12/02/2008 - 1:48pm

The difference is that in a playoff system, there's a clear delineation as to when the games start to "count", i.e. when it truly becomes single-elimination. One of the things that really annoys me is when BCS defenders say that in the current system, the regular season is a tournament. I've never heard of a "tournament" where it's single-elimination for some teams and double-elimination for others.

As for the comparison to the NFL, if they did it the same way the NCAA does, the Giants would have never been "invited" to the playoffs last year. After all, 3 of the 4 teams that the Giants beat in the postseason were teams that they had already lost to in the regular season (including one team that they had lost to TWICE). So clearly, the NFL believes in mulligans. Again, however, the difference is that in the NFL everyone knows when the games start to "count", i.e. when the mulligans end. The annoying thing about the college game is that you never know whether a result is going to "count" or not, because they arbitrarily change the rules in the middle of the "tournament".

48
by Eddo :: Tue, 12/02/2008 - 2:20pm

There's a clear delineation in your case, as well: the SEC conference championship game is when the elimination games start.

And the NFL uses head-to-head as their first tiebreaker anyways. It didn't come into play for the Giants, since they had a better outright record than the teams below them in playoff seeding, so I don't see your point.

You also make it seem like an 8-4 Florida team could beat a 12-0 Alabama team and be rated higher in the BCS standings. This is clearly not the case, as a four-loss team would not be rated ahead of a one-loss team. You also make it seem like a non conference champion has no shot at the national title; 2003 Oklahoma and 2001(?) Nebraska are counter-evidence.

53
by MC2 :: Tue, 12/02/2008 - 2:55pm

So the conference championship games are the start of single-elimination and regular season losses don't count? Tell that to USC or Penn State. Also, why is it single-elimination for Texas and Texas Tech and double-elimination for Oklahoma? What kind of "tournament" is that?

My point about the Giants is that based on college standards, the question of "Who's better: Cowboys or Giants?" had already been answered on the field (TWICE) during the regular season, so why give the Giants another shot?

As for your last point, it seems like a gross and deliberate distortion of my argument. I never claimed that a 4-loss team could finish ahead of a one-loss team. But we have certainly seen cases where a 2-loss team finished ahead of a 1-loss team, for example. Also, you say that non-conference champs aren't automatically excluded from contention, but that seems to be the exact reason why USC is being excluded this year.

Finally, you ignore my most important point, which is that in the current college system, important disputes like the ones we've been debating here are basically determined "on the fly", using purely ad hoc criteria. Can a 2-loss team be ranked ahead of a 1-loss team? Sometimes yes, sometimes no. Can a team that doesn't win its conference play for the national championship? Sometimes yes, sometimes no. Do head-to-head results matter? Maybe. Then again, maybe not.

It would be much simpler and much less arbitrary to simply have a playoff sytem with certain predetermined criteria. If you meet the criteria, you're in. If not, you're out. If that results in an odd number of teams getting in, then randomly draw for a first-round bye(s). Would it be perfect? No. Would it still lead to some people feeling that the best team got left out? Probably. But at least everyone would know the deal right from the start, as opposed to the current system, where you make the rules up as you go, virtually guaranteeing controversy every year. Apparently, that's what makes some people like it.

56
by Eddo :: Tue, 12/02/2008 - 3:56pm

Sure, a playoff system would solve a lot of problems, but that's not what we have now. So, when you're only taking two teams to play for the national championship, head-to-head is as good a tiebreaker as any. Maybe Alabama's loss to a stronger is indicative they're a "better" team, but Florida's head-to-head win makes them more "deserving". Let's say Alabama then wins the national championship game. Couldn't Florida then claim they were better by virtue of beating Alabama?

There's no perfect way, but head-to-head is at least letting the teams settle it amongst themselves as opposed to having third parties decide.

58
by Eddo :: Tue, 12/02/2008 - 4:07pm

And, in regards to your point about ad-hoc criteria, that's what happens when you narrow the playoff field to two teams. With a twelve team field, like the NFL, you can just let sex in from each conference, comfortable that one of the two or three worst teams will only win the championship once in a blue moon. However, when you have a single game decide the whole thing, you must make sure you get the two best teams, and there's not hard-and-fast definition of what makes a team the best.

62
by MC2 :: Tue, 12/02/2008 - 7:10pm

Exactly. That's precisely the problem with the current system, and it's also the reason why I get so annoyed when people make arguments like, "Oh, we don't need a tournament; the regular season IS like a tournament" - yeah, like one of those tournaments where they just make up new rules as they go along? Some tournament.

Equally annoying is the claim that a playoff would "take the excitement out of the regular season" - yeah, that's why nobody pays any attention to the NFL regular season, right? It only takes the excitement out of the regular season if you let in too many teams, like they do in basketball. The only thing an 8-team playoff (of the kind I've proposed above) would take away from the regular season is all the lobbying and politicking that goes on under the current system.

17
by Steve (not verified) :: Mon, 12/01/2008 - 8:27pm

Using the BCS rankings to break a conference tie is a terrible idea. Besides the issues raised with out of conference coaches and writers deciding the winner, you have the problem of out of conference games coming into play. Conference results should be determined solely on the basis of things that happen in conference. If Tech, Texas or Oklahoma had lost an OOC game that would have knocked them out of consideration, regardless of who they played OOC or who they lost to.

20
by ChrisH :: Mon, 12/01/2008 - 9:46pm

Spending some time thinking about this (after watching my Beavers leave their defense in Arizona last week), and then watching ESPN while I worked today, I just grew more angry listening to the people talk. They say "The polls got it right, but then here come the computers messing it up again". If you only want the computers to agree with the polls for the BCS, then what's the purpose of the computers? I'm lacking too much sympathy for Texas for three reasons:

- They could have objected to the Big 12 when this tiebreak was put into place, but they didn't (as far as I've heard)
- They could have beaten TTU, which is a worse loss than Oklahoma suffered, but they didn't
- They know that the BCS computers care about who you play, so they could have played a harder non-conference schedule than Oklahoma (who, come on, played winless Washington and still has a higher rating!) but they didn't.

Teams have known that the BCS will use computer rankings for a decade now and still get mad every year about it. They know computers care about SOS for non-conference teams, but still schedule awful non-conference teams to pad their wins. If the BCS wants to try to make things more just, here is what I would do:

- No polls until October (they can't stop the AP, but they can stop the coaches if they want to be in the BCS)
- Require all teams that wish to qualify for a BCS bowl to do the following: Play a BCS conference school at home and on the road from outside your conference, every year. Those SEC teams that refuse to travel, give them no choice.
- Most computers leave out I-AA teams from their ratings, but the BCS should punish for scheduling them. Schedule a non-BCS team instead, play at home, I don't care. But if you schedule a I-AA team, you get .05 taken off your BCS score (if you schedule two, .12)

For teams that aren't going to make the BCS and just want to win more games (see: Syracuse, Duke, etc..), you can schedule like you do now, and try to get to 6 wins for a bowl, and it'll be fine. However, the days of Kansas State playing 3 NE Syracuse Valley State Polytech to pad their record and then complaining when they get left out would be over.

Most good teams should do this already, but I'd like to hear objections/other suggestions for what the BCS could do to try to fix problems in the future. If I could, I'd make all the teams play 9 conference games (for SEC/ACC/Big12 people that complain about USC not having to play a title game, they also play every conference team, every year, which no other major conference team does) since we went to 12 games. I just realized that this new scheduling mandate would have killed Oregon State this year (and we played Penn State and Utah on the road, but only one is in a BCS conference), so perhaps it has to be tweaked more, but those teams with only 8 conference games should easily pull it off. Maybe just mandate a minimum of 10 games against other BCS conference teams?

21
by Eddo :: Mon, 12/01/2008 - 10:24pm

ChrisH, I like just about every suggestion you make (more regulation is a good thing!), but I do disagree with this statement, which you use to justify your lack of sympathy for Texas:
"They know that the BCS computers care about who you play, so they could have played a harder non-conference schedule than Oklahoma (who, come on, played winless Washington and still has a higher rating!) but they didn't."
If anything, Texas schedules better games than most other top-flight programs (remember their home-and-home with Ohio State?), and definitely better than Oklahoma. This year was kind of fluky; Texas played Arkansas, a school that has been near the top of the SEC in recent years, but had a down year this year. Oklahoma played Cincinnati and TCU, both of which had very good years this year, but are not traditionally as strong as Arkansas. It's hard to fault Texas for not having the foresight years ago that Arkansas would not be as good as TCU and Cincinnati. (I say this as an Illinois alum who has no stake in the Big XII.)

I don't mean to diminish your post; I really like the idea about requiring games against other BCS schools in order to make a BCS bowl. One question, though: what about non-BCS schools like Utah? Would you require them to have two games against BCS conference schools in order to qualify? I imagine very few BCS schools would schedule an excellent non-BCS school when they're already forced to play two out-of-conference BCS foes.

24
by Irish Boy (not verified) :: Mon, 12/01/2008 - 10:48pm

"I imagine very few BCS schools would schedule an excellent non-BCS school when they're already forced to play two out-of-conference BCS foes."

Isn't that pathetic? Our school (Illinois) seems to have no trouble doing it, scheduling Missouri and Cincinatti next year away from home AND Fresno State. Why can't the LSUs of the world find the time to schedule a team with a pulse?

27
by sam :: Mon, 12/01/2008 - 11:41pm

Because LSU goes into every season thinking they are already going to play a ton of good teams.

--
sam! or the original sam from the old FO

28
by SJT (not verified) :: Tue, 12/02/2008 - 12:31am

So does Texas, Florida, Oklahoma, etc, yet those teams still manage to schedule out of conference games against BCS conference teams.

38
by ChrisH :: Tue, 12/02/2008 - 11:08am

Yes, they do. One other benefit of forcing these teams to play other BCS teams is that we would be able to get a far better idea of how good the conferences actually are relative to each other. At the start of this year, we all thought LSU was still really good heading into SEC play since they tore through their opponents. Maybe if they had been forced to play other BCS teams, we could have seen them struggle in those games early, and known that the SEC wasn't quite as strong this year as we thought it was, and known that the Big 12 was very strong (at least on offense) this year.

41
by ChrisH :: Tue, 12/02/2008 - 12:20pm

One question, though: what about non-BCS schools like Utah? Would you require them to have two games against BCS conference schools in order to qualify? I imagine very few BCS schools would schedule an excellent non-BCS school when they're already forced to play two out-of-conference BCS foes.

This is something that would need to be ironed out as well. This year, does anyone really have a good idea of how good Ball State is? They played Indiana, which is a BCS school, but this year is a really awful one. Utah played @Michigan and Oregon State. BYU scheduled Washington and UCLA. Boise State played @Oregon. TCU played Oklahoma and Stanford. So, it seems that 3/5ths of the teams already had 2 BCS opponents, and the others each had one. I do believe it would be harder, possibly, if teams had to schedule two BCS opponents.

However, say you are a lesser team (Washington State, Syracuse, Northwestern, Vanderbilt, Duke, Stanford, etc...) that has virtually no shot to make a BCS bowl, you can still schedule these games and it counts for them as a BCS team. Of course, some of these teams might not want to schedule them on the off chance they can get into a BCS game and risk losing that money (however, taking the safe money of a lesser bowl seems like a better bet to me). Perhaps we change it to this:

- For teams with 9 conference games (only the Pac-10 currently), require you to play another BCS conference school at least once that year to qualify, and in a 5 year span to play a minimum of 2 on the road.
- For teams with 8 conference games, require the home and away game each year against another BCS conference school.

I would want non-BCS conference schools to have to play 2 games against a BCS conference team, since otherwise it gives a team like Ball State far more incentive to try to schedule as easy as possible, knowing that if they do go undefeated, the huge BCS game windfall more than makes up for not getting paid to go visit a BCS team and likely lose. If a team doesn't want to take the risk of losing against a decent team, they don't get the benefits of having to play a hard schedule. Yes, possibly unfair to some, but we've seen other teams play harder schedules and do well in BCS games (Boise State, Utah) and teams that play pathetically easy schedules make it and get destroyed (Hawaii) by the first good team they played all year.

23
by Irish Boy (not verified) :: Mon, 12/01/2008 - 10:36pm

I agree with all of that except that I'd merely mandate 8 games and not nine, not counting championship games. Assuming a schedule of 8 in-conference games and 4 out of conference games keeps most of the conferences where they are at right now while allowing you to implement the "two BCS team" rule and still leave room to play WAC and MWC teams and whatnot.

25
by Steve (not verified) :: Mon, 12/01/2008 - 10:49pm

I like a lot of your points, but I also had a few quibbles.

Even if Texas had the foresight to object to the use of BCS standings as the 5th tiebreaker, one dissent does not a rule make.

Why is Texas's last second, miracle play road loss to TTU "worse" than Oklahoma losing by 10 points on a neutral field? If anything, I'd say its the very "best" loss that any of the teams have.

Non-conference scheduling is hard to do. Teams have to have open dates and a willingness to play, and these schedules are usually made years in advance. I'm willing to bet that when Oklahoma made its OOC schedule they were counting on TCU and Cincinnati as "easy" wins. Also, Texas is the only one of the three which did not schedule any 1-AA teams this year.

I don't think any of this should matter, though, since OOC results shouldn't determine conference rankings. Maybe they could have taken the ooc games out of the computer polls for the purposes of this tiebreaker.

37
by ChrisH :: Tue, 12/02/2008 - 11:05am

Why is Texas's last second, miracle play road loss to TTU "worse" than Oklahoma losing by 10 points on a neutral field? If anything, I'd say its the very "best" loss that any of the teams have.

The computers will take it as the worst loss, since they lost to the worst team of the 3 teams (TTU played the weakest non-conference schedule of the three, and so typically has the lowest computer rankings of the three). The computers are made to ignore margin of victory (which is pointless, since it was designed to prevent teams from running up the score, but they do that anyway for the human polls) which also leads to this. I imagine if margin of victory was included, that TTU would have suffered the worst loss of the three, but the computers will probably see it as Texas.

Non-conference scheduling is hard to do. Teams have to have open dates and a willingness to play, and these schedules are usually made years in advance. I'm willing to bet that when Oklahoma made its OOC schedule they were counting on TCU and Cincinnati as "easy" wins. Also, Texas is the only one of the three which did not schedule any 1-AA teams this year.

In general Texas has done a better job of this than most teams (the home and home with Ohio State), and the chronic offenders seem to be in the SEC. For example:

LSU: Troy, App State, North Texas, Tulane (0 OOC BCS teams).
Alabama: Tulane, Western Kentucky, Arkansas State, Clemson (1 OOC BCS team)
Georgia: Georgia Southern, Central Michigan, Arizona State, Georgia Tech (hey, they played 2 OOC BCS opponents, yea!)

For reference, here's every OOC team that USC has scheduled over the past 5 years (excluding Notre Dame): Virginia, Ohio State, Idaho, Nebraska, Arkansas, Nebraska, Hawaii, Arkansas, Fresno State, Virginia Tech, Colorado State, BYU.

Out of 12 games (technically 17, if you count Notre Dame, which you should), they played 7 of those games against BCS conferences, 2 MAC games, and 3 WAC games. Somehow, despite refusing to play I-AA teams at all, and scheduling home and home series (Arkansas, Nebraska, Ohio State), USC manages to make a profit and keep playing in BCS games. So while many schools are going to say that they need the money from I-AA games to fund their athletic departments, I'm sure there are more examples of teams that don't need this excuse, and this only applies to teams that want to make the BCS, not to teams that just want to make a bowl game.

I don't think any of this should matter, though, since OOC results shouldn't determine conference rankings. Maybe they could have taken the ooc games out of the computer polls for the purposes of this tiebreaker.

The problem there is that the teams have control over the OOC teams that the schedule, but they have no control over the in-conference games that they play. I fully believe that TTU and OU would beat Missouri, but only Texas got to play them, due to the rotation in the Big 12 schedule. Are you going to say that, because of the rotation not going their way this year, they don't even get a chance to win the conference?

42
by lionsbob :: Tue, 12/02/2008 - 12:49pm

Alabama is going to probably schedule one big out of conference game (as it has done in the past with UCLA,Oklahoma, Florida State, and Clemson. And likely have 3 non-BCS teams as the other out of conference games. I hope they leave off the FCS teams from now on. They plan on playing Virginia Tech next season. They have planned games against Georgia Tech and Penn State in the future.

The ACC had some odd scheduling. Every team but Wake Forest played a FCS team. 3 of them played 2 FCS teams (Florida State, Clemson, and Georgia Tech).

63
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 7:12pm

The computers are made to ignore margin of victory (which is pointless, since it was designed to prevent teams from running up the score, but they do that anyway for the human polls)

Yes, but a human can tell the difference between a 20-7 victory where the opponent's 7 point score came in the last seconds of the 4th quarter and a 20-7 victory where a bizarre fumble return for a touchdown turned a 13-7 game into a 20-7 victory.

More importantly, wins have to have merit. A triple-overtime victory is still a victory. You can't marginalize it away to nothing.

This is why it's ideal to combine a purely descriptive ranking (the statistical rankings) and a semipredictive ranking (human polls).

32
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 5:44am

but still schedule awful non-conference teams to pad their wins.

This is so completely wrong, it's almost funny. Teams don't schedule awful non-conference teams to pad wins. They schedule awful non-conference teams because they're the only teams that will accept lopsided contracts (i.e. 'always away' - at the major team's venue). Teams have to schedule 7 home games a year - which means at least two patsies a year, assuming 4 home and 4 away conference games, and home/away split between the non-patsy non-conference schedule.

Athletic directors have stated this, almost word-for-word, in interviews, press releases, and letters to boosters. It's life.

Teams like Illinois, Indiana schedule harder not because they're ethically superior, or care about wins less. It's because they make less money at home than the larger schools in their conference, so the penalty of going on the road is less - which means you can schedule harder opponents.

Most computers leave out I-AA teams from their ratings

Sorry. In fact, only two of the statistical rankings exclude I-AA teams, and they're the two that are either insane (Billingsley) or completely opaque in their methodology (Anderson-Hester).

Play a BCS conference school at home and on the road from outside your conference, every year.

Neither possible nor fair. The Big 12/SEC play more in-conference games than the Big East, for instance. In addition, densifying the connections between BCS conferences means that you'll, by necessity, increasingly disconnect the non-BCS conferences.

If you're going to do that, you have to split Division IA entirely. Possible, but not likely.

However, the days of Kansas State playing 3 NE Syracuse Valley State Polytech to pad their record and then complaining when they get left out would be over.

As would the days of LSU (and Kansas State, likely), y'know, playing football. Cupcakes are part of the reality of college football economics.

The safest - as in, changes that could actually happen - changes to college football would be:

  • Go back to maximum 1/every 4 years Division I-AA team for bowl eligibility.
  • Require the statistical rankings to exclude all Division I-AA games and teams from rankings provided to the BCS. All of them - wins and losses. Let the voters punish a team if they lose to a Division IAA opponent. They might have a clue what to do - the statistical rankings can't.
  • Require that a conference have no more than 8 games against another conference (or against Division IAA), on average, over a 5 year period to maintain a BCS automatic berth. This one might need to be toned down at first, and then strengthened as scheduling becomes more tractable.
  • Evaluate voters on a yearly basis: if a voter exceeds a certain bias threshold for any one team (or possibly conference) from the given poll at large over the previous 5-year period, exclude his vote for the current year for the BCS rankings. This is eminently doable - the Coaches Polls have published final ballots and the Harris is commissioned by the BCS.
  • Require all statistical rankings to disclose methodology. Clarify what the statistical rankings are supposed to be doing - that is, ranking a team's season, rather than determining the best current team. Have the statistical rankings evaluated by a committee each year. Massey, for instance, should be yanked unless he removes weighting of recent games, and Billingsley should be removed due to lack of statistical merit.
34
by Travis :: Tue, 12/02/2008 - 9:23am

A minor clarification - the Billingsley rankings exclude I-AA teams, but include I-A games against I-AA teams. See his Week 1 standings, in which various teams got .150 (for allowing more than 7 points), .300 (for allowing 7), .400 (for allowing 2-3), or .650 (for a shutout) ratings points for beating I-AA teams. San Diego St., who lost to Cal Poly, lost 16.09 ratings points.

Yes, his system is insane.

36
by Dennis :: Tue, 12/02/2008 - 10:34am

"Teams have known that the BCS will use computer rankings for a decade now and still get mad every year about it. They know computers care about SOS for non-conference teams, but still schedule awful non-conference teams to pad their wins."

The problem is the human rankings have 2/3 the weight, and they put much more weight on number of losses than strength of schedule. And you're still better in the computer rankings with a win over a bad team than a loss to a good team. So a lot of teams aren't willing to take the risk of scheduling a good non-conference opponent.

49
by rjsen (not verified) :: Tue, 12/02/2008 - 2:25pm

The problem, still, is that this assumes any random BCS team is better than any random non-BCS FBS team, which is obviously untrue. If a team played Boise State, Utah, and Ball State this year, they would have had a much tougher slate than a team playing Syracuse, Washington, and Baylor, yet the latter can in some odd way be considered a harder schedule because all the teams are from BCS conferences.

How about this: ditch the notion of BCS conferences, and remove all conference tie-ins to bowl games. Keep the existing BCS formulas with some refinements (i.e., excluding all games against FCS teams). At the end of the year, #1 and #2 play in the title game, just like now. The 4 other BCS bowl games rotate matchups of #3 vs #4, #5 vs #6, #7 vs #8, and #9 vs #10. If a conference has 4 teams in the top 10, they get 4 BCS bowl berths; if a conference has none, they get none. Who cares if the Rose Bowl is Big-10 vs. PAC-10? It should just be the best possible matchup, whether that's USC-Penn State or Texas Tech-Utah. This should extend past the BCS as well -- the next tier of bowls (Cotton, Gator, Holiday, etc.) should rotate matchups between the 11-20 teams, and so on.

I know some would be upset with this system since it doesn't honor traditional pairings, but the fact is that the BCS has already done that. It's time to take that idea to it's logical conclusion and make all bowls as good as they can be based on how good the teams are, not what conference they play in.

31
by Pat (not verified) :: Tue, 12/02/2008 - 5:13am

It was not even intended to provide midseason rankings, as the computer rankings on which it relies for one-third of its formula weight need an entire season's worth of data to approach accuracy.

Unless you're Ken Massey, and have decided that what a team did at the beginning of the season is pretty much useless. How this makes sense in a ranking system intended to rank how a team performed over the year, I have no idea.

50
by asp_j (not verified) :: Tue, 12/02/2008 - 2:28pm

I get it. It's basically ELO-lite, in an attempt to reward teams that are massively better in the second half of the season than the first, like Florida. Personally, if it's only one of six computer polls I don't see the harm, kind of how the NCAA basketball comitee looks at the last 12 games of every at-large candidate, and left off Dayton last year even if they had an RPI of 32, because one of their top players got injured late in the season and they showed they weren't the same team without him.

I'd note in passing that:
* A playoff system would have the same bias, even more so (you win out your last 2-3 games and you win the championship)
* I don't know if it's such a bad idea. From what I understand, the point of the computer polls is to try to evaluate the teams from a variety of perspectives, and I think it's fair that one of them values recent performance. Like I said, it's like how they look at the last 12 games in basketball. Put it another away -- if a team was cruising at #1 all season but lost their star player and completely tanked their last 3 games, would you really want them in the CG?

52
by ChrisH :: Tue, 12/02/2008 - 2:41pm

I believe this also goes along with how you think voters should fill out their ballot: Who has played the best all year, who is playing the best now, or in order of who you would favor to win a game played on a neutral field right now? I don't believe the AP/Coaches/Harris tell the voters which criteria they should follow. If you go off #1, then you put Alabama #1 since they've won every game. If you go by #2, maybe you pick Oklahoma or Florida, who are scoring at will it seems, no matter who they play. Going by #3, many people would pick USC, who probably watched Oklahoma give up 41 points on Saturday and think about how they've given up 49 points total since the start of October.

I think there is room for both types of computer polls, since we are going to average them all out. Myself, I tend to lean towards choosing the teams that are playing great now (criteria #2) since the team at the start of the season might have nothing to do with the team that you see at the end of the year. Now, if that team lost 2 games early, then I don't put them in the title game, but if they had one early loss and since then have been more dominant than anyone else, I imagine they are probably the best team now, and have dealt best with injuries and made the best adjustments of anyone.

65
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 11:13pm

I don't believe the AP/Coaches/Harris tell the voters which criteria they should follow.

Which is fine, because they're human. They can make decisions on a case-by-case basis. Statistics can't.

I think there is room for both types of computer polls, since we are going to average them all out.

No. Statistical polls aren't human. For one, you can't use margin of victory, because a statistical poll will never have enough information to make a sane evaluation. Instead, you'll just get a biased result - biased for a type of team rather than a single team, biased for a style of play, but biased nonetheless. Thus, without margin of victory, you have no way of evaluating "how well" a team is playing now.

If teams played a lot of games, you could probably weight recent results because won/loss would be a decent proxy for team strength. But it isn't - which is why we have statistical rankings in the first place - and so weighting recent results is completely and totally unjustified.

Myself, I tend to lean towards choosing the teams that are playing great now

Again, you're not really understanding the problem. You have no way of evaluating who's playing great now. You have only wins and losses. While it may be true that the team is different now, you have no way of knowing that. Throwing out early results is entirely unjustifiable.

55
by Pat (filler) (not verified) :: Tue, 12/02/2008 - 3:48pm

I get it. It's basically ELO-lite, in an attempt to reward teams that are massively better in the second half of the season than the first, like Florida.

You can't use margin of victory (it's biased) which means that the only way you know a team is "massively better" is because they're winning against strong competition. Except that has nothing to do with a team's performance, and everything to do with the order of their scheduling. The order of a team's scheduling is determined years in advance, and is usually constrained by many other factors.

More importantly, why is the second half of the season more important than the first? Doesn't this give an automatic advantage to conferences with championship games and schools with long-standing rivalries with strong schools late in the year (c.f. UGA/GT, FSU/UF, in many years OSU/UM and USC/ND)?

if a team was cruising at #1 all season but lost their star player and completely tanked their last 3 games, would you really want them in the CG?

1) Did they win those last three games? Then yes. They earned it. (You can't win 3 games and 'completely tank.') Did they lose? Then obviously no.

2) Take it farther. What if they were cruising at #1 all season long, but lost their star player in the second to last game, and then struggled in the last game, but still won. Would you kick them out of the championship game then?

3) Take it even farther. What if they lost their star player a day before the championship game. Would you kick them out of the championship game then?

Obviously, "no" on the last one. Which means the second one should likely still be "no."

What are the rankings for? They're to tell the difference between 5 different 11-1 teams. Which means they're being used as an improved substitute for winning percentage. Therefore you have to rank a team's season - not how they will perform in the future.

Otherwise there's no logical justification for the BCS to exist.

43
by parker (not verified) :: Tue, 12/02/2008 - 12:54pm

How many posters railing against the Big 12 tie-breaker process really think that Texas would have a better record than OU in a 10 team round robin? To me that is ulimate question. The computers do what they're programmed to do. The Coaches do what is politically best for them. The Harris poll voters try to balance these two.

I have no problem with voting in an attempt to game the system. Isn't that why we have human voters incorporated into the system. They get a chance to analyze a dynamic system, while computers exist in a static world. This process is exactly what saved us from a Michigan/Ohio St. championship game in 05. Anyone remember Michigan getting "snubbed"?

If year after year we are getting a championship game in which 2 of the best 4 teams are being matched for a national championship game then I can't really see the beef.

Under the old bowl system we would probably be looking at

Rose Bowl- Penn St. vs. USC
Sugar Bowl- Florida vs. Texas Tech
Fiesta Bowl- Texas vs. Alabama
Orange Bowl- Ohio St. vs. Oklahoma

Those would all be great match ups. Those match ups are probably better than our current crop of match ups which have to include Cincinatti and Utah. But, we would all be left frustrated that the best game that could be put together would be Texas v. Alabama.

Instead we get to see what Utah is made of against one of the conference elite and we get to at least have the two teams of the top 3 teams play for the championship instead of having random voters deciding that since Texas beat Alabama and got jipped out the big 12 championship that they are in fact the best team in the country.

Our current system is the lesser of two evils. I have no specific feeling toward a playoff system being added. I would not be opposed to a playoff system but at the same time I am not opposed to the bowl system. Its all exciting for me as a fan. Excitement is excitement, and this is a pass/fail test. You pass if I get excited, I don't need measure my excitement.

60
by Eddo :: Tue, 12/02/2008 - 4:21pm

"Under the old bowl system we would probably be looking at

Rose Bowl- Penn St. vs. USC
Sugar Bowl- Florida vs. Texas Tech
Fiesta Bowl- Texas vs. Alabama
Orange Bowl- Ohio St. vs. Oklahoma"

Close, but no. The Big East and ACC had tie-ins under the old system, too. Most likely, it would have looked like this (assuming the higher-ranked teams wins any remaining games):

Rose Bowl- Penn State (Big Ten) vs. USC (Pac-10)
Sugar Bowl- Alabama (SEC) vs. Texas (at-large)
Fiesta Bowl- Oklahoma (Big XII) vs. Florida (at-large)
Orange Bowl- Boston College (ACC) vs. Cincinnati (Big East)

I'm not sure about where the at-large teams would be placed, but you get the idea. Cincinnati and Boston College (or Virginia Tech) would still be guaranteed spots and Ohio State and Texas Tech would be left out, as Florida and Texas are more attractive at-large candidates.

44
by RickD :: Tue, 12/02/2008 - 12:58pm

At some point, people are going to have to admit that the "neutral field" where Texas beat Oklahoma was, in fact, in Dallas. While Dallas is not Austin, it's a far cry from what one thinks of when the phrase "neutral field" is used.

51
by asp_j (not verified) :: Tue, 12/02/2008 - 2:32pm

* Dallas is almost exactly halfway between Norman and Austin (in a straight line, no less)
* Dallas has TONS of OU alums, and the stadium is almost always split 50-50 at those games. Seriously, of all the neutral site rivarly games I can think of, I think this one is pretty much the one with the truest neutral site (second would probably be KC for Missouri-Kansas). Florida-Georgia this ain't.

45
by RickD :: Tue, 12/02/2008 - 1:01pm

BTW, you had already hit both parties when you criticized "the Bush administration and Congress". No need to bring in John Kerry to "balance" things.

54
by pete (not verified) :: Tue, 12/02/2008 - 3:04pm

Boise State, Utah, and Ball State did not lose a game, yet none of those three will get a chance to play for the title, and only one of them will get a chance to play in a BCS bowl game. Texas lost a game, and will definitely be going to a BCS game and still has an outside chance to play for the national title.

Or you could simply tell the Big 12 they are idiots for having a conference title game matching their best team vs their fifth best team. But Texas knew the rules going into the year. If they did not win the division based on the tiebreaker in place, they would not get the chance to play for the Big 12 title. The easiest way to win your division is to not lose, AT ALL.

As a fan of the BCS, disaster was averted when OU beat Oklahoma State. Because if the Cowboys won, guess what, Texas still wasnt going to the Big 12 title game! The Tech/Mizzou showdown wouldve been rendered meaningless as Texas (a team that didnt win its conference) would be awaiting the SEC winner in Miami.