Amazon.com Widgets
Membership: Join Now : Login

The BCS worked, so shut up!

[Corrections: USC lost to Cal in Berkeley, not at the Coliseum; corrected 2003.12.09]

I was still coming back from the Big XII Championship when I heard the BCS Selection show on the radio, and the more I heard, the angrier I got. Most of the arguments I've heard today about why Oklahoma "does not belong" in the Sugar Bowl, playing for the national championship, were heard on that show, and many of them originated with World Class Asshat Terry Bowden.

Bowden, for those unaware, is the son of Florida State head coach Bobby Bowden, and a former NCAA head coach himself. He has admitted that during his tenure at Auburn, boosters funneled tens of thousands of dollars to players, while he was certifying to the NCAA that his team - already on probation from before his arrival - was clean. Terry Bowden serves as ABC's anchor analyst for NCAA football, opining on all teams from his seat in ABC's Times Square building, including healthy doses of praise for his father and Florida State, usually when the Seminoles deserve it least.

When announcing the BCS bowl selections, Bowden had a conniption because both the AP (writers) and ESPN/USA Today (coaches) polls list USC as #1, LSU as #2, and OU as #3. That's because LSU won its conference championship Saturday night, and USC won its final regular season game, but OU lost the Big XII Championship to Kansas State in a game I'll talk about elsewhere.

It was OU's first defeat of the season, making OU, LSU, and USC the only three serious one-loss schools in NCAA football. There are no undefeated teams. Three other teams have only one loss each, but because they played easy schedules compared to the major conference schools, they get no respect from the poll voters, and their strength of schedule points hurt them in BCS ratings.

Schedule sanity

Miami of Ohio finished 11th in the BCS, below three-loss (but Big XII champions) Kansas State, largely because the highest-ranked team the school defeated was #24 Bowling Green. Their only loss was to Iowa, #13 in the final rankings, in the first game of the season, but after that, and discarding Bowling Green, Miami of Ohio never played a team ranked higher than #46. Similarly, one-loss TCU beat teams ranked #35, #65, #67, #70, #79, #83, #86, #89, #97, #116, #117, and lost to #27-ranked Southern Mississippi. Note that there are only 117 teams in NCAA Division I football, so two of TCU's wins came against the two worst teams in the sport. All but two of TCU's opponents ranked in the bottom half of NCAA football, and of the two that didn't, TCU was 1-1. One-loss Boise State lost to #47 Oregon State, the highest-ranking opponent the team faced all year. They defeated teams ranked #57, #61, #64, #75, #76, #78, #87, #101, #109, #114, #116, and a Division II school that doesn't even count for BCS purposes. All of Boise State's wins were over teams in the bottom half of the sport, and three were in the bottom ten.

Oklahoma defeated teams ranked #6, #21, #33, #37, #40, #52, #56, #57, #69, #71, #92, #93, and lost to #10 Kansas State. Six of Oklahoma's twelve wins were against bottom-half opponents, but by definition, half the teams in football are in the bottom half, so that's not too anomalous. Plus, when the schedule was set, everyone thought Alabama (#71) and UCLA (#52), and maybe Fresno State (#57) would turn out to be tougher. No one really imagined Colorado would finish #56, either, or Texas Tech #40; both were projected as top 25 teams.

LSU had a tough schedule, beating #12 Georgia twice, plus #19, #38, #39, #60, #71, #78, #89, #95. LSU's loss was to #15 Florida on October 11 - OU-Texas day. USC beat #16, #39, #47, #48, #52, #55, #64, #72, #73, #75, #89, and lost only to #43 California in overtime.

Of these three, the BCS says Oklahoma had the toughest schedule, though all three had schedules in the top third of teams. BCS strength of schedule ("SOS") is two-thirds how well the teams you played do, and one-third how well the teams they played do. That's supposed to help correct for conference problems like Baylor. Baylor went 3-9 this year, indicating it was a weak opponent for OU and Texas, but the teams that beat Baylor included both OU and Texas (#1 and #6), #10 Kansas State, #37 North Texas, #40 Texas Tech, and #21 Oklahoma State. That's probably how OU got the #11 hardest schedule by BCS rules. LSU had #29, and USC #37. They're all strong - all three schedules are separated by a total of 3.6% in won-loss against opponents and opponents' opponents.

The BCS translates that to a ranking, though, and each rank in SOS adds 0.04 to your BCS score. Because USC finished eight places behind LSU, that added 0.32 more to USC's final score than to LSU's score. The final difference between the two was just 0.16 points, so if something had changed to bump USC up four or more places higher in SOS, it would be going to the Sugar Bowl. BCS expert Jerry Palm says that if either Notre Dame or Hawaii had won on Saturday, USC would have made the big show. It also would have happened if USC had ranked just one position higher in one of the six computer rankings.

But that didn't happen, and USC goes to the Rose Bowl, not the Sugar Bowl for the National Championship. This has the Asshat's panties in a wad, because USC is #1 and LSU is #2 in both polls. Oklahoma is #3 after the KSU loss, and the asshat is livid that the title game is not USC vs. LSU.

Poll protection

In being livid, though, he and his fellow asshatians fail to mention that OU didn't beat out USC for #2 in the BCS. OU was #1, by a fairly wide margin. The contest was between USC and LSU for #2, and LSU's schedule won it for them. Even if OU didn't get bonus "quality of win" points for having defeated #6 Texas, the gap between OU and LSU would still be more than twice as large as the gap between LSU and USC.

Why? One of the entire points of creating the BCS was to avoid the absolute incoherent ravings of the voters. Earlier this season, poll voters had TCU ranked as high as #6, despite the fact that TCU had only played one top half team all year. Forget about playing top ten teams; TCU didn't play top half teams. Purdue beat Iowa one week, and both teams won the next week, but the voters moved Iowa ahead of Purdue.

You may be thinking, "well, maybe Iowa was the better team overall, and they deserved a higher ranking." That's correct - but that's not how the polls are "supposed" to work. For decades, the unwritten rule has been that teams who lose their games that week fall in the polls, but high-ranking teams do not fall unless they lose, or unless a similarly-ranked team wins a big game. If you're #6 and you lose a game, you fall at least one spot even if you've previously beaten the former #7 and #8 teams. That's Just How It's Done.

Look at the three one-loss teams again. OU was #1 all season, and lost its final game to #10-ranked Kansas State. USC lost to #43 California in overtime in October, and LSU lost to #15 Florida on the same day. If all three of those losses had occurred on December 6, do you think USC would be ranked the highest of the one-loss teams? Do you think Oklahoma would be third?

Of course not. It's not even in question. OU had the harder schedule and beat higher-ranked teams like Texas, considered the third-best two-loss team in the country. OU's only loss was to the Big XII champions. LSU's only loss was to a top twenty team. USC had the weakest schedule of the three and lost to a Cal Golden Bears team that was never ranked higher than #41, was ranked #59 at the time, and was previously ranked as low as #70.

If OU had lost to Texas on October 11 (when LSU and USC both experienced their losses), there's no question OU would still be ranked #1 of the one-loss teams (presuming Texas found a way to lose somewhere else; let's not get too hypothetical). If USC had lost to Cal this past weekend, it would have dropped out of the top five, much less the top three, although I argue that's probably the wrong move. It's just How Things Are Done in the polls.

If USC had a stronger schedule, perhaps losing to a who-dat team wouldn't make the difference. If USC had lost to #16 Washington State instead of #43 Cal, they might be making Sugar Bowl plans. But the truth, painful though it is for USC fans, is that their team has the weakest record of the three one-loss teams, and that's why they're not playing for the championship, period.

The BCS rankings did exactly what they were supposed to do, and sportswriters like Asshat Bowden can't stand it. He fumed and ranted on the show that if the polls rate the same teams #1 and #2, the computers should be thrown out and those two teams should play for the title. Or, in plain English, "If sportswriters and coaches happen to agree, all other objective measurements should be discarded, because I know better than all of you."

Conference FUD

Bowden also whined that after 2001, when Nebraska snuck in as #2 after an embarassing loss to Colorado on the day after Thanksgiving, the BCS should have modified its system so that only conference champions can play for the title. Even the mighty Jesse Taylor of Pandagon says he believes that, saying, "Here's my issue - if you can't win your conference, you shouldn't be playing for the national championship. If you aren't the best team in your conference, how can you be the best team in the nation? All other considerations aside, this is like putting the loser in the NFC Championship up against the winner from the AFC."

Of course, Asshat and his east coast compatriots love this idea, for one simple reason: only the Big XII and the SEC have title games. The other BCS conferences - the Big Ten, the Pac 10, the ACC, and the Big East - all decide their champions solely by schedule. In Bowden's world, where Jesse has sadly taken up residence, the Big XII and SEC representatives would be ineligible for the national championship game unless they could win their last game of the season, everything else notwithstanding. USC, or daddy's Florida State team, would be able to clinch a championship, lose the last game, and possibly still make it to the championship game.

And this year is not like 2001, despite Asshat-enabler John Saunders saying so. It is not like Nebraska squeaking in at #2 despite not winning the Big XII championship because - and read this carefully:

Nebraska did not even win its division in 2001.

Nebraska was undefeated heading to Boulder on 2001.11.23, and Colorado had two losses, but only one of those was a conference loss, to Texas. Colorado drubbed Nebraska 62-36, in what was obviously a defensive struggle [sigh]. After that game, both Nebraska and Colorado had one conference loss each, and Colorado won the tiebreaker because it beat Nebraska head-to-head. Colorado was the Big XII North champion, and two weeks later, defeated Texas 39-37 for the Big XII Championship and the Fiesta Bowl berth. Nebraska did not lose the conference championship; Nebraska did not play for the conference championship.. This year, Oklahoma clinched the Big XII South on 2003.11.08, and lost the title game in Kansas City. A loss is a loss, but the situations are not the same.

When every conference has a title game, it would be fair to restrict the national title game to conference winners - a mini playoff, just like so many clueless "junior NFL" fans want. Until then, no. Or - since division championships are determined purely by schedule, just like conference championships in the conferences with no title game, it would be perfectly fair to require participants in the title game to have won their divisions and played in the title game.

Week by week

But Asshat and his enablers aren't saying that, because that would still rule Oklahoma in, and they don't want that. They've been forced to say nice things about Oklahoma all season because Oklahoma hadn't lost, and has won most games in disturbingly dominant fashion, such as beating #6 Texas 65-14 on the same day LSU and USC were saying, "wha' hoppen?" The BCS system was designed to look at an entire season, not individual weeks, and the sportswriters hate it. Hate hate hate hate hate hate it. It robs them of their power on a weekly basis to anoint someone and crush someone else.

For example, on ESPN's Sunday morning SportsCenter, Mark May was all over OU quarterback Jason White and how his performance should lose the Heisman Trophy. Most sportswriters weren't wild about White anyway, inventing any number of bogus excuses why an East-coast player (like Pitt's Larry Fitzgerald, a sophomore on a four-loss team) should get it instead. May used his national bully pulpit to say that White's performance should knock him out of the running, but that KSU's running back Darren Sproles did so well that he should be included in the top five Heisman ballots and have a real shot at winning the trophy, despite a season that could have been better (and his team's three-loss record).

Did you get that? One bad performance negates an entire record-setting season for Jason White, but one good performance overcomes any of Sproles' difficulties and puts him in the race. This is what sportswriters do and love - change the story every week. The polls have always done the same thing, and the coaches mostly go along because That's How It's Done. You lose, you fall.

The polls haven't made a serious attempt to distinguish between closely-matched teams for decades, and when all of those are at the top, it's a problem. That's why the BCS system includes seven computer rankings (with each team's lowest score discarded) and a strength-of-schedule component.

SOS looks at the whole year, by the way, not just the most recent week. That's pissing off King Kaufman of Salon, who says:

Let's say you beat a team that's 5-0 and ranked third, and the next week that team loses its quarterback, top running back and best defender for the season in an unfortunate car wreck, then loses the rest of its games. Your victory back at midseason gets less and less impressive every week, even though the team that's losing now is vastly inferior to the great team you beat. The system thinks you beat a 5-7 team.

This is an exaggeration, in the sense that it's never happened, though it could. The point is that the BCS is a system that lets the No. 3 team pass the No. 2 team while the No. 2 team is in the process of beating a 7-4 team in a major conference by three and a half touchdowns and the No. 1 team is getting beat like grandma's rug, all because of something that happened in the Boise State-Hawaii game. That's a bad system.

Kaufman manages to leave out that it works both ways. If you get beat by the #7 team in the country in week three, but that team goes on to lose six more games, you weren't beaten by a good team. The SOS calculation should take that into account, and it does, and that's why. Last year, OU played #11-ranked Iowa State in Norman, and beat the Cyclones pretty badly. It sent the team into a tailspin, and Iowa State lost four of its next five games, finishing 7-6 and ranked #43. What do you want to bet that if the rules went the way that Kaufman now wants, that a year ago he'd have been bitching that OU got SOS credit for beating an Iowa State team that had clearly peaked?

The BCS future

The sportswriters hate how the BCS system balances the entire season, and they're constantly pressuring the conference presidents to tinker with it. In the 2000 season, OU played massively-overranked Florida State for the national title in the Orange Bowl. All the sportswriters could talk about, though, was whether Miami would have a claim to a share of the title when FSU beat OU, since Miami had already beaten FSU. OU won the game 13-2, and it would have been a shutout had OU not taken a deliberate safety on a muffed punt late in the 4th quarter.

But nonetheless, at the end of the season, the BCS presidents tinkered with the formula and added "quality of win" points, improving your score if you beat a highly-ranked team. Had that been applied in 2000, the title game would have been OU vs. Miami, just like the writers wanted.

In the 2001 season, Nebraska completely gamed the system by booking eight home games against relatively weak opponents and winning handily. Nebraska's first real road test was at Colorado, and the Huskers tanked, but the strength of the computer rankings was enough to get Nebraska into the Rose Bowl to play for the title. Why? Because several of the computers used the margin of victory as a factor in ranking teams, and Nebraska tended to beat the crap out of its weaker opponents. Nebraska's toughest game of the year was against Oklahoma, but it was in Lincoln, and the vegetable peelers pulled it out, 20-10. Other than OU, Nebraska's highest-ranked opponent before Colorado was #29 Texas Tech. Nebraska averaged nearly 35 points per game at home, compared to just under 14 for the opponents. The Huskers outscored home opponents 278-110. On the road, it was 171-79, with 62 of those points coming from Colorado. Nebraska's only other road games were against its weakest Big XII opponents - Missouri, Kansas, and Baylor. But it worked beautifully - the computers kept Nebraska at #2 and got them to the Rose Bowl, where they were thoroughly embarrassed again. So, after the 2001 season, the BCS required that computer rankings no longer consider margin of victory. One ranking dropped out completely because its author refused to change his formula; three others added special BCS versions that don't consider point margin.

In 2002, everything worked well - there were two undefeated teams at the end of the season, and they were invited to the championship game. No adjustments necessary.

This year, they're already clamoring for a new rule that would have kept Oklahoma out. Some want it limited to conference champions, but Kaufman and others want a rule that a consensus #1 in the polls gets invited automatically. That would have made this year's title game OU vs. USC, but the sportswriters really want USC vs. LSU. Kaufman says that every "logical" way to rank the top three puts USC and LSU in the title game.

He's flat wrong, just like Asshat and his enablers are. By any measurement other than "did you lose that one game this weekend," OU is clearly #2 if not #1. If USC wanted a berth in the Sugar Bowl, it shouldn't blame Notre Dame or Hawaii or the computers - it should have beaten Cal, a team it clearly overmatched.

Extra demerits, however, go to USC coach Pete Carroll. When Asshat-wannabe John Saunders interviewed Carroll on the BCS selection show, he asked him how he'd get his team excited "since the BCS obviously treated your team so unfairly." There wasn't a damn thing unfair about it, but Carroll didn't take the high road. He said that the writers' poll is not obliged to vote the Sugar Bowl winner as #1, and since USC is #1 in that poll, it intends to stay there by winning the Rose Bowl and claim its national championship like in the old days, where finishing #1 in either poll gave you a share of the crown. (That's how OU claims its 1974 and 1975 championships - one poll only, since the team was on probation. Then, as now, the coaches poll does not rank teams on probation.)

That's the kind of crisp, logical coaching that got Carroll fired at New England after taking a Super Bowl team through three straight years of decline, losing six of its last eight games and finishing at the bottom of its division just before Carroll got the door. It's hypocritical as all hell, because this situation could easily have been reversed. It could have been USC vs. OU in the Sugar Bowl with LSU ranked #2, and LSU saying that if it won its bowl and USC beat OU, that LSU should be champions in the coaches poll. You think Pete Carroll would have said, "That's fine, we'd be happy to share the title?" Hell, no. He'd say, "This is the title game, and the winner gets the national championship trophy, and that's what we're playing for, and they can talk about whatever they want." Breathtaking hypocrisy that the football gods will not smile upon.

What's worse is that if it had gone the way that Asshat and Kaufman and that gang want, they'd be outraged that OU wasn't in the title game. "Here's a team that's ranked #1 all season, sets records, has dominated this sport like a Miami team of the 90s, and with just a single loss is being excluded from the title game in favor of teams with weaker schedules and more embarassing losses? It's criminal," they'd say. But that's what they want - to change the story every week. The BCS system prevents that, and that's why, at all costs, they want it to go.

It's not going to go. The Big XII, Big Ten, Big East, and ACC are all happy with the system. Without the BCS, Ohio State would have been in the Rose Bowl last year and could not have played Miami for the championship, since the Rose Bowl wants only Big 10 vs. Pac 10 matchups. The Big XII wouldn't have sent teams to the title games three of the past four years without help, as the Big XII champion would always be in the Fiesta bowl. The Pac 10 is less happy, thanks to this year and 2001's snub of Oregon, but the SEC is happy that LSU gets to play for the title.

The bowls make a lot of money from it and get tourists to their festivals (which is why most communities built bowl games in the first place), and NCAA university presidents do not want a playoff, or any other reason for people to treat NCAA football like a junior NFL. The endless arguments that schools would make playoffs work took a near fatal blow this month, as one-loss TCU passed on the GMAC bowl, the #2 bowl for its conference, because the game schedule would conflict with final exams. TCU opted to play its bowl in its home stadium, for much less money, after finals were over. Half the argument for a playoff is that it would let talented smaller schools have a chance, but TCU's academics-first attitude (one common at more schools than you'd believe) pretty much shuts that down.

Still, every time the BCS title game is not what the sportswriters demand, the BCS presidents meet the following April and tinker with it so that it would have produced what the sportswriters wanted. Kaufman's article says this is unsatisfactory because it's "reactionary" and not "proactive," but that's the entire point. The BCS could stand up and defend its system like I have here, but they won't, because then the sportswriters will turn on them even more.

The truth is simple: the BCS criteria involves winning and scheduling. If you schedule decent teams and win all your games, you're going to play for the national championship. When you start losing games, you relinquish control of your own destiny, and that's not the BCS's fault. It's yours.

The only way OU is not the best one-loss team in the country is if the primary measurement is when a team lost games. The polls have used that as the primary ranking for decades, but the BCS system was designed to not do that. It worked just as it should have, taking the turn-on-a-dime power away from sportswriters and the coaches who follow them. Until they get it back, or until the BCS makes people understand why relying on such whims was a bad idea, the sportswriters will demand blood every time they don't get their way.

Asshats.

Discuss
[ Print This Page ]