BCS Years in Review: 2003, Nightmare of Split National Championship

Use your ← → (arrow) keys to browse more stories
BCS Years in Review: 2003, Nightmare of Split National Championship
Stephen Dunn/Getty Images

Part 6 of a series: Over the next few weeks, I will be reviewing each of the 16 seasons since the Bowl Championship Series came into existence in 1998. Here is a look back at who got lucky, who got robbed, what could've been, what should've been and other controversies of the day. The series will appear throughout December and January.

Part 1: 1998, A New Beginning for College Football

Part 2: 1999, FSU Ends Michael Vick's Quest for Perfection

Part 3: 2000, FSU-Miami Sows Seeds of Controversy

Part 4: 2001, Nebraska Fiasco Rocks College Football

Part 5: 2002, Controversy On-Field Mars Perfect Ending

---

The epic Miami-Ohio State showdown in the 2003 Fiesta Bowl, even with the attendant controversy at the end of the game, gave the BCS a huge sigh of relief. "The system works!" went the battle cry.

No, it didn't, and most certainly in the 2003 season, when the BCS was met with catastrophe—the very scenario the system was created to avoid: a split national championship.

In its first five years of existence, while there were disagreements and debates about certain teams' merits to be included in the championship game, there had never been a case where the BCS champion was deemed unworthy or not been crowned by the Associated Press, which maintained its independence.

But in 2003, it all happened. Going into the final weekend of the season, three teams were vying for two spots in the Sugar Bowl. USC had one loss—at Cal in triple overtime, 34-31. As did LSU—to Florida at home, 19-7. No. 1 Oklahoma was undefeated going into the Big 12 title game against Kansas State.

Even before the games were played on that final Saturday, word was that the Sooners would stay No. 1, even if they lost the game.

The computers favored Oklahoma by a wide margin, and since all other major conference teams besides USC and LSU had at least two losses, Oklahoma would not drop to lower than No. 3 in the human polls. Put it together, the Big 12 title game was a mere exhibition with very little riding on it.

And the Sooners played like it, getting pasted by Kansas State, 35-7. After LSU beat Georgia in the SEC title game and USC romped past Oregon State, as expected, the Trojans ascended to No. 1 in both polls, while the Tigers moved up to No. 2.

Which team would've won USC-LSU matchup in 2003?

Submit Vote vote to see results

In the penultimate BCS standings, USC had a comfortable lead on LSU (6.90 vs. 8.43). The Trojans were ranked higher in the human polls and computer rankings and also had better strength-of-schedule ratings. The expectation was that USC would play Oklahoma in the Sugar Bowl with LSU left to howl.

But then Boise State beat Hawaii in the final regular-season game of the season in the middle of the Pacific.

Say what?

Exactly right, that's what happened. Combined with Syracuse's 38-12 pasting of Notre Dame, the Tigers got enough of a boost to nudge out USC by .16 of a percentage point, getting Oklahoma as their date, while the Trojans were left with a Rose Bowl berth against No. 4 Michigan.

A confluence of events made this happen, and all of it came from BCS's meddling with its standings:

1. Eliminating margin of victory from all computer formulas: This completely unnecessary step was taken gradually, out of an irrational fear of teams running up the scores to impress computers. By 2003, all BCS computers had agreed to abide by this restriction, in some cases reluctantly.

Because of this, Oklahoma's four-touchdown debacle was weighed the same as a road loss in triple overtime. Since the BCS neglected to remove the voters' eyeballs, the polls appropriately knocked the Sooners down to No. 3. But the computers overwhelmingly still went for Oklahoma.

2. Keeping strength of schedule (SOS) as a component: Strangely, with all the tweaking in the first five years, the BCS never touched this.

First of all, it's hardly an objective tool, almost as arbitrary as the human polls. The formula being used was a poor imitation of the RPI (used by the NCAA basketball selection committee) and called for an absurd division of 25 to produce the SOS number. Besides, all computer rankings have formulas for strength of schedule of their own, so at the very least, it's redundant. 

What was the worst part of BCS standings?

Submit Vote vote to see results

3. If you can't beat them...tank them: Notre Dame was routed by the Trojans, 45-14, in South Bend, so the Irish returned the favor...by getting blown out at Syracuse on the final Saturday of the season. Notre Dame's loss dealt USC's SOS rating a fatal blow. That, combined with Boise State's win over Hawaii, another team beaten by USC earlier in the year, catapulted LSU over USC.

The Tigers, by beating Georgia for the second time in the season, saw their SOS rating jump from 54th to 29th, while USC's held at 37th. LSU edged USC by .30 of a percentage point in the SOS ratings, the difference it needed to seal its hold on the No. 2 spot.

 

For the record, USC played at Auburn and Notre Dame and played home games against Hawaii and BYU while LSU faced Louisiana-Monroe, Louisiana Tech and I-AA Western Illinois at home and Arizona on the road. Just how LSU's schedule could be considered to be among the top 30 in the country showed the flaw in the SOS ratings.

At the end, USC impressively defeated Michigan, 28-14, in the Rose Bowl and held on to the No. 1 AP ranking and a share of the national championship. In a rather sloppy and uninspired game, LSU defeated Oklahoma, 21-14, for the BCS title—though not without one last bit of drama.

Despite a mandate to vote for the BCS title game-winner No. 1 in the final poll, three coaches (Oregon's Mike Bellotti, Illinois' Ron Turner and South Carolina's Lou Holtz) broke the contractual agreement and cast their No. 1 ballots for USC.

Nick Saban's Tigers got their half of the title but were quickly forgotten as USC romped to the BCS title games the following two seasons. The general acknowledgement that the Trojans won "back-to-back" national championships in 2003 and 2004 left many LSU fans embittered for quite some time.

Final BCS Standings: 1. Oklahoma, 2. LSU, 3. USC, 4. Michigan.

Final 2003 BCS Standings
Rank Team Poll Avg. Computer SoS Loss Q-W BCS
1 Oklahoma 3 1.17 11 1 -0.5 5.11
2 LSU 2 1.83 29 1 - 5.99
3 USC 1 2.67 37 1 - 6.15

BCS Guru

Alternative Methods

Using post-2003 formula: 1. USC, 2. LSU, 3. Oklahoma 4. Michigan.

Using 1998-2000 formula: 1. USC, 2. LSU, 3. Oklahoma, 4. Michigan.

Likely four-team playoff: USC vs. Michigan; LSU vs. Oklahoma.

If there were another game, then USC-LSU would've been the natural "real" championship game. And it could have happened (read on).

 

Controversies

The Snub of Miami (Ohio): After losing the season opener to Iowa, 21-3, junior quarterback Ben Roethlisberger led the Red Hawks to 12 consecutive victories, beating Bowling Green for a second time in the Mid-American title game. Bowling Green, incidentally, lost only three times—twice to Miami and once at Ohio State, 24-17.

But despite a No. 11 ranking in the BCS standings, Miami never had a chance for an at-large berth. Even though the Red Hawks were "eligible," they received no consideration from the four BCS bowls and ended up finishing a 13-1 season by routing Louisville in the GMAC Bowl.

Until before the 2006 season, a non-BCS school had to be ranked in the top six for a guaranteed spot in a BCS bowl game. Under congressional pressure and with the advent of a separate BCS National Championship Game, the standard for "mid-major" schools were relaxed to guarantee a spot for any team in the top 12. Too late for Miami and Big Ben, though.

The Texas Shaft: Despite being ranked No. 5 in both polls, the Longhorns were shut out of a BCS bowl berth thanks to Kansas State's upset victory over Oklahoma.

With the Big 12's two slots already spoken for, Texas had to settle for the Holiday Bowl, where it was beaten by Washington State. But a year later, the 'Horns managed to avoid a return trip to San Diego by being engulfed in yet another BCS controversy.

Ironically, after being the BCS victim the first two years of its existence, Bill Snyder's Wildcats finally made its BCS bowl debut by earning the Big 12's automatic bid. K-State lost to Ohio State, 35-28, in a wild Fiesta Bowl.

The Extra Game: On Jan. 9, 2004, Ted Waitt, CEO of Gateway Computers offered a $31 million package for a national championship game between USC and LSU. Despite vocal support from both schools, the NCAA did not consider the offer.

2003 BCS Bowl Matchups
Bowl Score Attendance TV Rating
Sugar Bowl* #2 LSU 21, #1 Oklahoma 14 79,342 14.8
Rose Bowl #3 USC 28, #4 Michigan 14 93,849 14.3
Orange Bowl #9 Miami 16, #7 Florida St. 14 76,739 9.1
Fiesta Bowl #5 Ohio St. 35, #10 Kansas St. 28 73,425 8.5

* BCS Championship Game

BCS formula review: With no audible criticism of its formula—thanks to having two, and only two, undefeated teams in 2002—the BCS for the first time in four years maintained the same formula with only a slight adjustment to decrease the value of a "quality win."

This little-known adjustment actually threw some extra fuel on the controversy because LSU would've gotten an extra .40 points with its victory over Georgia, giving it a much more robust final lead of .56 over USC.

 

Final analysis: Just how close was the LSU-USC spread? Had one of the four computers that ranked LSU No. 2 and USC No. 3 switched places for those teams, the Trojans would've gotten the coveted No. 2 BCS slot by .01 of a percentage point.

Or, if the Tigers had not made the quantum leap from 54th to 29th in the SOS rating in the space of one week—let's say they finished 34th instead, then USC would've been ahead by .04 of a percentage point. Yes, had any of the teams that USC played won one more game or any of LSU's opponents lost one more game, then it would've been a different outcome.

But still not a just outcome.

Should margin of victory matter in computer ratings?

Submit Vote vote to see results

LSU was not undeserving of a spot in the BCS title game. That's not the issue at all. The "correct" result should've been a USC-LSU title game in the Sugar Bowl. The team that didn't belong obviously was Oklahoma, which failed to win its own conference after getting blown out in the Big 12 title game—in fact, a game that was not as close as the four-touchdown spread suggested.

An easy remedy could've been found, as early as 2001, after Nebraska somehow secured passage to the BCS title game without even winning its division, let along the conference.

The old bowl alignment was always arranged to match conference champions in the most prestigious bowls, and therefore a pre-requisite of winning one's conference should not have been unreasonable for teams vying to play in the BCS title game. In all, three nonconference champions played for the BCS title in its 16 years of existence.

But the BCS refused to insert this one amendment—throughout its entire history and extending into the new College Football Playoff.

After the catastrophe of 2003, when the No. 1 team in both polls was denied a place in its title game that resulted in a split championship, the BCS was forced back to the drawing board and smashed it up. A brand-new formula was concocted to appease an increasingly angry and skeptical public.

The formula overhaul, however, would not save the BCS from another controversy in 2004.

 

Follow on Twitter @BCSGuru

Load More Stories

Follow B/R on Facebook

Out of Bounds

College Football

Subscribe Now

We will never share your email address

Thanks for signing up.