America at Large:In my youth, and in fact for most of my adult life, college football supremacy was an elusive goal invariably described as the "mythical national championship", writes George Kimball

Because college football involves hundreds of teams organised in dozens of "conferences", a conventional league system, with each team playing every other team, would be impracticable. And in the absence of a workable post-season play-off system akin to those in other sports, American colleges rarely settled matters of ultimate supremacy on the field of battle.

Teams would plod through a nine- or 10-game regular season, and if they were among the best dozen teams in the country they might be lucky enough to be invited to one of the five or six bowl games on New Year's Day.

Back then, collegiate gridiron success was tabulated by not one but two weekly polls - that of the Associated Press, whose electorate was composed of sportswriters and editors, and United Press International's, which maintained a panel of college football coaches. (With the demise of UPI, the latter poll was resurrected by USA Today.) Sometimes the two polls couldn't even agree, in which case two schools might each claim the mythical national championship.

Over the last half-century the number of post-season bowl games has proliferated to over 30, meaning just about any team with a winning record qualifies for one of the available slots. (Six wins has become the benchmark, meaning that in an age in which 12-game regular seasons are not uncommon, at least a few 6-6 teams will be playing in bowl games.) Considerably more than honour is at stake. A college team invited to a bowl game can expect a payout ranging from half-a-million dollars from a lesser bowl to the $17 million guaranteed to each of the schools invited to BCS games.

Usually, the $17 million is divided among member institutions' conferences, meaning the biggest post-season winner won't be Ohio State, which will share its spoils with other Big Ten members, or Florida, whose Southeastern Conference brethren will share in the booty, but 11th-ranked Notre Dame, which belongs to no conference and hence keeps all the $4.5 million the Fighting Irish will receive for playing LSU in a Sugar Bowl collision of twice-beaten teams.

Nine years ago, after several imperfect experiments, the NCAA and the operators of the largest bowl games huddled together to devise the Bowl Championship Series, which was supposed to settle these matters once and for all: the top 10 (eventually 12) teams would qualify for invitation to the more prestigious bowl games, with the pair adjudged the two best teams in the country matched in a National Championship game with an official trophy for the winner.

The BCS employs a complicated formula in which computerised ratings, the results of the coaches' USA Today poll, and a Harris interactive poll are accorded equal weight in determining rankings.

This system proved to be efficient only if there were two, and exactly two, qualified claimants at the end of a regular season.

(That is exactly what occurred last year, when Texas and Southern California were matched in the Rose Bowl, and produced a thrilling 41-38 game won by the latter in the final 19 seconds.)

The imperfections of the system were demonstrated two years ago when Auburn finished 12-0 but were voted out of a spot in the Championship Game, in which USC played Oklahoma in the Orange Bowl.

This year the foot is on the other shoe. No one disputes that Ohio State, having completed a perfect 12-0 season, have earned a right to play for the title, but the identity of the opposition became a subject of considerable dispute.

As potential challengers fell by the wayside, Ohio State held down the number one spot for most of the season, and by mid-October their traditional rivals Michigan had emerged as number two. When the two met in Columbus on November 18th, the game lived up to expectations, Ohio State winning 42-39.

Now, logically, the best team in the country should beat the second-best, especially on their own field, and using the odds-makers' traditional yardstick, they ought to do it by just about three points.

The defeat, nonetheless, relegated Michigan to third place in the ensuing week's rankings, as USC, who also had one loss, slipped past them to claim the number two spot. Michigan, their regular season completed, could only sit back and wait.

Then last weekend in California, USC were upset by crosstown rival UCLA, while in the Southeastern Conference Championship Game, number four, Florida, beat Arkansas 38-28. And when the final alignments were announced Sunday evening, Florida had leapfrogged both USC and Michigan into the opponent's role against Ohio State in the January 4th championship game in Arizona.

The computer, it turned out, had ranked Michigan number two but was overridden by the mortals, as the coaches and Harris voters opted for Florida. In the end the issue was decided by .0105 of a percentage point.

Some on both human-comprised panels had expressed distaste for a rematch. Michigan had already had their chance and come up short, they said. Why not give another once-beaten team a chance? Here's the problem: having lost narrowly on Ohio State's home field, Michigan might have a chance of beating them in a neutral venue. Florida probably couldn't beat Ohio State if they played them in Gainesville.

Former Michigan State coach George Perles admitted to having voted for the Wolverines in the Harris poll "because they're from the state of Michigan, and I just so happen to live there", while South Carolina coach Steve Spurrier, who held the same position at Florida for nine years, defended his equally provincial vote by proclaiming, "Heck, I'm a Gator!"

The controversy was only deepened on Monday when Ohio State coach Jim Tressel revealed he had abstained in the final week's voting. Realising that whichever way he cast what would have been the deciding vote he would be roundly vilified by the odd man out, he opted to recuse himself from the process.

Now, obviously, this whole issue could be streamlined if the NCAA adopted a post-season tournament, as it does in basketball, baseball, and 21 other intercollegiate sports. (Although, whether it was a 16-team, an eight-team, or even a four-team bracket, you'd still have wails of protest from whomever didn't get picked.)

College presidents and other educators have steadfastly resisted such a formula on the grounds it would interfere with studies and further prolong what was once an autumnal pursuit but now stretches from August into January. Given the NCAA basketball example (in which 65 teams participate in an elimination tournament that runs into April), the argument strikes one as both short-sighted and hypocritical.