From Polls to Playoff: A brief history of the evolution of how college football determines the national champion

by selmaborntidefan
August 1, 2014

If Walter Camp were alive today to see what his sport has become, he would no doubt be stunned beyond belief. What began as an amateur fitness program among the Ivy League schools in New England has evolved into a multi-million dollar enterprise watched by millions of fans every Saturday from late August until early January. Fans argue over which school is better among teams that did not play while simultaneously crowing about their school’s victories over the teams they did play. In fact, part of the never-ending argument of college football centers upon presumed great games that will never be played (e.g. the 1991 Washington Huskies against the Miami Hurricanes). That is part of the charm of the innocence of the game, and it is about to change forever.

2014 will see the first-ever formal playoff at the highest level of college football. Playoffs have been held for years among the lower divisions, but this will be the first time that the FBS (formerly Division I-A) presents a round of semi-final playoffs followed by a recognized championship game. Unlike other sports that have had playoffs almost from day one, college football has maintained a poll system, followed by a bowl system, and then a series of computer/human selections to determine the two most worthy adversaries in their post-season.

Whether this advancement is primarily positive or negative will not be known for perhaps a decade a more. But college football insists upon staying unique; rather than using the polls that have helped determine match-ups for over fifty years, a selection committee will determine whom they deem the most worthy opponents for the four-team playoff. This promises a level of controversy to be sure but then again, what would college football be without controversy? Controversy might as well be the middle name of this sport.

Over the next several weeks we will take a look back at how college football has determined their national champions, controversies that have influenced change, and consider both the results and ramifications that have come out of these controversies. There is no better place to begin than with the advent of the first Associated Press poll in 1936.

The Advent of the Poll Era

The first Associated Press college football poll was compiled and organized by Charles Woodroof, a former SEC director of media relations, in November 1934. Although this was the first actual poll, the AP did not conduct weekly polls until October 19, 1936. Much as the originator of baseball is not completely known, it is uncertain who precisely came up with the idea. Alan J. Gould, whose newspapers would eventually win 14 Pulitzer Prizes under his guidance, is credited with the invention, but some sources list Cy Sherman of The Lincoln Star as suggesting that Gould poll the sportswriters of papers affiliated with the Associated Press. While the AP was hardly the first selector (the Dickinson system existed in the 1920s while the Helms Athletic Foundation began in 1936 as well) or ranking system, it had the influence via weekly exposure in the newspapers and quickly became the gold standard. Polls prior to the AP were primarily regionally based, but the use of the AP ensured a national presence and proliferation (as well as eventual acceptance) of the polls as authoritative. The AP final polls were conducted at the conclusion of the regular season but prior to any bowl games. This would eventually cause an obvious problem in selection but nobody was overly concerned with this in 1936 because there was no reason to be.

Poll Controversies That Weren’t – And One That Was

The first recognized AP national champions of college football in 1936 were the University of Minnesota Golden Gophers, coached by Bernie Bierman. Minnesota got through the schedule with a record of 7-1 and was awarded the championship. There was no controversy at the time. It did not matter that the nation’s only undefeated team, Alabama, finished fourth. Minnesota was the champion, and that is how it was. Minnesota did play a somewhat more difficult schedule than Alabama did but there was an obvious problem anachronistically speaking: Minnesota’s loss was to another 7-1 team, the Northwestern Wildcats. How did Minnesota wind up ranked ahead of Northwestern? The Wildcats made the mistake of losing their season finale to Notre Dame, 26-6. The bias inherent in the polls at the time is amazing. In the AP poll of November 9, Alabama had one first place vote and a number four ranking. The Tide then beat 5-5-1 Georgia Tech by four points on the road – and somehow dropped four spots in the ranking to number eight. They also fell behind LSU, who boosted their ranking in the polls a week later by administering a 93-0 shellacking of SW Louisiana, an act which boosted the Tigers to number two. Despite being undefeated and playing a much tougher schedule, LSU’s September tie with 2-6-1 Texas hurt them more than Minnesota’s road loss to Northwestern, and Minnesota won the national title. The fact LSU lost the Sugar Bowl to Santa Clara is important to note for history but irrelevant; champions were voted on prior to the bowl games, which were considered exhibitions at the time. LSU was picked for the 1937 Sugar Bowl because they were considered the more guaranteed draw at the time. Remember, this was during the Great Depression and the interstate highway system did not exist.

Polls went on in this manner for the next twenty-eight years and there was nary a complaint about the champion being chosen at the end of the regular season. Part of this reluctance to criticize was likely based upon the fact that most bowl games or conferences had a “no repeat” rule for their conference champions. It was hardly fair to displace a national champion selected by sportswriters when that team had no control over the rules that prevented their attendance at the post-season exhibitions. Notre Dame, professing to maintain the integrity of amateurism, refused to attend bowl games until 1970. This failure to include the bowl games in the final analysis led to the first real poll controversy of the AP poll era.

It should be noted before continuing that any researcher with an open mind could easily find fault with the national champions selected in the first thirty years of the AP poll. For example, Pitt won the 1937 national championship but it’s fair to ask “why.” The University of California Golden Bears not only had the same record as Pitt did, but they also beat the undefeated Crimson Tide of Alabama in the 1938 Rose Bowl, 13-0. The Tide had only lost three games in five years at that point. However, because the polls were prior to the Rose Bowl that game was not taken into account, and Pitt won the title. In 1938, TCU somehow was selected over Tennessee, despite the fact that the Vols played a schedule twice as tough as the Horned Frogs and went undefeated as well. Ohio State won it in 1942 despite the fact that Georgia won two more games, the schedules were about even, and the Bulldogs knocked off UCLA, 9-0, in the Rose Bowl. Army’s selection over undefeated Alabama in 1945 – despite Alabama’s drilling of USC in the Rose Bowl – likely owed as much to their tougher schedule and immediate post-WW II surge of patriotism as it did to the exclusion of the bowl games in the final vote. This was simply the way the champion was chosen. And then in 1947 came the first evident problem in the system. This is an interesting case study because it was the beginning of a pattern. The moment a flaw was detected in the system, the rush to fix the flaw never seemed to contemplate whether or not the potential flaws in the new setup were even more egregious. In short, the AP was about to find out it could please nobody.

Coming out of WWII, a number of people turned to religion as well as football. This increased the popularity of “the” major Catholic football school in the nation, Notre Dame, a school already legendary with its Four Horsemen, Gipper, and Knute Rockne. The Irish immediately won three national titles in four years, but it was the second title along with the AP’s action that caused major controversy. On December 8, 1947, the AP announced Notre Dame had been voted national champion. It was hard to argue the selection because only two teams finished the nation with an unblemished record, Notre Dame and Michigan. (SMU and Penn also finished undefeated but with ties). The Irish ended their year with a 38-7 demolition of unbeaten and third-ranked USC in California. Notre Dame’s dismantling of the Trojans seemingly justified the national title. And it would have were it not for what happened next: Michigan went out to Pasadena and not only beat the Trojans, they did so even more impressively than Notre Dame had, waxing USC to the tune of 49-0 in what was essentially a home game for the Trojans. The Wolverines held USC to an amazingly low 133 yards, and the Trojans only crossed the fifty-yard line twice the entire game (once on a 73-yard drive in the second quarter that was most of their offense for the entire game). The call for a co-champion began early with a football writer named Pete Rozelle (yes – THAT Pete Rozelle) saying that his polling of the sportswriters in the press box showed overwhelming support for a co-championship. The legendary Grantland Rice said that Michigan was the best team he had seen all year. And the controversy grew from there.

Perhaps had it been any other school, this would not have been so big. This, however, was the mighty Michigan Wolverines, and an immediate outcry arose for the AP to crown Michigan co-champions along with Notre Dame. The argument of Michigan was simple: they don’t play bowl games and we do, we played the same toughness of schedule, and we beat three common opponents by more points. Notre Dame’s argument was even simpler: this is how the champion has always been determined, and we won – plus, we’re the defending national champions and we haven’t lost. In a strange twist of irony, Irish supporters also complained that Michigan had intentionally run the score up on USC (keep reading, it’s about to be very important).

But there was another element to the controversy: Notre Dame and Michigan had begun the year at 1-2. Michigan took the top spot from the Irish in the second poll (October 13, 1947) after three impressive victories: 55-0 over Michigan State, 49-13 over Stanford, and 69-0 over Pitt. Since Notre Dame had “only” beaten Pitt by a score of 40-6, Michigan’s accomplishment was considered more impressive so they moved ahead of the Irish. The two teams had one other common opponent, Northwestern. Michigan drilled the Wildcats, 49-21, while Notre Dame survived a 3-6 Northwestern, 26-19. The latter game has an amusing side note in that Notre Dame led, 26-12, and sent in quarterback Frank Tripucka to pass from his own five-yard line late in the game in a desperate effort to run up the score. Northwestern intercepted the pass and returned it for a touchdown. But one can hardly fault the Irish since the pollsters had already made it clear by moving Michigan ahead of the Irish that big scores would be rewarded. It did, however, make their later charge against Michigan running up the score reek of hypocrisy.

It is defensible that the voters moved two teams around in the polls that could not play when it certainly appeared one was vastly superior to the other. The problem actually began when these same voters put Notre Dame back at number one after the Irish pounded 2-5-2 Tulane, 59-6, while Michigan was beating 2-6-1 Ohio State by “only” three touchdowns. The Irish held on and were voted the championship.

So heated was the controversy that rather than defend their process, the AP made the exact same mistake the BCS would later make: they decided to change a system that had already been accepted, in place, and worked for years. (But at least the BCS didn’t try to change the system to “fix” previous results). The AP decided to hold a revote of the sportswriters. Given all the data – four common opponents, all beaten far more impressively by Michigan than by Notre Dame – the AP writers picked Michigan, with a final points margin of 226-119. Twelve voters called it dead even. For perhaps the only time in history, the mystique of Notre Dame did not influence the outcome of the poll. Except it did, because right after announcing Michigan won the second AP poll, the Associated Press suddenly decided that it was going to stick with the results of the December 8 poll and declare Notre Dame the sole national champion! It is theoretically possible for an organization to screw up even worse than that, but it’s hard to see how. If they were going to stick with the first poll anyway, why even have the second vote?

It should be noted for history that Michigan DOES claim the 1947 national championship as one of their recognized titles. The Helms Foundation split their title while the NCF gave their version to Michigan. So deep was the rift (despite nice words at the time) that the two schools did not meet one another again on the gridiron for thirty years.

The 1947 controversy notwithstanding, things proceeded smoothly for the AP for the next seventeen years. Multiple times teams won the AP title and then lost the bowl games, including three times in a five-year span (1950 Oklahoma, 1951 Tennessee, 1953 Maryland). A new selector appeared on the scene in 1950, the United Press. The AP countered by presenting, for the first time, a pre-season poll of rankings. Prior to 1950, these polls had not been released until after at least two or three games had been played. This new invention, a good idea that would eventually go haywire as well, gave the pecking order for teams at the start of the season. The other new invention, United Press, would prove in time to be a mixed bag.