Sunday, December 7, 2008

The BCS vs. the Old Poll System: Consensus or Controversy

Following up on the previous post, I thought it would be worthwhile to look more closely at the major polls and see just how frequently there was controversy regarding the crowning of a national championship. Prior to 1950, the only national poll was that sponsored by the Associated Press. The BCS came into being after 1997, so we'll focus on the years 1950 - 1997 to see if the polls provided consensus or controversy.

In the decade that was the 1950's, the two major national polls were the AP and the coaches' poll sponsored by the United Press (later the United Press International). During this decade, both polls agreed eight times on a national champion, disagreeing only in 1954 (AP choosing Ohio State, with the UP choosing UCLA) and 1957 (AP choosing Auburn and UP choosing Ohio State).

In the 1960's, both major polls agreed on champions nine times, disagreeing only once in 1965 when the AP chose Alabama and the UPI favored Michigan State.

The 70's saw the most controversy, with the AP and UPI agreeing six times and disagreeing on four. Split national championships were awarded in 1970 (Nebraska/Texas), 1973 (Notre Dame/Alabama), 1974 (Oklahoma/USC), and 1978 (Alabama/USC). In fairness, 1974 was the first year the UPI instituted the rule that teams on NCAA probation would not be ranked in its poll; otherwise, Oklahoma may have been the consensus in both polls.

The 80's was a decade clear of controversy, at least as far as the polls were concerned. In all 10 years, both AP and UPI polls agreed on their national champions.

In the 90's, the USA Today took over sponsorship of the coaches' poll from the UPI (1991) and for the eight years of consideration, 1990 - 1997, consensus was had in five. The AP gave national championships to Colorado in 1990, Miami in 1991, and Michigan in 1997 while the UPI/USA Today polls crowned Georgia Tech, Washington, and Nebraska respectively.

So, in a total of 48 years with two major polls, a split national championship occurred exactly 10 times, or approximately 20%. This means that in 48 years before the BCS, a consensus national champion was crowned without controversy about 80% of the time.

So how is the BCS's record on controversy in the decade in which it has existed?
  • After the 1998 regular season: Tennessee and Tulane were the only Div. 1-A schools to finish the regular season undefeated, but Tulane coming from Conference USA was not only rejected for the BCS title game, it was excluded from all BCS bowls. And even though Kansas State was ranked #3 in the BCS standings, it too was excluded from the four BCS bowls.

  • After 1999: Not too much controversy, with undefeated Florida State paired with undefeated Virginia Tech for the BCS championship. Third ranked Nebraska had a single loss by four points to Texas, which it avenged by beating the Longhorns in the Big XII title game by a convincing 22-6 score. Though the one loss would mean the Huskers had no shot at the national championship, many at the time believed they were the strongest team in the country as they dominated the defending champions Tennessee in a Fiesta Bowl that was not nearly so close as the 31-21 final score would indicate. And Kansas State again got the shaft, ranked #6 in the final BCS standings, but denied a BCS bowl bid.

  • After 2000: Oklahoma finished the regular season undefeated and was the clear #1 choice. Florida State, Miami, and Washington all finished with a single loss and clear cases to be matched against the Sooners. Florida State was ultimately chosen to face Oklahoma in the BCS championship game, even though their one regular-season loss was to Miami. And while Florida State lost to the Sooners, Miami and Washington each won their bowl games convincingly, adding to the controversy.

  • After 2001: Not only was Nebraska ranked #4 in both the AP and coaches' polls, it had not won its conference, having been blown out by Colorado in its last regular season game. Oregon had one loss and was the Pac 10 champion, but it was the Huskers stunningly selected to play Miami for the national championship. A 37-14 spanking by Miami in the Rose Bowl, combined with Oregon's 38-16 drumming of Colorado in the Fiesta Bowl left many convinced the BCS got it flat wrong.

  • After 2002: This was arguably the best year for the BCS thus far, creating a national championship pairing that could not have happened under the old bowl system. It took double overtime for undefeated Ohio State to beat previously-undefeated Miami in a game for the ages.

  • After 2003: This was arguably the worst year for the BCS thus far in terms of controversy. Oklahoma had a good regular season but was thumped by Kansas State 35-7 in the Big XII championship. The Sooners were loved by computers though and selected to play against one-loss LSU for the BCS championship, even though Pac 10 champion USC also had only one loss, and was rated #1 by both the AP and coaches' polls. And after winning the Rose Bowl against Michigan, the AP kept USC rated #1, creating a split national championship between USC and LSU - the very thing the BCS was created to avoid. The controversy spilled over into the final coaches poll, when three voters broke contractual obligations to vote USC #1 over LSU.

  • After 2004: another year filled with controversy over the selection of two teams to play for the BCS national championship. Fully five teams finished the regular season undefeated: BCS teams USC, Oklahoma, and Auburn, and mid-majors Utah and Boise State. Auburn was hurt by a significantly lower preseason ranking than USC or Oklahoma, but proved its worth winning the Sugar Bowl against Virginia Tech. USC crushed Oklahoma in the BCS national championship, and Utah cruised past Pitt leaving three undefeated teams following the bowls, and more question marks as to whether the correct two teams were paired for the championship.

  • After 2005: another clean year, as none would argue over the pairing of undefeated Texas and undefeated USC. Like after the 2002 regular season, this was a terrific matchup that could not have happened under the old bowl system. Had they played and won separate bowl games, Texas and USC would surely have shared a split national title.

  • After 2006: Ohio State finished the regular season undefeated, ranked #1 in the AP, and a clear choice for the BCS national championship game. Controversy came in the selection of the Buckeye's opponent, with Boise State having also finished the season undefeated, and four teams each with one loss. Ultimately one-loss Florida crushed Ohio State to win the BCS championship while leaving Boise State with its victory over Oklahoma in the Fiesta Bowl the only undefeated Div. 1-A school following the bowls.

  • After 2007: Strong finishes by ACC champion Virginia Tech and Big XII champion Oklahoma as well as an undefeated season by WAC champion Hawaii were overlooked in favor of two-loss LSU, which had lost its final regular season game. For the second straight year, Ohio State, ranked #1 in the BCS standings, lost to the SEC champion in the BCS title game.

  • After 2008: Oklahoma/Florida may very well be the correct pairing, but this year can be hardly considered free of controversy. In a monster Big XII South division, Oklahoma, Texas, and Texas Tech, each with one loss has a legitimate claim to a bid in the title game. Count the same for Pac 10 champion USC and Big 10 champion Penn State. Utah went undefeated in the Mountain West which arguably, top-to-bottom, was better this year than either the Pac 10 or Big 10. Throw in undefeated Boise State, and you have seven teams with reasonable arguments to participate in the BCS national championship game.
In eleven attempts, the BCS has had three relatively controversy-free years in 1999, 2002, and 2005. Significant controversy in the selection of two teams to face off for the BCS national championship has occurred in eight out eleven years, or roughly 73% of the total attempts.

National champions decided by polls saw consensus about 80% of the time. National champions decided by the BCS have seen consensus only about 27% of the time. A playoff tournament would be better than either, but if one were to compare the current BCS system to the old poll system, wouldn't one want the system with significantly less controversy in determining its national champion?



No comments: