Did the BCS Rankings Work?
In their second year of existence, the Bowl Championship Series rankings
took shots from a few cynical sportswriters and skeptical broadcasters,
but when the dust cleared, the final BCS rankings called for the same national
championship matchup as the polls: Florida State versus Virginia
So, did the BCS rankings serve college football well in 1999?
They certainly did; in fact, it is hard to imagine anything serving it
AndersonSports has long recognized that the BCS has a dual function
to fulfill: it must answer the calls of justice (Which two teams
deserve to play for the national championship?) and of politics (Which
two teams do fans and the sports media want to see play for the national
championship?)—and the two are not necessarily synonymous.
This season, the overwhelming consensus was that Virginia Tech should
be Florida State's opponent in the Sugar Bowl, host of the 1999-2000 national
championship game. The BCS reflected this consensus, but it did not
do so by accident; it did so by design.
The BCS rankings are designed to correct only those errors in the polls that
are relatively egregious. The polls' ranking (11-0) Virginia
Tech over (11-1) Nebraska was not such an error. While the
Cornhuskers were, in the opinion of these pages, more deserving
of a Sugar Bowl invitation than were the Hokies (see Why
Nebraska Was #2 in the Pre-Bowl Rankings), the two teams' accomplishments
were relatively comparable, and a reasonable argument could certainly
be made on behalf of either—or, for that matter, on behalf of #4
(in the Anderson & Hester/Seattle Times Rankings) Alabama.
In truth, it was noticeably less clear which team should be invited
to play Florida State than it was that Florida State was far ahead
of the field: the gap between the accomplishments of #1 Florida
State (11-0 with a nation-leading 8 wins versus the top-50) and
#2 Nebraska was larger than the gap between those of Nebraska and
#8 Tennessee. Thus, faced with two relatively comparable teams
in Virginia Tech and Nebraska, the BCS rankings called for the matchup
that most college football fans wanted to see. If the polls
(the best barometers of fan sentiment) had gone the other way, if
both had ranked Nebraska ahead of Virginia Tech, then the BCS would
instead have extended a Sugar Bowl invitation to Nebraska (Nebraska
would then have beaten out Virginia Tech by a tally of 6.42 to 7.12
in the BCS rankings), and again the BCS would have given the fans
the game they wanted.
But the BCS rankings are not merely a reflection of the polls; they
are an improvement upon them. In the final regular season rankings
of 1999 (the only rankings that really matter), in literally every instance
in which the BCS differed from the average of the A.P. and coaches' polls,
the BCS was right and the polls were wrong. In opposition to the
polls, the BCS was right to rank Texas and Texas A&M ahead of Southern
Miss; Penn State ahead of Minnesota and Marshall; Kansas State, Tennessee,
and Alabama ahead of Wisconsin; and Alabama ahead of Tennessee (Alabama
and Tennessee were tied in the polls).
Let us examine two of these instances in more detail. As noted,
in relation to the polls, the BCS leapfrogged Penn State over Minnesota
and Marshall. Leaving Marshall aside for the moment, Penn State went
9-3 versus the nation's 14th-toughest schedule; Minnesota went 8-3 versus
the 58th-toughest schedule. The Nittany Lions played 7 top-50 teams,
including 2 top-10 teams; the Golden Gophers played 5 top-50 teams, none
ranked in the top-10. Aside from their head-to-head matchup (a heroic
upset victory by the Gophers), each team played 4 games versus teams ranked
between #11-50: Minnesota went 1-3; Penn State went 4-0. Bringing
Marshall back into the discussion, the Thundering Herd went undefeated
but played the 2nd-easiest schedule in the country (#113 out of 114).
Marshall played only 1 team ranked in the top-75, 6 teams ranked in the
bottom-20 (#95-114), and one 1-AA team. The evidence overwhelmingly
suggests that Penn State was the class of this trio: BCS 1, polls
Alabama was ranked ahead of Wisconsin by the BCS, but behind by both
polls. The Crimson Tide went 10-2 versus the nation's 6th-toughest
schedule; the Badgers went 9-2 versus the 66th-toughest schedule.
No bowl-bound team played more than 8 games versus the top-50—except for
Alabama, which played 10, double the number played by Wisconsin.
Alabama played 4 games versus the top-10; Wisconsin played 2. Alabama's
4 weakest opponents were #73 LSU, #52 Vanderbilt, #44 Auburn, and #43 Houston;
Wisconsin's were 1-AA Murray State, #113 Ball State, #97 Iowa, and #91
Cincinnati, to which Wisconsin lost. How the polls could rank a team
that won a nation-leading (tied with Florida State) 8 games versus the
top-50, that won a nation-leading (tied with Nebraska) 3 games versus the
top-15, that beat otherwise-9-1 Florida twice and went 10-2 against
a schedule that would have swallowed most teams whole, behind a
9-2 team that avoided Penn State and Minnesota in conference and lost to
Cincinnati out of conference, is beyond us: BCS 2, polls 0.
(Further comparisons between the BCS rankings and the polls yield further
victories for the BCS.)
In its first two years in operation, the BCS has given college football
fans the national championship matchups they want, while improving upon
the rankings of the polls. In the process, it has ensured the ongoing
splendor of college football's regular season—the most meaningful regular
season in all of sports; and it has preserved intact the richness and tradition
of college football's bowl games—the most unique events in all of sports.
Has the BCS worked? Like a charm.