Politics Foreign Affairs Culture Fellows Program

The BCS Would Have Ranked Florida State #3

Systems based on the discretion of small groups of experts will always breed controversy.

Credit: Wikimedia Commons

If college football selected teams for its playoff the same way that it used to select them for its national championship game under the Bowl Championship Series (BCS) system, then the Florida State Seminoles would be heading to New Orleans and a semifinal berth in the Sugar Bowl. Since college football now uses an “expert” committee instead of the old BCS formula, however, the Seminoles were left out of this season’s four-team playoff despite their perfect record.

Former Southeastern Conference Commissioner Roy Kramer was the architect of the BCS. Kramer felt strongly that the BCS needed to base its rankings on more than mere subjective opinion. Rather than simply using the writers’ or coaches’ polls, or the verdict of a subjective committee, Kramer insisted that the rankings include an objective element as well—computer rankings that couldn’t be influenced by the names on the uniforms and would instead rank teams entirely based on their success on the field during the season in question.


After a fair amount of trial and error through the first six seasons of the BCS, which did feature some controversy, the system was nicely simplified and had an error-free run across its final decade. The BCS delivered a largely controversy-free national championship game for ten consecutive years, from the 2004 season through the 2014 bowl games.

This year, however, controversy abounds—and the fact that the sport eschewed Kramer’s wise design and switched to rule-by-committee is a big reason why. Florida State went 13–0 as a member of the Atlantic Coast Conference, one of college football’s five major conferences.  Granted, the ACC was only the #4 conference this season—according to the Anderson & Hester Rankings, which I co-created with Chris Hester, and which were part of the BCS throughout its 16-year run. Still, it’s very hard to go undefeated, and in doing so the Seminoles earned the #3 spot in the Anderson & Hester Rankings—behind #1 Washington and #2 Michigan, which both went undefeated versus tougher schedules.

Two of the BCS computer rankings no longer publish the version of their rankings that the BCS used. (Sagarin and Massey now only publish rankings that incorporate margin-of-victory, which the BCS didn’t allow after the 2001 season because winning by large margins isn’t the object of the game.) Using the remaining four BCS computer rankings—Anderson & Hester, Billingsley, Colley, and Wolfe—as well as the AP and coaches’ polls, I’ve computed what the BCS standings likely would have looked like going into this year’s bowl games. (The AP poll was used for several years by the BCS before being replaced by the now-defunct Harris poll, which was very similar to the AP in results. Also, the highest and lowest computer rankings for each team were dropped by the BCS, leaving the middle four. Since the highest or lowest could have been Massey or Sagarin, I computed the computer average both with and without dropping the high and low rankings, and split the difference.)

Each poll made up one-third of the BCS formula, and the computers made up the other third. Here’s how the estimated BCS standings would have looked this season:

1. Michigan (13–0): #1 in the polls (.992), #2 in the computers (.960), .981 total


2. Washington (13–0): #2 in the polls (.962), #1 in the computers (1.000), .974 total

3. Florida State (13–0): #4 in the polls (.876), #3 in the computers (.910), .887 total

4. Alabama (12–1): #5 in the polls (.865), #4 in the computers (.875), .868 total

5. Texas (12–1): #3 in the polls (.879), #5 in the computers (.825), .861 total

6. Georgia (12–1): #6 in the polls (.816), #7 in the computers (.770), .801 total

7. Ohio State (11–1): #7 in the polls (.765), #6 in the computers (.810), .780 total

8. Oregon (11–2): #8 in the polls (.713), #8 in the computers (.720), .715 total

9. Penn State (10–2): #10 in the polls (.617), #9 in the computers (.660), .631 total

10. Mississippi (10–2): #11 in the polls (.605), #10 in the computers (.630), .613 total

So #1 Michigan and #2 Washington would have been separated by just .007 in the BCS standings, with #4 Alabama and #5 Texas being separated by that same slim margin. Michigan would still be playing Alabama in the Rose Bowl, as will be happening on New Year’s Day under the current system, and Washington would still be playing in the Sugar Bowl that night. But instead of playing Texas (#3 in the committee’s rankings), the Huskies would instead be playing Florida State (#3 in the BCS rankings). The Seminoles would have finished #3 in the computer part of the BCS rankings, and #4 in the poll part of the BCS rankings, yet they finished #5 in the committee’s rankings.  

It’s also worth noting that if we were back in the pre-playoff, pre-BCS days, the top two teams—Michigan and Washington—would be playing a de facto national championship game in the Rose Bowl.

When college football expands to a 12-team playoff next season—at the partial expense of its glorious regular season—it should use the BCS formula to select and seed those teams. Otherwise, what happened this season will continue to happen. Subjectivity breeds bias, which in college football always seems to work against teams from conferences other than the Southeastern Conference and the Big Ten. With the committee in charge, this pattern will continue.  

The committee admitted that it docked Florida State because the Seminoles lost their starting quarterback late in the season. Yet the Seminoles beat archrival Florida on the road two weeks ago with a backup quarterback. Then they beat Louisville (currently #18 in the Anderson & Hester Rankings and #15 in the committee’s rankings) this past weekend with their third-string quarterback. Yet the committee acted as if beating a top-15 team (in the committee’s own estimation) on a neutral field was no accomplishment at all.  

In the rare moments when it tries to ground its subjectivity in something more concrete, the committee appears to use a poorly designed strength-of-schedule rating published by ESPN that apparently takes into account whether teams play weak Football Bowl Subdivision (FBS) teams (the 133 teams who vie for bowl bids) but not whether they play teams that aren’t even in the FBS—like when Alabama played Chattanooga three weeks ago. As a result, Alabama’s strength-of-schedule was rated much higher than it should have been—5th-toughest—when in reality it was more like the 23rd-toughest, as it is rated by the Anderson & Hester Rankings (which don’t give teams a pass for playing non-FBS teams).  

Indeed, the teams with the eight toughest schedules in the country according to ESPN’s strength-of-schedule ratings are all SEC or Big Ten teams that played a non-FBS opponent. Only one of those teams’ schedules (Michigan State’s) is ranked as having been one of the ten toughest schedules by the Anderson & Hester Rankings.  

Washington, which certainly played one of the nation’s toughest schedules (and didn’t play any non-FBS teams), is the unanimous #1 team across the four BCS computer rankings. Indeed, the Huskies finished with the highest pre-bowl rating in the 30-year history of the Anderson & Hester Rankings. No other team over the past three decades played as tough of a pre-bowl schedule as Washington’s and managed to go undefeated. The Huskies survived 10 games in the nation’s top conference (the Pac-12 edged the Big Ten and the SEC for that honor in the Anderson & Hester Rankings), handing #8 Oregon its only 2 losses along the way. In the final game in the storied history of the Pac-12 Conference—which was raided in mercenary fashion by its former conference ally, the Big Ten, and will cease to exist next year—the Huskies beat the Ducks in a thriller on Friday night, 34-31.  

Which invites the question: Couldn’t we keep the Pac-12 and ditch the committee?