Bowl Championship Series
Entire books have been written about the BCS, so there's no need to re-litigate its history here. Instead, I'll highlight two areas: the computer rankings that comprised a part of it, and how some fans have managed to keep it "alive" in recent years.
From 1998 to 2003: Lowest Score is Best
These were the years of annual tweaks, adjustments, and refinements. [details coming soon]
From 2004 to 2013: Highest Score is Best
During the mature era of the BCS, [details coming soon]
Component 1: Human Polls
The Bowl Alliance (1992-1994) and Bowl Coalition (1995-1997) used only the Associated Press poll and the Coaches poll (USA Today/ESPN/various titles over the years) to create their overall ranking. The Bowl Championship Series added a third component, a computer average, beginning in 1998.
In 2005, the Associated Press and the writers pulled out of the BCS and were replaced by the Harris Interactive poll of former players and coaches and current and former writers (and maybe more?).
Component 2: BCS Computers
Before 2002, systems were allowed to include margin of victory in their calculations, but the BCS managers required removal of that feature from 2002 onward. Jeff Sagarin chose to create a new, "politically correct" system that did not include score information and continued to participate; however, David Rothamn declined to modify his code and withdrew.
| Computer System | 1998 | 1999-2000 | 2001 | 2002-2003 | 2004-2013 |
| Anderson & Hester (Seattle Times) | ✔️ | ✔️ | ✔️ | ✔️ | ✔️ |
| Richard Billingsley | ✔️ | ✔️ | ✔️ | ✔️ | |
| Wes Colley ("Colley Matrix") | ✔️ | ✔️ | ✔️ | ||
| Dunkel Index | ✔️ | ||||
| Kenneth Massey | ✔️ | ✔️ | ✔️ | ✔️ | |
| Herman Matthews/Scripps Howard | ✔️ | ✔️ | |||
| New York Times | ✔️ | ✔️ | ✔️ | ||
| David Rothman | ✔️ | ✔️ | |||
| Jeff Sagarin | ✔️ | ✔️ | ✔️ | ✔️ | ✔️ |
| Peter Wolfe | ✔️ | ✔️ | ✔️ |
Additional Components
In the early years, there were additional components that "adjusted" the human-plus-computer average. The formula in those years was "lowest total wins," and these additional components were:
- strength of schedule was added in the spirit of "penalty points," with more points added for weaker schedules. Examples:
- #1 overall national schedule = +0.04
- #2 = +0.08
- #25 = +1.00
- #50 = +2.00
- losses were also added as "penalty points," with one point added per loss
- (Commentary: this means that the managers felt like a loss was worth being dropped the equivalent of one additional place in the polls; pretty serious!)
- quality wins over a top team were subtracted as a bonus in a few years, under two different formulas
- In 2001, minus 1.5 for a win over #1; minus 1.4 for a win over #2; and so on
- In 2002-2003, minus 1.0 for a win over #1; minus 0.9 for a win over #2; and so on
The system was "pure" from 2004 onward, with none of these additional components used -- just human polls and computers.
| Computer System | 1998 | 1999-2000 | 2001 | 2002-2003 | 2004-2013 |
| Schedule Strength (#1 = 0.04; #2 = 0.08, ...) |
✔️ | ✔️ | ✔️ | ✔️ | n/a |
| Losses (+1 for each loss) |
✔️ | ✔️ | ✔️ | ✔️ | n/a |
| Quality Win Bonus (Over #1 = -1.5; over #2 = -1.4; ...) |
✔️ | n/a | |||
| Quality Win Bonus (Over #1 = -1.0; over #2 = -0.9; ...) |
✔️ | n/a |
BCS Rankings Today
Asher Feldman at @BCSKnowHow (Twitter account required, sorry), produces weekly "simulated BCS" standings as if the BCS were still in use. He uses the AP poll instead of the Harris Interactive poll, but otherwise follows the exact same methodology that was in place from 2004 to 2013. He provides great details and analysis, so you should check out his feed. For now, I'll only leave below his image comparing the final top 25 of the simulated BCS and the CFP committee at the end of the 2023 season:
Quick Links
• Want to create your own ranking?
• Want to browse other ranking systems?
• ...including this author's?
• More to add here...