By overwhelming popular demand, we at Barking Carnival will publish a basic BCS primer. This primer is not as interested in the question of why the BCS exists in the first place, simply with explaining the mechanics of the standings and selection rules.
Components
There are three components to the BCS standings. The first is the SIDs' and Team Managers' Coaches' Poll. The second is the Harris Poll. The third and final component is the computer rankings. A team's scores in all three components are averaged to determine the team's final BCS score for that week.
Coaches' Poll
The Coaches' Poll consist of Division 1-A head coaches or whatever employee or coworker they designate that week. For the 2008 season, there are 61 voters in the Coaches' Poll. A team's Coaches' Poll score is determined by taking their total points in the poll and dividing by the maximum possible points. The maximum is simply 25 points per voter multiplied by the 61 voters for a total of 1,525 points. Therefore, improving your ranking by one place on every voter's ballot is worth .0133 BCS points. Improving your ranking by one place on a single voter's ballot is worth 0.0002 BCS points.
Harris Poll
The Harris Poll is the second human poll, commissioned through Harris Interactive by the BCS after the AP Poll requested that they no longer be used in the BCS calculations. There are 114 panelists in the Harris Poll, and usually they all vote. Occasionally there will be fewer than 114 votes in the weekly poll, and the calculations account for this. Points are scored in this poll in the same manner as the Coaches' Poll. Therefore improving your ranking by one place on every voter's ballot is also worth .0133 points here, but because of the larger number of voters, improving your ranking by one place on a single voter's ballot is worth only .0001 points.
Computer Rankings
The final component, the computer rankings, is further divided into the six individual computer ratings systems. In alphabetical order, these systems are the Anderson-Hester, Billingsley, Colley Matrix, Massey, Sagarin, and Wolfe systems. While the final three systems listed offer ratings sets that use the actual scores of the games in their calculations (or formerly offered them in Wolfe's case), the BCS has opted for the politically correct option of using only wins and losses with no score data allowed to be used. Sagarin's ELO_Chess column, then, is the rating that is used for the BCS system.
A team's computer score is determined by first throwing out the high and low computer score for each team. The remaining four computer rankings are then added (25 points for a #1 ranking, etc.) and then divided by 100. Therefore, improving your team's position one spot on each computer is worth .040 BCS points. However, improving your team's ranking by a single spot on a single computer is only worth .010 points if that computer was counted both before and after the change. It's important to note the difference here between the human polls and the computer polls; it doesn't matter how close two teams are in a computer system, only ordinal rank matters.
BCS Bowl Selection Criteria
The selection procedures are available on the BCS website, but let's go through them in order, which is important.
1.) #1 and #2 in the final regular season BCS standings play in the National Championship Game.
This is the first rule and the most important. No matter what else happens, this rule is always followed.
2.) The champions of the ACC, Big East, Big Ten, Big 12, Pac-10 and SEC conferences will have automatic berths in one of the participating bowls after the 2008 through 2013 regular seasons.
And here is the first point of confusion. It was long assumed that because of the well-known rule that only two teams from any conference could go to the BCS that if a conference managed to have the #1 and #2 teams but neither as champion that the conference champion would miss out on the BCS. However, it was recently stated that this is not the case and it's important to note that the two teams per conference rule is not mentioned here. Therefore rule #2 is followed as well no matter what.
3.) The champion of a non-BCS automatic berth conference will qualify for a BCS bowl if they are in the Top 12 of the final standings, or if they are in the Top 16 and are also ahead of any BCS conference champion. However, only one team may qualify under this rule, and if more than one meet the criteria then only the highest ranked one will go to a BCS Bowl.
4.) Notre Dame qualifies if they finish in the Top 8 of the final BCS Standings.
5.) If there are still any at-large spots open after rules 1-4, then the #3 team in the final BCS standings automatically qualifies for a BCS bowl as long as there aren't already two teams from their conference in the BCS.
This rule does include the two teams per conference exception. But not the BCS conference champion rule. This and rule #6 were instituted after Kansas State got left out of the BCS in 1998.
6.) If there are still any at-large spots open after rules 1-5, then the #4 team in the final BCS standings automatically qualifies for a BCS bowl as long as there aren't already two teams from their conference in the BCS.
At-Large Eligibility
Generally speaking, a team must have nine wins and be in the Top 14 of the BCS standings to be selected as an at-large team. There are rules dealing with cases where this doesn't provide the BCS with enough eligible teams, possibly via teams with only eight wins being in the Top 14, but that hasn't happened yet so I won't cover it here.
Bowl Selection Procedure
Finally, there is the actual selection procedure. After the top two teams are in the championship game, the conference tie-ins are accounted for. The ACC is tied into the Orange Bowl, The Big Ten and Pac-10 to the Rose Bowl, the Big 12 to the Fiesta Bowl, and the SEC to the Sugar Bowl.
When one of the bowls with a tie-in loses their champion to the championship game, they select first from the at-large teams and automatic qualifiers. If two teams lose their champion, then of course the bowl that lost #1 selects first. After this is taken care of, the order of selection for the rest of the bowls changes from year-to-year. The next two seasons are shown below:
January 2009 games: Fiesta, Sugar, Orange
January 2010 games: Orange, Fiesta, Sugar
And that's how it works.