When considering "Who has the harder schedule"........is it what a team was ranked when another team plays them, or where any given team is ranked at the end of the year- after all their games are played- that counts. Example: LSU played 4 Top Ten teams, on the road last year..........or did we? Only two were actually "Top Ten" in the final poll. Your thoughts??
If you are refering to strength of schedule in the context of a computer poll, it's virtually entirely based off of their end of the year result. If you are talking about in a poll voter's mind, who knows? Seems to be somewhere in between.
I'm pretty sure it's where the opponents are at the end of the season. I remember in 2003 that USC got knocked out of the national title game because their strength of schedule dropped when Hawaii, a team they bombed early in the year, lost its final regular season game to Boise State.
Ranking has no effect on SOS. Its a combination of your opponent's records and your opponent's opponent's records. Your opponent's records make up two thirds of the formula, while your opponent's opponent's records make up one third. For example, you play a two 6-6 teams, for a combined 12-12 record, it carries the same weight as playing a 12-0 team and an 0-12 team, also totaling a 12-12 combined record. You add all that up throughout the course of the season, and whoever played the schedule with the most combined wins, has the toughest schedule. This used to be, but is no longer an individual component of the BCS formula. It is, however, still used in some of the computer polls.
I understand all that and didn't mean in it terms of the BCS formula......just straight up. It does matter when making an argument for being "More deserving" of a title shot and in the AP polls, and for bragging rights. Actually, in those terms, it would make a difference wouldn't it?
Well, for bragging rights, ranking would be more important than combined record, I would suppose. The fundamental flaw in the SOS as calculated by the computers is that 99% of humans would agree that beating 2, unranked, 6-6 teams, is much easier than beating an 0-12 last place team AND and undefeated, 12-0, #1 team in the nation. The pollsters and talking heads will definitely give you more credit for the latter, while the computers rank them identically.
This is true about the old strength of schedule formula used in the BCS as the SOS component. However, there's the misconception that this is the formula that is also used by all the computers. This is not the case, and as a matter of fact, each of the computers has it's own SOS system. Only one computer poll has made it's formula public (Colley Matrix), but many others discuss some of the theory behind their rankings. At least one system (Peter Wolfe) factors road games into the mix. Richard Billingsley uses a combination of gametime ranking of a team and final ranking of a team in his strength of opponent rankings. His belief is that the hype around a game at the time (such as this weekends 2 vs. 9 game with VT, or 8 vs. 17 according to his rankings) should be noted, but also there should be a penalty if it turns out VT is overrated (i.e. they lose 4 or more games this year).
Well, just goes to show how much I've kept up with the BCS in the last few years. Ignore the man behind the curtain; there's nothing to see here.
