The formula (not "he") puts equal emphasis on all wins and losses. If a team plays a weaker schedule, they have to beat those teams by more points in order to stay ranked near or above a team that plays a more difficult schedule. That's how it should be.
By taking into account margin of victory. Maybe it would make more sense to you if you understood how it works. Winning a game gives you a positive credit. Losing a game gives you a negative number. Those numbers are averaged to create your final rating.
Winning and losing still matters. The difference is that his system accounts for the degree to which a team wins or loses. Wins-only systems do not.
Where are you seeing those rankings? You keep mis-quoting them. I really don't think you understand what you're looking at.
If the other teams play competitively against that team and do well in their own OOC games, that is what you should expect.
Maybe not. However, statistics form a solid factual basis.
Right. That's why you brought up that mathematicians want to boycott the BCS.
This is like Saddam Hussein saying that the only election that is meaningful is one that gives him 100% of the vote. Just the opposite, a rating system is completely meaningless if the author has an output in mind already that it must calculate. The point of a rating is to measure some criteria to compare. Not to satisfy biases and other pre-conceived notions about what the data should mean.
Hopefully, you pick a thing or two up from whichever university you end up going to. I sincerely wish you luck there. We'll need it.
This is partially true. The way it works seems to be a mystery to most. I wouldn't have understood it at all had I not learned about it during the process of building my own rating program.
Those who are familiar with the math (and I'm not the only one) tend to agree with Sagarin's predictor type of rating. As was explained in a recent article the flaws of the current BCS formulas are used as an example of what not to do in elementary college math courses.
Vegas also uses Sagarin's ratings to set starting lines.
Then take a clue. Be happy expressing your opinion as it is and stop getting into the business of trying to claim more knowledge than others. Don't start wars you can't finish.
It does not discriminate against those types of teams because if those teams beat the high-scoring teams, their margin of victory score is averaged into their score.
If it does discriminate against Alabama, why does it rank them at #5, higher than both human polls and the BCS?
The human polls are a disaster. Coaches don't have time to watch the games they would need to in order to make informed decisions. I've personally emailed with voters from the Harris poll who had made unexpected selections, only to hear excuses along the lines of "I didn't have time to watch that game because my daughter graduated. I'll fix next week."
We simply don't need humans to make this call when the game of football already provides us with a concrete statistical goal: points. The programs and players deserve a rating system that is just as concrete as the rules of the game itself.
We don't vote on whether or not a team deserves 6 points each time they enter the end-zone.
This is a good example of when humans fail. It does damage a teams ability when they lose key players. But, their rating at the end of the season should still reflect a proportional average between what they did before and after that player was lost. Otherwise, you get a bunch of people trying to guess at how good a replacement player is that may never have played a college game in their life.
