Connecting you to your favorite North Texas sports teams as well as sports news around the globe. The procedure in selecting which teams is where it gets a little confusing. There are two different types of teams AQ-teams automatic qualifying and non-AQ. Ten teams are chosen to participate in one of five bowl games. Bowl organizers are contractually obligated to host the champion of one of the five designated conferences.
Why these specific conferences? What about the others? Well, all conferences had an opportunity to earn automatic qualifications during a four-year evaluation during the seasons. The above mentioned were the five that met the criteria. The champion team from each AQ-conference will automatically earn a trip to one of the five BCS bowl games -- at least until , when their contract ends. The AQ conferences have contracts with bowl organizers and T.
The biasness of the BCS system creates an enormous challenge for any team in the seven non-AQ conferences to reach a championship game.
These standings are an intricate portion of the BCS. Each computer ranking uses different algorithms in its formula. Elo Chess is based only on wins and losses and where the game was played while Predictor further incorporates margin of victory. Sagarin is bound by the BCS to use only Elo Chess method in his rankings for picking a national title game BCS forbids using margin of victory in rankings , but he admits that the Predictor is more accurate in picking upcoming games.
He says on his USA Today page :. However, it is less acurate in its predictions for upcoming games than is the Pure Points, in which the score margin is the only thing that matters.
One of the most interesting quotes in the paper pertaining to Elo is:. Just like the BCS. In all seriousness though, the Elo Chess rating system is well established rating system that has been applied to many forms of competition over the years.
So, overall I generally approve of this computer ranking. And besides, Nate Silver has used Elo , so it must be valid, right? According to Wired. According to their website their rankings are distinct in four ways:. Unlike the polls, these rankings do not reward teams for running up scores. Teams are rewarded for beating quality opponents, which is the object of the game.
Posting large margins of victory, which is not the object of the game, is not considered. Unlike the polls, these rankings do not prejudge teams. These rankings compute the most accurate strength of schedule ratings. These rankings provide the most accurate conference ratings. Each conference is rated according to its non-conference won-lost record and the difficulty of its non-conference schedule. First of all, none of the BCS computer models use margin of victory in their formulas now because the BCS prohibits it.
Everyone else is doing it because they have to. This was true, prior to the BCS deciding that margin of victory could not be used in determining the rankings. They all let the data determine the rankings with no preconceived notion of strength this is at least true by the end of the season. The next two claims are simply subjective statements of superiority.
How are they measuring accuracy? There is just no way to factually back up these claims. In fact, Sagarin points out that his Pure Points model, which does incorporate margin of victory, is more accurate than his Elo Chess rating, which does not. So, by choosing not to consider margin of victory at the time when it was allowed in the rankings could possibly have made their rankings less accurate than the other systems that were considering margin of victory.
And the fact that they are using Excel makes me even more skeptical. So no one has any idea what they are doing. Finally, an article in Wired. And while mathematical rankings predate the BCS by decades, the new process promised to be the first system combining consensus opinion with numerical analysis into a single ranking. Sure, go ahead and call the human polls art, but calling the computer models science a pretty liberal use of the word. Actually, I take that back. In review, this computer model, whose methodology is not public, was created by two former students of the University of Washington who, by their own admission, created this system because they felt as if Washington was being snubbed nationally in the ratings and is endorsed by the Pac Bloom was amused that I had such a vast amount of information.
How I ended up getting in the BCS was probably because they were impressed with my research. He goes on to say :. My system is probably more different from the other computer systems. But my system is not purely mathematically based. My rankings are closer-ly related to human voters, an improved AP poll, if you will. It reacts to games more like a human voter but does it without biases like the name of team, the conference they play in, etc.
The core of my system is not something you see in most computers. Just to remind you, the BCS is using this guys rankings to decide which team gets to go to the national championship in football and, indirectly, what school gets the biggest cut of the BCS money pie.
Did I mention this is insane? Wes Colley received a Ph. He even has a very simple method for predicting elections and he got 49 out of 50 states correct this year. As a result of this, a mistake was found in the final BCS standings of the season because Wes Colley had failed to include a game between Appalachian State and Western Illinois.
Let that sink in a little bit. However, rather than criticize Colley, I think he should be applauded for putting his model in a position where someone could find and point out this mistake.
This means, that for all we know, the other five computer models could be constantly making mistakes. Or even just making stuff up entirely. How would we know? Anyway, Colley lists five advantages of his method on his web site :. Summarizing these points, results from the Colley Matrix are entirely based on performance in the season of interest with no past performance considered, strength of schedule plays a major roll, and only wins and losses are considered.
Neither margin of victory nor location of the game is factored into the rankings and the method uses simple statistical methods. When a friend starts talking about strength of schedule three teams removed, they are using the Anderson and Hester approach to college football rankings. He also takes into account location, wins, and losses. All computer results do not factor margin of victory in college football games per BCS rules.
Margin of victory is thought to influence bias through computer results as found in human polls. Once the six computers have produced their Top 25, the top ranking and lowest ranking are removed for each team. If Oklahoma State received one first place ranking, four second place rankings, and one third place ranking, the first and third place votes would be removed leaving four second place finishes per the computer listings.
In this example OSU has received 96 points after four second place rankings 4 x The total, 96, is then divided by a perfect score of for a final total. All statistics can be skewed but taking into account how well one team stops the run but not the pass or how another team excels with their passing attack would seemingly make the formulas more robust and interesting.
A Top 10 team may be able to run the table within their conference but could have trouble against another team outside of their conference due to schemes. For example the University of Houston is No. On a linear comparison the computers should be able to calculate how well or poorly Houston would perform against other Top 10 teams such as LSU, Oklahoma State, and Alabama. Part of what makes college football fun is the debatable stance each fan, pundit, or non-biased person has towards the best and worse teams in college football.
0コメント