Seaweed: Your post is very interesting, but a lot of the information you provided seems irrelevant to me. Are you honestly suggesting that the NFL is going to manipulate the outcome of a game that is heavily scrutinized by millions of fans at the risk of ruining the integrity of the game so that a slightly more popular and more marketable team will win the superbowl? Do you really think this is a reasonable long term marketing strategy. If so, how come low profile teams win major sporting events so frequently? Do you genuinely believe that unemployment statistics influence the outcome of the superbowl game? If I were to research weather, demographics, topography, and various other random statistics, I'm sure I could find many variables correlated with prior superbowl success. A trend isn't impressive unless it is actually the CAUSE of the correlation. If the team with more characters in its name is 6-0 when playing on Wednesday when Oprah weighs < 145 lbs and the dow industrial average is greater than 12,650, do you want to bet on that? Do you really think that The patriot's coaching staff needs a petty sense of personal revenge to get motivated for the superbowl? How can you be "92%" confident in this bet unless the line is several standard errors away from a fair line? Even if you legitimately thought the patriots should be favored by 17, you wouldn't have a 92% chance to cover (the standard error in point differential is about 14.4, so the probability of covering would be only about 84%; source:
http://www.stat.auckland.ac.nz/~iase...45_Everson.pdf)