tag:blogger.com,1999:blog-38600807.post5725409090401185323..comments2018-06-02T14:19:34.554-04:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): The Patriots and the Conjunction FallacyUnknownnoreply@blogger.comBlogger4125tag:blogger.com,1999:blog-38600807.post-30072870587364263752008-01-09T10:27:00.000-05:002008-01-09T10:27:00.000-05:00Actually, although you forgot to consider the vig,...Actually, although you forgot to consider the vig, your assumptions are correct and maybe even strengthened more when the vig is included.<BR/><BR/>There is usually a high price (vig) to pay when looking at future odds (probabilities) from the wise guys. Look at the current odds to win the AFC for the 4 remaining teams. (Odds are from vegasinsider.com)<BR/><BR/>NE - 1 to 8<BR/>IND - 5 to 2<BR/>SD - 8 to 1<BR/>JAX - 10 to 1<BR/><BR/>These translate into probabilities of .89, .29, .11, and .10 respectively.<BR/>The sum of these probabilities is 1.39 from which one can only conclude that the vig for this set is 28% -> (1.39 - 1)/1.39<BR/>Decreasing the probabilities by 28% would result in adjusted probabilities of .69, .21, .08, and .07 if the vig was distributed evenly for each team.<BR/>However, I believe, as you suggested, that there is an over reach of play on the Pats and therefore more of the vig is probably slanted towards them - less for others. The increased vig on the Pats would adjust their probability to win down further. Similiarly, the vig for the NFC is 22%.jaginmahttps://www.blogger.com/profile/01982619125717380440noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-55835769833108161342008-01-09T09:53:00.000-05:002008-01-09T09:53:00.000-05:00j holz-No, he used the method to suggest a possibl...j holz-No, he used the method to suggest a possible mismatch between the spread and over/under point total, and the tradesports win probability. But he did raise the possibility that the method may not hold when one team is a strong favorite.<BR/><BR/>jaginma- You're right. My bust. And that would make almost everything I wrote wrong.<BR/><BR/>But assuming the vig is equal for both bets (the 0.60 to win the SB, and the 0.69 to win the AFC), wouldn't they would cancel out in the math:<BR/><BR/>(0.60*vig)/(0.69*vig) = (0.60/0.69) = 0.87 for the SB individual gameBrian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-50154045088513211092008-01-09T07:54:00.000-05:002008-01-09T07:54:00.000-05:00I may be off base, but when the odds makers set a ...I may be off base, but when the odds makers set a money line for a game, it also includes a vig. Thus, although a 9 to 4 odds may produce a probability of .69 without a vig, a 10% vig would result in a .76 to .77 probability while a 15% vig is .81 probability.jaginmahttps://www.blogger.com/profile/01982619125717380440noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-86056172657802755832008-01-09T04:52:00.000-05:002008-01-09T04:52:00.000-05:00Maybe I'm not reading it right, but doesn't Phil B...Maybe I'm not reading it right, but doesn't Phil Birnbaum decry that exact method of calculating win% in the post that you linked to?<BR/><BR/>I think this weekend's moneyline is a much better indicator of NE's chances of winning, and they're 7-1 favorites there. But I do agree that the futures lines offer an arbitrage opportunity.j holzhttps://www.blogger.com/profile/13428814047654767163noreply@blogger.com