I will continue to track the Generic Points Favored metrics (GPF, oGPF, dGPF). There are also four new columns, corresponding to each team's probability of advancing to the next playoff round. The last column is just the probability of a Superbowl victory converted into Vegas-style odds (it's the "fair" payoff for a $100 bet).
The probabilities are actually calculated explicitly (there are only 32 ways each conference can play out, so it's not necessary to do a simulation).
The win probability for each game is calculated according to the following formula:
win probability = 1 / (1 + exp(line/7))
A negative line indicates the team is favored For the current week's games, I use the actual betting line. For the remaining playoff rounds, I first predict a point spread using each team's current GPF (Generic Points Favored). That predicted point spread is then converted into a probability, using the formula above.
As mentioned previously, the probabilities are calculated explicitly, rather than with a simulation.
The Superbowl favorite, as of January 5, is the New England Patriots, with a probability of 22.9%. Denver is right behind them, with a probability of 22.3% (New England's 1.5 GPF edge over the Broncos just outweighs the Broncos' home field advantage). The NFC favorite is the San Francisco 49ers, with a probability of 17.7% (the NFC teams are stronger on average than the AFC, effectively "splitting the vote" in the NFC).
According to the futures odds at 5dimes.eu, the favorite is the Denver Broncos, getting +290 odds (the "fair" odds, according to my probabilities, are +347). It is possible that the market has started favoring Denver more than what my rankings reflect. We won't know for sure until the lines for the Divisional Round games are released. I wouldn't be surprised if Denver's GPF jumps up.
Comparing my "fair" odds to the 5dimes futures odds, there are three positive ROI bets: the Patriots, Falcons, and 49ers.
Long shots continue to be horrible bets. The Colts are currently getting +6800 at 5dimes, but a fair payoff, according to my probabilities, should be +50000, nearly an order of magnitude higher.
Futures vs. Moneyline Bets
To get a sense of how bad that +6800 payoff is for the Colts, let's pretend you just bet the moneyline each week on the Colts, rolling over any profits into the next week's bet. The moneyline for the Colts/Ravens game is +255. The Raven are not the best team in the AFC, so a conservative estimate of the moneyline for the Colts' future playoff opponents would be +255 (I can guarantee it will be higher for either a Broncos or Patriots matchup).
So, a +255 payoff means that your bankroll multiplies by a factor of 3.55 for each win. The Colts would have to win four times to win the Superbowl, so your bankroll would multiply by a factor of 159. Meaning that a starting bet of $100 would yield an ending bankroll of $15,900. In contrast, the futures payoff of +6800 would give you an ending bankroll of $6,900, less than half of our conservative estimate. And, with the moneyline, you can cash out easily after each round. The only drawback is that you need to make a bet each week. I imagine that these long shot futures bets are most popular for the casual gamblers who aren't looking to manage a bankroll.
As with the regular season rankings, these probabilities will be updated daily to reflect game results and the latest line movements. Enjoy.