Here’s the data for this week:

I get to pick for Dad again, so two shots…This week is back to normal a bit, last week the games were very predictable, and it basically would take perfect picks to win the league. Best bet looks like Philly upset over Oakland (18% picked, 55% win probability). As expected, the ANS win probability model, which has a few weeks of Michael Vick data built into its analysis and doesn’t know he’s concussed, picked Philly a lot higher (69%). It looks like Nick Foles will start for Philly, so I’m going to go with Philly all-in as one pick. The other games to consider are Buffalo over KC (12%, 40%), Houston over Indy (8%, 42%), and Cleveland over Baltimore (29%, 59%). ANS picks Cleveland very high (76%), but it doesn’t know about Jason Campbell starting instead of Brandon Weeden. Maybe that’s cause for the win probability to adjust up even more!
I’m going to go Cleveland with the 2nd all in pick. The only concern is 29% picked them, so it may not be enough contrarianess.