Time to re-evaluate things.

Now, first things first: a man admits when he has made mistakes, and clearly, I have made a few here this year. While I still think my pick of the Giants was defensible, clearly Kansas City and Philadelphia are not good teams. It is a fools errand to stake your suicide pool life on bad teams. But more importantly, I have been using SYSTEMS approach to this when I should be using a MATHEMATICAL approach. We are changing gears and taking things up a notch. My mathematical approach will be based on the following basic assumptions.

The first: a team's scoring differential is a better description of its relative skill than its won-loss record. A MODIFIED won-loss record can be calculated using a team's scoring differential and Bill James Pythagorean expectation formula.

The second: the odds of a team winning a particular matchup can be roughly calculated based on the MODIFIED won-loss record using the log5 method, also developed by James.

What does this approach tell us about the matchups this week? Here, according to this approach, are the top 5 matchups to consider this week.

Teams used: Arizona, Detroit, Tennessee, Tampa Bay, NY Giants, Pittsburgh, New Orleans, Baltimore, Kansas City, Philadelphia

#1: Green Bay over Tampa Bay. Now, obviously, we don't need some fancy formula to tell us Green Bay is a powerhouse. What this does tell us is that they have the highest odds of winning this week, at 87%.

#2: New England over Kansas City (80%). And that's before we consider the fact that KC is starting their backup quarterback!.

#3: San Fransisco over Arizona (79%). The old system would scare us off this matchup because it is a divisional game. My eyes tell me Arizona has looked frisky. And since I still have NE and GB on my (admittedly marked up) board, I'll steer clear for now.

#4: Detroit over Carolina (74%). I've used Detroit, but it will be interesting to see how this plays out.

#5: Dallas over Washington (72%). Not only a divisional game, but a road game.

For what it's worth, the system predicted the Jets as a 61% favorite over the Broncos. Of course, some mitigating factors (short week for the Jets, road game for the Jets, Jesus for the Broncos) would have indicated that the Jets might be in more trouble than that.

Too close to mess with: Buffalo (58%) against Miami, Cleveland (51%) against Jacksonville, Baltimore (57%) against Cincinnati, Atlanta (50%) against Tennessee.

The pick: New England