Why are the Mets so Pythagen-screwed this season?  Normally a team's run differential -- the difference between the number of runs its hitters score and the number its pitchers give up -- can pretty reliably be used to extrapolate its W-L record.  Why not now?

Rob Neyer runs the "expected" standings using this method on ESPN:

http://sports.espn.go.com/mlb/neyer/index

And indeed, this tells us some things we already knew: the Phillies and Nationals are not as good as they've looked over the first 2/3 of the season, and the Nats especially are already enjoying a sharp and predictable fall back to earth over their last couple of weeks.

All season long, though, the Mets have been drastically underperforming this metric.  What's the story: bad bullpen pitching, lots of Met hitting in blowout games, or sheer bad luck?  Does Pythagoras bear the Amazin's no love?