The Mets have the 2nd worst Ultimate Zone Rating (UZR) in baseball, at -20.1 runs.  Only the embarrassing Nationals measure out worse.  This number jumps off the page because the Omar Minaya Mets have been strong defensively.  They were worth +27.1, +6.8, and +9.8 defensive runs in 2008, 2007, and 2006, respectively. Team ERA outperformed team FIP in each of these seasons, consistent with a team that plays above average defense.  We've seen that UZR correlates decently with FIP-ERA.  Based on how poorly UZR rates Mets defense this year, one would expect team ERA to underperform team FIP.  This is not the case, as ERA is once again better than FIP by 0.11.  Compare to the Nationals, whose ERA is worse than their FIP by 0.52.  It's always prudent to seek another opinion on defense statistics, so I looked at plus/minus (+/-) and Revised Zone Rating (RZR) to maybe figure out what the heck is going on here.

John Dewan's +/- is a play-by-play stat, calculated similarly to UZR.  It is available, with subscription, at BillJamesOnline.net.  Like UZR, it aims to measure how many runs above or below average a player is worth defensively at their position.  Pitcher and catcher defense is not measured.  From 2006-2008, both systems agreed that the Mets were above average in the field.  This year, +/- has the Mets at +5, a 25 run (or 2.5 win) difference from UZR.  Here is the breakdown between infield and outfield:

Both agree that infield defense has been poor, but the gap in outfield defense is ~19 runs.  Let's get a third opinion.

RZR is generally not held in as high a regard as UZR or +/-, but is nonetheless useful.  It doesn't measure defensive prowess on a runs basis and a detailed explanation can be found here.  A higher number is preferable with RZR.  Here is the RZR breakdown for Mets infield and outfield, with MLB average RZR for infield, outfield, and overall listed as well:

RZR generally agrees with +/- here, measuring the Mets infield slightly below average, the outfield above average, and defense overall above average.  The disparity between UZR and RZR regarding outfielders once again stands out.  R.J. Anderson noted at Fangraphs that UZR is still using Shea Stadium park factors for its calculations.  He proposed that this might be having an effect on Carlos Beltran's poor UZR this season, but ultimately decided that such a small sample isn't enough to make any conclusions.  Still, the 4 players who have played the most outfield innings for the Mets this season all rate worse per UZR than +/-.  It would be useful if a home/away UZR breakdown was available at Fangraphs, but it isn't (yet).  Also useful would be an analysis of these stats for all team's outfields to provide some context, but teamwide +/- data is not readily available.

Defense stats are inexact but always improving.  I agree with R.J. that less than half a season's data is not much to go on either.  However, it might be wise to take Mets OF UZR with an even larger grain of salt than usual, at least until Citi Field park factors are used in its calculation.

(Note: All stats through Sunday's games)