Add new comment

PCA is calculated year to year because I don't have access to an automatically updating daily statistical archive that feeds me up to the second info like some of the big sites.  And I haven't had time to do the full calcs since 2006 for all of PCA and since 2007 for DNRA (hopefully you remember the basic ideas behind DNRA, if not all of the details).
I can calcualte it for individual people when I need to, though, and for fielding metrics, I have reasonably accurate back-of-the-envelop calculations for in-season analysis.
PCA's defensiv emodel starts with marginal runs saved.
Seattle has allowed 255 runs in 69 games.  The league is averaging 297 runs in that time span.  The replacement level used by PCA is .250 at the team level (and the basic linear assumption that this replacement level implies holds up very well for team-level analysis for the most part).  Using the standard PythagenPat equations, we expect the Mariners, if they have an average offense, to win at a .569 clip.  Meaning an equivalent offense would score 346 runs.  The margin for an equivalent offense is 162 runs.  So the Mariners as a team are 184 runs above the .250 margin for defense.
Then you figure out the pitching component using DNRA (if they had an average defense, they would allow how many runs with the same DIPS pitching line - tweaked to account for things like pitcher assists, park home run factors etc) and assume that the defense explains the difference.
The Mariners have a team DNRA this year (thus far) of 3.98 - good for 6th in the AL, heavily weighed down by Vargas and a mediocre bullpen, FWIW.  That means if the Mariners had an average defense, they'd have allowed 275 runs...not 255.  Meaning the team defense is 20 runs above average (good for 4th in the AL).
20 runs above average...how does that compare to the fielding margin.  Keep in mind that the margin for each half* (pitching accounts for about 60% of the defensive game...not the 70% that James estimated...at least as the numbers appear to tell me when I do the analysis on my own...I could be wrong, but that's where PCA is at the moment) of the defensive game is higher than the margin for the whole team.  PCA uses a .350 margin for pitching in isolation (meaning we're giving the pitchers about 8.9 wins above margin when you take thte total of 18.6 WAM and break out the pitching marginal credit vs. the fielding marginal credit) and a .350 margin for each of the defensive units except the catching position, whose margin is .400 - this is based on empirical analysis of the distribution of marginal unit winning percentages when calculated assuming the rest of the defense was average.
The catcher (Olivo) comes out barely above average so far (PCA does not yet have a sophisticated technique for trying to put value on game calling, though I have not given up hope on that front yet) and has thus far earned 0.9 or so defensive WAM (bare in mind again...my magin is lower than fangraphs' replacement level) and Giminez/Moore virtually no value.
The infield has been weighed down by Figgins' lackadasical play at third and Wilson's mental problems at second, but Ryan and Smoak are above average for their positions...the outfield has been led by Gutierrez/Saunders and, yes, Ichiro.  The 18.6 wins of defensive creidt got split about this way (based on marginal value):
8.9 wins to the pitchers (a bit of a lower estimate than fangraphs...DNRA is not as bullish as FIP and PCA gives less credit to pitchers for run prevention than the WAR system does anyway - assuming a larger role for defense)
0.9 wins to the catchers (no metric in PCA for pitching selection and game calling...though I have not given up hope for trying to measure that some day)
3.6 wins to the outfield
5.2 wins to the infield/pitchers for infield defense
back-of-the-envelop estimates for the fielding staff::
Ryan: +1.7 WAM (on pace for a very nice season in the vecinity of 4-4.5 defensive wins...puts him in contention for a gold glove, typically)
Figgins: +0.7 WAM (horrible year defensively, thus far...but may yet improve)
Smoak: +1.0 WAM (on pace to be in the top five in the AL in PCA defense at first)
Wilson: +0.2 WAM
Rodriguez: +0.5 WAM
Kennedy: +0.8 WAM (on pace, prorated to be a near-average defensive second baseman...not bad for an old guy)
The Entire Pitching Staff: +0.3 WAM (pitchers on this team are fielding rather poorly as a unit - especially Pineda, Bedard and the formerly outstanding fielding Felix Hernandez)
Saunders: 1.0 WAM (in about 1/3 of a season...not too shabby)
Gutierrez: 0.8 WAM (considering the partial playing time...very impressive)
Ichiro: 1.2 WAM (PCA still thinks Ichiro's on pace to best the typical average of 2.0 wins/162 in RF)
Langerhans: 0.3 WAM (not good in CF but very good in LF)
Bradley: -0.2 WAM (yep...he was that bad)
Wilson: 0.0 WAM
Peguero: 0.3 WAM
Halman/Carp: 0.2 WAM (in very limited PT)
These aren't precise...just based on play made distribution vs. league average broken down by position and the players at the position after the value is apportioned to the units.