November 18th will mark the 10-year anniversary of Voros McCracken's revolutionary notion that pitchers have little control over the outcome of balls put into play against them.  Once met with unbridled skepticism, McCracken's theories have grown in popularity over the past decade to the point where people who use stats like WHIP (walks plus hits per innings pitched) are laughed out of analytical circles for not using stats like FIP (Fielding Independent Pitching ERA).

But every so often, we see pitchers that serve as exceptions to Voros' rule, particularly in small sample sizes.  Take Rich Harden's first career start against the Colorado Rockies as an example.

No player on the Rockies' roster had ever faced Rich Harden before Wednesday's game.  When that happens, most people will tell you that the advantage goes to the pitcher.  After the first inning, it certainly seemed that way; Harden used 13 pitches to retire the first three Colorado batters, each on swinging strikeouts.

Garrett Atkins also struck out swinging to begin the second inning, but only after a 14-pitch at-bat that wore down the easily-fatigued Harden.  The Rockies would go on to collect five hits, draw four walks, and score four runs against Harden before he left the game with a whopping 92 pitched thrown in three innings.

But here's the weird part:  Harden struck out eight batters in nine innings, and the other out that he recorded came from an outfield assist.  That means that the Rockies had a batting average on balls in play of 1.000 against Harden (4-for-4, as one hit was a home run).

There are two ways to interpret this, of course.  One is that Harden's strikeout ratio had no correlation with his success, meaning that McCracken was full of it.  The other is that even though Harden had dominant stuff, once he allowed balls to be put in play anything could happen.

Interestingly, the night before, Javier Vazquez experienced a similarly "unlucky" start.  He fanned a dozen batters in half a dozen innings, but nevertheless allowed three runs on five hits in a losing effort.  Unlike Harden, he did not allow a homer and walked just one man. His defensive-independent pitching would prognosticate far better results than a 4.50 ERA, but a 4.50 ERA is what he got for that start.

The thing is, Vazquez has this sort of thing happens to him all the time.  The only time he has had a higher FIP than ERA since 2004 was in his renaissance 2007 season, and even then, it was only .18 runs lower.  His other FIP-ERA scores were -.08, -.46, -.90, -.86, and -1.69 so far this year.  You could blame his defense or his ballpark, but Vazquez is pitching for his fourth team in six years.  It may be time to acknowledge that Vazquez is an exception to Voros' tenets.

Harden's anti-McCracken start, on the other hand, appears to be an aberration.  His FIP-ERA scores from 2004-2008, splitting 2008 between his time in each league: -.26, .39, .05, 1.51, .55, 1.33.  While Vazquez consistently underperforms given his excellent peripheral stats, Harden consistently overperforms his already-terrific peripherals.

So what can we take from these back-to-back starts?  Is Javy's bad luck continuing while Harden's good luck is starting to even out?  Has Harden simply benefited from more favorable pitching environments since 2004?  Is there something about Harden that allows him to suppress hits on balls in play while there is something about Vazquez that fosters them?    

Nearly 10 years after Voros McCracken revolutionized the way we think about pitching and defense, we still do not have completely satisfactory answers to these questions.