In the National League this year, just about one in ten fly balls turn into home-runs - the rate is 9.9%. However, that broad statistic covers a wide range of individual variation. Among major-league pitchers with 30+ innings of work so far, the rate varies from 26.1% [the Indians' T.J. House, down to seven who sit at 0.0%, having not yet allowed a homer, led by the Royals' Wade Davis who has thrown 648 pitches, and hasn't lost one yet. Both Brandon McCarthy and Addison Reed are in the top ten of the 315 qualifying pitchers: Reed is #6, at 19.6%, and even after keeping the ball in the park for the fourth consecutive game yesterday, McCarthy is still #8 at 18.8%,
Standard sabermetric theory says that pitchers don't have a great deal of control over what happens to the ball once it comes off a hitter's bat. If it's a ground-ball, will it be hit to a fielder or find a hole? If a fly-ball, will it pass the fence or drop harmlessly into an outfielder's glove? These kind of things aren't generally seen as being due to the pitcher, so if one has a BABIP (batting average on balls in play) significantly greater than league average, it's probably the case that they're being unlucky, and that number should drop eventually, taking things like ERA with it. The same goes for HR/FB%. So Addison Reed will improve once those fly balls stop leaving the yard. But is that inevitable?
Except, there does appear to be evidence that HR/FB% isn't all about luck, at least to the same extent as other peripheral numbers. Some parks are easier to hit balls out of than others, though it isn't necessarily the ones you might think (Coors Field is more triples friendly than home-run friendly). If you look at the numbers since 2005 for teams pitching in their home parks, the HR/FB rate ranges from 9.5% for the Giants in AT&T to 11.6% for the Pirates in PNC. That 22% difference between bottom and top, is higher than other batted-ball metrics. Line-drive rate for the same situation, varies less than 5% (between 19.6% and 20.5%) and BABIP has a 6.6% range.
We see something similar if we look at individual numbers. It's estimated that it takes about 600 fly-balls for HR/FB% to normalize, which is typically at least three seasons' worth of data. But even if we look at more than that, again since 2005, there is a broad range of home-run rates. With 500+ innings pitched, they vary from Brett Myers 14.2%, all the way down to Clayton Kershaw, at less than half that, 6.7%. Either Kershaw's beard does indeed belong to the luckiest leprechaun in baseball, fluking his way to those two Cy Youngs, or we need to accept pitchers can have a significant effect on HR/FB%. And if Kershaw can suppress it, presumably Reed can increase it too.
This makes sense: Watch the Home-Run Derby next week, and more than 9.9% of the fly-balls will leave the yard. That's because "bad" pitches, e.g. hanging breaking balls, cement-mixer sliders, fastballs "right down Broadway" (to quote Grace), are more likely to be souvenirs. Of course, not all home-runs are hit off mistakes - sometimes, you have to tip your cap, and those would appear to be more controlled by luck. But the more bad offerings thrown, the higher HR/FB% would seem to be. Think of it as a scale, with Kershaw at one end, and a Home-Run Derby pitcher at the other. Reed, particularly a one-pitch Reed, unable to locate his slider, has been closer to the latter.
Who you face is also a factor, since hitters do have more control over their HR/FB% than pitchers. Since 2005, with 1,500 PAs, you will not be surprised to hear that the three leaders are Ryan Howard, Giancarlo Stanton and Adam Dunn. The top 10 also includes the likes of Mark Reynolds, Prince Fielder and Albert Pujols. Hard to be sure if Reed has faced particularly tough batters: giving up a home-run to Troy-boy is hardly unique. On the other hand, he is responsible for the only home-run hit by Kevin Frandsen in over 200 at-bats since last August.
So, how much of Reed's struggles are his own, and how much are bad luck? [With all due respect to McCarthy, and I wish him all the best, he is now the Yankees' problem, so I don't really care whether he gets fixed or not] Looking at the difference between Addison in 2013 and 2014, a couple of things stand out. He has significantly moved away from his slider and change-up. Last year, those represented about two out of five pitches he threw. This year, barely one of five, with the change-up all but gone entirely. Also, the difference in velocity between his fastball and slider has narrowed, from 9.1 mph to 7.5 mph. In particular, his fastball has been generally trending down since April.
Personally, I think it's a little from column A, a little from column B. I have genuine concerns about Reed's overall effectiveness and the declining velocity, but I do think the results have been exacerbated by some bad luck, and I don't think the 20% home-run rate will continue in the second-half. However, how far the baseball gods let it descend, and how much is actually down to Reed, will remain a matter of concern, until proven otherwise.