Earlier, Jeff:
[I]f Charlie Furbush is unusually homer-prone, then that's bad... I know it's strange to talk about someone being "unusually homer-prone" since a lot of writers tie themselves into knots trying to find exceptions to batted ball theory that mostly don't exist, but it isn't out of the question that Furbush could have a problem. Pitchers don't share the same level of ability to prevent fly balls from leaving. Maybe Furbush is below-average in that regard.
Maybe he is, Jeff. Maybe he is. We'd like for him not to be because then he would be a pitcher likely to allow fewer home runs. Since he pitches for the Mariners, allowing fewer home runs is a goal we prefer. Dingers! actually only applies to our hitters. Jeff went on to pose two questions following from the notice of Furbush's continued elevated home run rate. Is it a fluke? Observationally that is difficult to prove or disprove, but there is possible supporting evidence from a step forward in the baseball community's understanding of home run rates.
Mike Fast: The other big issue that I have is that directionality of air balls matters a lot, and pitchers have been shown to have a repeatable skill for this. Pulled air balls go out of the park quite frequently, and opposite field air balls almost never do.
In the linked post, Jeff took that quote and looked at Michael Pineda, noting that Pineda appeared to limit the number of pulled batted balls quite a bit. However, Jeff didn't have the data available to do a systematic evaluation. I now do thanks to a pitch-by-pitch database and in light of the Charlie Furbush discussion highlighted above, I went investigating.
I don't have perfect data for this, nor did I have the patience to construct an optimal model for the data that I do have. I'd prefer to do something in the mold of a three-zone classification of bucketing batted balls into either pulled, up the middle or the other way. However, the data is on a xy-coordinate grid and what's needed is a radial one, like a radar map would have. Therefore, in want of a quick method, I merely drew a line from home plate through second base and split batted balls into either pulled or pushed.
The MLB average (2007-11) for balls landing on the pulled side has been 58.3%. For the probable starters of the 2012 Mariners' rotation, here are their rates.
| Pitcher | Batted Balls | Pull% |
|---|---|---|
| MLB AVERAGE | ---- | 58.3 |
| Felix Hernandez | 3223 | 60.1 |
| Jason Vargas | 1708 | 61.3 |
| Kevin Millwood | 2643 | 54.1 |
| Hector Noesi | 178 | 57.3 |
| Blake Beavan | 376 | 60.1 |
| Charlie Furbush | 269 | 67.3 |
That's potentially good news about Kevin Millwood I suppose. It's worth pointing out that Texas, Baltimore and Colorado have three of baseball's most home run friendly parks. Some figures from now former Mariner starting pitchers include Michael Pineda at 55.5%, Erik Bedard at 55.9% and Doug Fister at 59.2%. Charlie Furbush stands out here for his very high rate of pulled batted balls, albeit with a head-shakily small sample size. In fact, over the period from 2007 to present day, Furbush has the fourth highest pull% among pitchers with at least 250 batted balls.
Furbush's exceeding high pull rate helps to shed light on answering that first question on the matter of Furbush's home run rates. The higher than average rates do seem likely to be partially a product of something that Furbush does. Keep in mind that pull% itself likely needs a decent amount of regression as well. However, there's a more important, but unanswerable, second question. Will that continue in 2012 and beyond? That's as unknowable as just how long past the use by date I can really get by with those eggs in my fridge.