Investigating Foul Balls as a Skill or Signal of Skill

It's often said that good hitters have an ability to protect the plate and extend at-bats by fouling off pitches. We've all heard some version of this, typically during broadcasts and almost certainly during sequences where a hitter manages to work a pitcher by fouling off a large number of pitches. 

A few folks have looked at this claim, in some shape or form. 

Dan Fox, writing at Hardball Times in 2005, looked at a number of claims regarding hitter talent, one of which was fouling off pitches. Looking at data from 2000 to 2004, Fox found little evidence that fouling off pitches makes hitters more successful.

A few years later, John Walsh--also writing at Hardball Times--took a deeper look into "bat-handling" or "place-hitting" skills. He specifically looked at whether hitters showed an ability to foul off pitches with two strikes. He found little evidence for the skill, noting that while there was a 50% increase in the rate of fouls per pitch between two strikes and less than two strikes the effect was largely due to the increased swing rate of hitters when the are in a two strike count.

The most robust analysis that I could find looking at whether hitting foul balls was a skill as well as an indicator of "outcome" talent (e.g. actual production by hitters) was conducted in 2008 by Pizza Cutter (otherwise known as Russell Carleton). Cutter found that foul ball rates seemed to be a consistent skill (i.e. something that players replicated over time), and that high two-strike foul ball rates tend to correlate with decreased strikeouts, walks, and home runs.

Knowing all this I decided to update things a bit, looking at batter data from 2008 to 2011 for hitters with >250 plate appearances in each season.

Here's what I found:

• The Year-to-Year correlation of Foul Ball Percentage (FB%)* was .72, which is pretty good when we think about hitting statistics.** It suggests, like Pizza Cutter's study, that the ability to hit foul balls is a skill or talent that hitters generally replicate season to season.
• The ability to hit foul balls does not necessarily indicate that a hitter is more skilled. The correlation with Weighted On-Base Average (wOBA), while positive, was only .23. Similarly, the ability to get on base (OBP) had next to no relationship with FB% (.079). The one batted ball outcome that did show the strongest correlation was power. Isolated power (ISO) managed to correlate at .45 and HR/FB at .47. This makes sense as strikeouts showed the strongest overall relationship (.74) and we know that hitters that display greater power tend to have higher strikeout rates.
• The more curious finding was that contact percentage was negatively correlated with FB% (-.63). So the better a hitter is at making contact, the less foul balls they hit per swing. On the one hand you might think that hitting more foul balls would go hand in hand with contact percentage. But, as it turns out, that's not the case. Overall, better contact results in less foul balls. This is consistent with what Pizza Cutter found when he broke down foul balls into different counts. In one and two-strike counts, the correlation was also negative. However, it was positive for two-strike counts.

Here is the overall correlation table for the study:

In terms of leaders in this areas, Adam Dunn had the highest FB% over the time period with 57%. Some notable hitters on the lower end of the spectrum include Vernon Wells (40%), Jose Reyes (40%), and John Jaso (37%). You can find the full list here.

This refresh does not have anything to say about two-strike counts. For that, I would refer folks to the two studies above by Walsh and Cutter. Also, Brian Mills looked at the question using PitchF/X data.

Overall, however, the results are consistent with what others have found--that while the ability to hit foul balls appears to be a talent that some hitters have and others less so, that ability only seems linked to hitter power, but not overall skill or ability.

----------

* All foul ball data courtesy of StatCorner. Foul ball percentage is the number of foul balls per swing, not per plate appearance.

** Unlike Pizza Cutter, I did not run an intra-class correlation here. I simply looked at the correlation from Year 1 to Year 2 of FB% for each hitter season in the sample (N=409).