It’s up to the hitter

This week, we’re going to continue our look at the length of plate appearances (previous installments: one, two, three). For the last few weeks we’ve been looking at whether or not pitchers have any control over how long a plate appearance lasts by looking at different groups of pitchers. Our basic conclusions so far:

  • The number of pitches per plate apperance is almost entirely a function of walk, strikeout, and in-play rates. If you want to estimate pitches per plate appearance, use Tango’s Pitch Count Estimator.
  • The difference in the length of plate appearances between the most hittable and least hittable pitchers is very, very small—amounting to no more than about five pitches per game.
  • Some pitchers, as grouped by the type and frequency of strikes thrown, do seem to induce hitters into making quicker contact, although the magnitude of this difference is about a quarter of a pitch per plate appearance between the most and least hittable pitchers.
  • Any advantage we credit to the more efficient pitchers pales in comparison to the fact that the efficient pitchers, on the whole, are also terrible pitchers.

    Now that we’re done with pitchers, let’s take a look at hitters. We’ll break down hitters’ plate appearances three different ways today and see if there are any differences. For the purposes of this study, we’re going to look at MLB hitters who logged more than 200 plate appearances last year—a sample of almost 350 hitters. For each measure we’ll look at the top and bottom 20 hitters (or around the top and bottom 6%). The three measures we’ll look at are:

    How they hit

    1) Percentage of plate appearances that result in a ball in play. We’ll call this measure “Contact.” Because Contact tells us what happens at the end of a plate appearance, it’s a results-oriented measure.

    Group        Value     Guys like...
    Top 20      >84.1%     Placido Polanco, Juan Pierre, Paul LoDuca, Kenji Johjima, Nomar Garciaparra
    Average      74.0%     Alex Rios, Derk Jeter, Dan Uggla, Miguel Olivo, Ken Griffey
    Bottom 20   <61.4%     Nick Swisher, Jason Bay, Travis Hafner, Pat Burrell, Adam Dunn

    2) Called strikes as a percentage of all strikes seen (called, swinging, foul, and in play). These are hitters who don't put a lot of swings on the strikes they see. We'll call this measure "Passive." Unlike Contact, this is an approach-oriented measure as it tells us about the evolution of the plate appearance and not simply its result.

    Group        Value     Guys like...
    Top 20      >36.2%     Jason Kendall, Kevin Youkilis, Scott Hatteberg, Joe Mauer, Pat Burrell
    Average      26.7%     Rich Aurilia, Raul Ibanez, Miguel Cabrera, Alex Rodriguez, Mike Piazza
    Bottom 20   <18.6%     Kenji Johjima, Wily Mo Pena, Moises Alou, Ivan Rodriguez, Vlad Guerrero

    3) Pitches per plate apperance. This is the bottom line—do guys who see a lot of pitches see a lot of pitches across all categories? Or do they only see a lot of pitches because they walk and strike out so much?

    Group        Value     Guys like...
    Top 20      >4.16      Mike Napoli, Jason Giambi, Frank Thomas, Travis Hafner, Adam Dunn
    Average      3.76      Alex Cora, Carlos Delgado, Albert Pujols, Derek Jeter, Jeff Kent
    Bottom 20   <3.32      Placido Polanco, Howie Kendrick, Yuniesky Betancourt, Nomar Garciaparra, Jay Payton

    Let's go straight to results, with a table that shows the number of pitches required for each type of outcome (walk, strikeout, hit or out on a ball in play) broken down by the three different measures.

                     ----Contact-----   -----Passive----   ---Pitches/PA---
    Result  Average  Top 20 Bottom 20   Top 20 Bottom 20   Top 20 Bottom 20
    BB       5.67     5.44    5.67       5.56    5.58       5.74    5.32
    K        4.81     4.71    4.98       4.91    4.71       5.05    4.62
    Hit      3.34     3.13    3.53       3.70    2.89       3.75    2.87
    Out      3.32     3.19    3.46       3.67    3.05       3.71    2.98
    All      3.75     3.38    4.16       4.07    3.37       4.29    3.23

    If you compare this table to its analogous table in last week's article, you will see that the differences among hitters are striking. Unlike pitchers, the differences between the top and bottom of each classification are quite large.

    The top contact hitters do everything quickly—walk, strike out, put the ball in play—as compared to the lowest contact hitters. One wouldn't guess that a bat-on-ball sort of hitter like Placido Polanco would strike out in fewer pitches than a three-true outcomes kind of guy like Pat Burrell—but he does. The majority of the hitters who come out on the bottom of the Contact measure are powerful sluggers whereas the top contact hitters are not, so it may just be that pitchers are more afraid to throw strikes to the low contact guys.

    The players who see the most pitches per plate appearance see more pitches across all categories, and the differences are not insignificant. Hitters who see a lot of pitches don't simply see a lot of pitches because they walk and strike out a lot. They see a lot of pitches because they wait to put the ball in play, and their walks and strikeouts require the pitcher to throw more pitches as well.

    The Passive classification is the most interesting one. Burrell and Jason Kendall coldn't be more different hitters when it comes to outcome, but both take a similar approach at the plate: wait for a strike you like, then swing. Our notions about Vlad Guerrero are reinforced as well; if he sees a strike, he'll swing at it. In the top and bottom of this category, we see only very slight differences in the length of plate appearances that result in a walk or strikeout. But the passive hitters take forever to put the ball in play—about six-tenths of a pitch greater than the least passive hitters in the league. That is an enormous difference—the largest that we've seen in any of our investigations.

    But is it really a surprise that the most passive hitters take forever to put the ball in play, whereas the least passive hitters will put the ball in play very quickly? Not really. What is surprising is that if we use this same measure on pitchers, we find basically no difference between the top and bottom:

                     -Passive Hitters--   -Passive Pitchers- 
    Result  Average   Top 20 Bottom 20     Top 30 Bottom 30
    BB       5.67      5.56    5.58         5.60    5.75
    K        4.81      4.91    4.71         4.75    4.88
    Hit      3.34      3.70    2.89         3.32    3.32
    Out      3.32      3.67    3.05         3.34    3.29
    All      3.75      4.07    3.37         3.78    3.76

    (Passive is just a label—I realize it doesn't make sense to call pitchers whose strikes do not induce swings passive. Also, I'm using the top and bottom 30 pitchers (about 6%) who threw greater than 200 pitches in 2006).

    So...what does it mean?

    Pitchers whose strikes are least likely to induce a swing are no more or less efficient than pitchers who whose strikes are most likely to induce a swing. For hitters, there is a big difference. What's going on?

    One might postulate that this phenomenon is consistent with the fact that the spread in skill for batting average on balls in play is narrower than for hitters. That is, the difference between the best and the worst hitters in terms of on-contact batting average is greater than the difference between the best and worst pitchers. When we dole out credit or blame for the result of a ball in play, the lion's share of goes to the hitter as opposed to the pitcher. Similarly, drawing out or quickly concluding a plate appearance appears to be more in the realm of the hitter than the pitcher. If you're of the oversimplified DIPS school of thought that the pitcher has no control over what happens once bat touches ball, think of this result as replacing "bat touches ball" with "ball leaves hand."

    The batter-pitcher matchup is at the core of baseball, and it isn't well understood. If there were a psychological component to baseball, surely it would be found there. We often hear about how a pitcher controls a plate appearance by spotting the ball and throwing off-speed pitches. But if the pitcher has little control over what happens once the ball leaves his hand, it may be more appropriate to talk about how a hitter controls a plate appearance. We see that different classes of hitters exhibit much greater variability in the length of plate appearance, controlling for outcome. That is, the number of pitches a hitter sees is not entirely determined by his walk, strikeout, and in play rates as is the case for pitchers. So at the very least, these results show that the hitter is more responsible than the pitcher for determining when a plate appearance ends.

    References & Resources
    Matthew Carruth and Sean Forman provided me with a lot of the data used to generate the results here. Other data were obtained from Retrosheet.

    A Hardball Times Update
    Goodbye for now.

    I would be remiss if I didn't mention Keith Woolner's study on historical pitch counts in Baseball Prospectus 2007 and a related discussion on The Book Blog.

  • Comments are closed.