The benefit of batting eighth

Baseball’s conventional wisdom suggests that batting eighth in the National League—that is, in front of the pitcher—has its pros and cons. With an extremely weak hitter on deck, opposing pitchers would seem less likely to throw strikes. To a patient No. 8 hitter, this is a benefit: He can simply wait it out and take his base. Less-patient hitters, however, can end up swinging at bad pitches.

There has been more talk than usual in 2008 about managing the pitcher’s spot in the lineup. This week, the Cardinals and the Brewers are facing off in St. Louis, and both Ned Yost and Tony LaRussa are slotting their pitcher in eighth. In this case, the wisdom about batting eighth must be shifted to the No. 7 spot, which may mean that a better hitter (in the case of the Brewers, J.J. Hardy) is facing the challenge.

What I want to know is: What effect does batting in front of the pitcher have on a hitter?

How to measure the effect

To know how batters handle batting in front of the pitcher, we need to have a point of comparison. In other words, we need a sample of players who have a reasonable number of plate appearances both batting in front of pitchers and batting in front of non-pitchers.

Players like that aren’t hard to come by. Not only do managers shift players in and out of the 8-hole, but even a batter who starts in that spot every day will regularly hit in front of a pinch-hitter. And, presumably, pitchers use a different strategy with Daryle Ward on deck than they do with Jake Peavy standing there.

With that group of batters in hand, we can determine, in the aggregate, how they manage batting both in front of a pitcher and in front of a non-pitcher. The effect turns out to be dramatic.

The algorithm

For today, I limited myself to play-by-play data from 2007. Because it’s hard to rack up very many at-bats in front of the pitcher, I used a low cut-off: I included a player if he had 20 or more plate appearances with the pitcher on deck. (I’m ignoring tactical shifts, such as the pitcher standing on deck and then being pulled for a pinch-hitter, or vice versa. If a pinch-hitter came in, I’m assuming that everyone knew a pinch-hitter was coming in.)

The vast majority of those players had more plate appearances in front of non-pitchers than in front of pitchers. To take just one example, Jack Wilson had more PA batting eighth than elsewhere, but he had only 182 PA with a pitcher on deck against 353 PA with a hitter on deck.

To isolate the effect, I used the typical approach employed in measuring Minor League Equivalencies: I equalized the number of PA. So, for Wilson, I reduced his 353 PA in front of non-pitchers to 182. This step keeps any one player from disproportionately affecting one of the two pools (“in front of pitchers” and “in front of non-pitchers”).

The results

Here is how our population compares on a few key statistics:

                  AVG    OBP     SLG     Stk%    P/PA
In Front of P     0.267  0.330   0.399   61.3%   3.67
In Front of Other 0.230  0.302   0.338   62.8%   3.74

There are quite a few surprises here. I expected that the on-base percentage would be much higher in front of the pitcher. It is, but not for the reason you might expect: Walk rate is actually a little higher without the pitcher batting behind them.

Also, the similarity in P/PA suggests that, as a whole, these hitters don’t change their approach when batting eighth—or, at least, there is no standard different approach that all hitters take in that spot.

The difference in strike percentage is the one comparison that wholeheartedly supports the conventional wisdom. That gap is meaningful, and it confirms that batters see fewer strikes when the pitcher is coming up. The spread might not be as dramatic as some assume, though.

With all these resemblances, how do we explain the huge differences in AVG and SLG? These wouldn’t be powerful hitters anywhere in the lineup, but they still manage to alter their performance quite a bit. One contrast that may provide a clue is strikeout rate: In front of the pitcher, these players strike out 14.1% of the time, whereas elsewhere, the corresponding number is 17.1%.

My guess is that two things are going on. First, pitchers are looking past the No. 8 hitter to the weak batsman on deck, so they are putting themselves in more hitter’s counts, thereby cutting down strikeouts and allowing more balls in play.

Second, in low-leverage situations with the pitcher on deck, the hurler on the mound is pitching to contact. Like the above, this would also lead to fewer strikeouts and more balls in play. This makes sense: In a situation, say, with the bases empty and two outs, it would be nice to retire the No. 8 hitter and have the pitcher leading off the next inning; it might be easier, though, to give the No. 8 hitter a chance to get himself out on two or three pitches. Either way, there’s very little chance of a run scoring.

Eight-spot all-stars

Fourteen players came to the plate at least 80 times last year with the pitcher on deck. Most of these are names you’ll recognize, and it’s interesting to see just how much of a boost some of them got with a weak hitter behind them:

                   In Front of Pitcher              In Front of Others
First   Last       PA    OBP     SLG     Stk%       OBP     SLG     Stk%
Alfredo Amezaga    103   0.359   0.371   61.3%      0.325   0.348   63.1%
Brad    Ausmus     188   0.332   0.354   60.8%      0.298   0.277   62.9%
Craig   Counsell   83    0.337   0.274   59.3%      0.308   0.287   60.7%
Tony    Graffanino 91    0.286   0.277   59.5%      0.293   0.421   60.5%
Chris   Iannetta   90    0.289   0.375   60.1%      0.346   0.319   60.3%
Cesar   Izturis    119   0.328   0.360   63.7%      0.290   0.274   64.9%
Jason   Kendall    89    0.382   0.397   61.9%      0.327   0.353   66.4%
David   Ross       143   0.282   0.466   59.4%      0.217   0.211   60.8%
Carlos  Ruiz       198   0.293   0.352   60.0%      0.356   0.409   57.3%
Brian   Schneider  80    0.300   0.286   57.0%      0.322   0.331   58.8%
Chris   Snyder     111   0.387   0.536   57.6%      0.323   0.360   58.9%
Yorvit  Torrealba  115   0.400   0.553   60.6%      0.292   0.290   62.0%
Omar    Vizquel    139   0.331   0.375   63.6%      0.289   0.285   60.8%
Jack    Wilson     182   0.357   0.506   62.7%      0.316   0.372   65.0%

Next steps

A better study would use multiple years of data—matched sets of 80-100 PA don’t cut it if we want to break down the findings any further. And presumably, breaking down is the direction we want to go.

The aggregate numbers, and the handful of players displayed above, agree on the generalities: Batters perform better ahead of the pitcher, and they don’t see as many strikes. But the effect is much stronger on some players than others, maybe just because we are working with small samples, but probably, to some extent, because different players handle the role better than others.