http://tinyurl.com/5d3jf9
by Jeff Sackmann
April 17, 2008
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?
棒球界對於國聯球隊打線的第八棒—也就是說，在投手前的一棒—有些傳統的看法。投手
不太會投出較甜的球給第八棒，因為接下來會對上明顯較差的打者。對一位有耐心的第八
棒，這會增加他上壘的機會；對一位球來就打的第八棒，很可能結果是追打壞球。
2008年投手的棒次引起了更多的討論。這週的紅雀釀酒人之戰，雙方教練都不約而同的把
投手排在第八棒。於是，原本的「第八棒」現在成了打擊更好的「第七棒」。
現在我想知道的是，在投手的前一棒打擊和其他棒次有什麼不同?
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.
首先，我們先來蒐集一些在第八棒和其他棒次都打過不少次的打者的資料，比較他們在不
同棒次下的打擊情況。
這樣的打者倒是不難找到。教練常會換不同的打者打第八棒，即使是萬年第八棒的也常常
會上來代打。此外，我們假設投手的投球策略會因為下一棒而有所不同。
從資料看來，「第八棒」對打擊的影響還不小。
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").
資料是以2007年所有超過20個「第八棒」打席的球員統計資料為準。在這裡暫時不考慮投
手換代打的可能性。
打多數打者在「第八棒」的機會都不太多。以Jack Wilson而言，他最常打的是「第八棒」
沒錯，但他在其他棒次的打席總和是「第八棒」打席的兩倍。
為了消除這個變因，我把其他棒次的打席縮減到和「第八棒」打席一樣多。這樣打者在這
兩種情況下的統計資料在比例上不會相差太多。
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.
就和大家想的一樣：「第八棒」的打擊成績確實比較好。但，有趣的是，「第八棒」的保
送並不像大家想的那樣多。另外，P/PA顯示打者的打擊策略並沒有隨棒次而改變。好球率
倒是和傳統觀念相同；「第八棒」看到的好球比較少。
那麼，是什麼原因造成「第八棒」和其他棒次打擊成績的巨大差異呢？我想，從三振率我
們可以看出一些端倪；「第八棒」：14.1%，其他棒次：17.1%
可能的原因：投手會比較不在意「第八棒」的打擊，因為下一棒幾乎就是自動出局數，此
外，當「第八棒」上場時通常也都是局面較輕鬆的時候，這時投手比較會投給打者打。這
樣一來「第八棒」能打到的甜球就變多了。
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%
有十四位打者打超過80次的「第八棒」，當有投手當墊背時，他們的成績都明顯提升了。
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.
現在這個只是最初步的分析，接下來我還會研究年更多年更多「第八棒」打席的資料，這
也是我們未來的目標。
從目前這一大串的數字中，我們可以同意這個說法：「第八棒」的打擊確實比較好。當然
，現在這明顯的數字差異也有可能是小樣本的誤差造成，但是，我想這差異性還是存在，
因為每個「第八棒」都打的特別好。