July 12, 2002

Aim For The Head

Long Plate Appearances Mailbag

by Keith Woolner

Judging from the feedback in my inbox, people are still having fun looking at the number of pitches in a plate appearance, so let's answer a few more questions from the mailbag.

E. O. writes:

I'm having a bit of an argument with a friend about the likelihood of striking out...I've said you're more likely to strike out the deeper you go into a PA, he's said that doesn't make sense. Obviously, you have the numbers there...could you give us the strikeout and walk rates for each pitch of a PA?

NP    SO_R    BB_R   SO/BB
3   0.1465       0     ---
4   0.2340  0.1294   1.808
5   0.2606  0.1613   1.616
6   0.2654  0.2024   1.311
7   0.2390  0.2306   1.036
8   0.2247  0.2413   0.931
9   0.2132  0.2466   0.865
10  0.2073  0.2463   0.842
11  0.2046  0.2491   0.821
12  0.2019  0.2718   0.743
13  0.2045  0.2386   0.857
14  0.1639  0.1639   1.000
15  0.2143  0.3214   0.667

Jeff Angus writes:

Based on your graph of Long PA events, it suggests that the more times a batter faces a pitcher in the same game, the more likely he is to have a long PA. The uptick in the last two innings would reflect (if this hunch is correct) be a higher frequency of facing-a-relief-pitch, that is, a "new" pitcher in this game.

One way to look at this is to look at the average pitch length for starters and relievers both early in the game (say, innings 1-6) and late in the game (innings 7 and beyond). I've done so in the table below.

Average NP/PA   EARLY (Inning 1-6)   LATE (Inning 7+)
Starters              3.62                3.48
Relievers             3.65                3.69

Contrary to what Jeff suggests, starting pitchers appear to be more efficient with their pitches later in the game, whereas starters in the first 6 innings, and relievers at any time, all average about 3.65 pitches per plate appearance. One important caveat is that we haven't controlled for the quality of pitcher. Good pitchers will throw into the 7th inning more often that bad pitchers, and thus good pitchers will tend to be overrepresented in the Starter/Late sample. A better comparison would be to look at only games where the starter goes 7 or more innings, and see how the average pitch length breaks down, or, alternately, to balance the contribution of pitchers across the two groups.

An anonymous reader writes:

[...] concerning long PA's, what about the BA, OBS, Slug, OPS for the NEXT batter? It's oft spouted wisdom that longer taking a lot of pitches helps the rest of the line up. Is it true?

In a word, no. Here's the OPS for batters following a plate appearance of different lengths (min 1000 PA).

PITCHES   AVG  OBP  SLG  OPS
1        .264 .333 .398 .731
2        .264 .333 .401 .733
3        .263 .331 .403 .734
4        .259 .326 .398 .724
5        .262 .328 .405 .732
6        .264 .327 .410 .738
7        .262 .324 .410 .733
8        .268 .328 .419 .747
9        .258 .319 .403 .722
10       .265 .326 .413 .739
11       .265 .322 .415 .737

Richard Chen writes:

How about the most pitches in a plate appearance with 2 outs and a runner on first? I remember ESPN's showing this situation with about 6 fouls on 3 and 2; when the batter finally ended the stalemate, the guy on first mock-collapsed.

Sounds like fun. There are 6 instances in the database where a batter saw 15 or more pitches with 2 outs and a runner on first. I'll save the longest (and probably the most interesting, considering the score and situation) for last… so let's start the countdown:

# 6. April 25, 1989, Giants at Cardinals, no score in the bottom of the first. David Robinson pitching to Pedro Guerrero, with Terry Pendleton on first.

(throw to first)
1. ball 1
2. ball 2
3. foul (strike 1)
4. called strike 2
5. foul
6. foul
7. foul
8. foul
9. foul
10. foul
11. foul
12. foul
(throw to first)
13. ball 3
14. foul
15. ball 4 (walks)

Note: One of only two such episodes where the batter reached base, and the only one where the pitcher threw over to first. This is also the only entry where there is one a runner on first. All the others have 2 runners on, or bases loaded.

# 5. June 2, 2000, Rockies at Brewers, no score in the top of the 1st. Jason Bere pitching to Brent Mayne with the bases loaded (Todd Helton on first, Jeff Cirillo on second, Tom Goodwin on third).

1. foul (strike 1)
2. foul (strike 2)
3. ball 1
4. foul
5. foul
6. ball 2
7. foul
8. foul
9. foul
10. foul
11. foul
12. ball 3
13. foul
14. foul
15. swinging strike 3 (strikes out)

Note: One of two bases-loaded situations in the list. The only swinging strikes on the list were on the first pitch, or the very last pitch.

# 4. July 30, 1992, Orioles at Yankees, Yanks lead 3-0 in the top of the 5th. Curt Young pitching to Jeff Tackett, Joe Orsulak on first, Leo Gomez on second.

1. called strike 1
2. ball 1
3. foul (strike 2)
4. ball 2
5. foul
6. ball 3
7. foul
8. foul
9. foul
10. foul
11. foul
12. foul
13. foul
14. foul
15. in play (grounds out, pitcher to first)

Note: Tackett is probably the least accomplished major league batter on the list, which is saying something given the next entry's batter. Entries #6, #4, and #3 each had a sequence of 8 straight foul balls--tops on this list.

# 3. April 9, 1997, Indians at Mariners, Mariners lead 3-0 in the bottom of the 1st. Bartolo Colon pitching to John Marzano, with the bases loaded (Russ Davis on first, Paul Sorrento on second, Jay Buhner on third).

1. swinging strike 1
2. foul (strike 2)
3. ball 1
4. ball 2
5. foul
6. ball 3
7. foul (runners going now and on every pitch hereafter)
8. foul
9. foul
10. foul
11. foul
12. foul
13. foul
14. foul
15. ball 4, (walks, forcing in a run)

Note: The only run scoring play on the list, forcing in a run. After starting off with an 0-2 count, Colon loses the battle 13 pitches later.

# 2. April 28, 2000. Rangers at Orioles, Rangers lead 2-0 in the bottom of the 3rd. Esteban Loaiza pitching to Albert Belle with Delino DeShields on first, Brady Anderson on second.

1. called strike 1
2. foul (strike 2)
3. foul
4. foul
5. ball 1
6. ball 2
7. foul
8. foul
9. foul
10. foul
11. foul
12. foul
13. foul
14. ball 3
15. foul
16. in play (pops up to first base)

Note: Ball 1 (pitch 5) and ball 3 (pitch 14) were called later than any other PA on the list that reached a full count.

# 1. June 3, 1989. Blue Jays at Red Sox, score tied at 11 in the bottom of the 9th. David Wells pitching to Marty Barrett, with Wade Boggs on first and Jody Reed on second.

1. called strike 1
2. foul (strike 2)
3. foul
4. ball 1
5. foul
6. foul
7. foul
8. foul
9. ball 2
10. foul
11. foul
12. foul
13. foul
14. foul
15. foul
16. foul
17. in play (grounds out, 5-3)

Note: This is the wacky game where the Red Sox blew a 10-0 lead after six innings, and were behind 11-10 going into the bottom of the ninth. The Bosox tied it in the bottom of the 9th, and a base hit here would have won the game, but it wasn't to be. The Blue Jays won the game 13-11 in 12 innings. Barrett is the only batter on the list whose plate appearance didn't run to a full count, and he swung at 14 pitches in this at bat, also tops on the list.

Keith Woolner is an author of Baseball Prospectus. You can contact him by clicking here.