The Importance Of Strike One (and Two, and Three…), Part 2by Craig Burley
October 15, 2004
In last week's piece, I discussed the overwhelming importance (and substantial benefits) of throwing a strike on a 0-0 count. As I mentioned then, it was researcher/reliever Carlos Gomez who first pointed this fact out to me, not by highlighting the benefits of strike one as baseball people so often do, but instead by showing me how small the risks are. Every pitcher should make it his mantra - less than 8% of first-pitch strikes turn into base hits.
Thanks to some data posted recently at Diamond Mind Baseball by Tom Tippett (the White Wizard of DMB), I was able to look not only at the results of throwing balls and strikes on 0-0, but also look at the data from every other count and do some analysis there as well. First, it's probably most instructive to just look at the table of data. “Strikes” here do not count foul balls on 2-strike counts, but only third strikes (strikeouts) plus balls put into play. This data is for all major league non-pitchers (i.e. no pitchers hitting are included in the data) for the entire 2003 season.
The data incorporates what happens on or after the pitch in question. Thus, “1-1 Strike” includes all 1-1 counts where the ball is put in play, as well as all 1-1 counts that then proceed to 1-2 (but not 0-2 counts that proceed to 1-2). The average non-pitcher in 2003 hit .268/.337/.430, so that is the baseline that hitters are compared to. Intentional walks are removed from this data.
AVG OBP SLG LWR/PA Benefit 0-0 Strike .261 .296 .411 -.029 -.069 0-0 Ball .280 .385 .459 .040 0-1 Strike .231 .256 .349 -.085 -.058 0-1 Ball .248 .319 .396 -.027 0-2 Strike .167 .179 .243 -.200 -.068 0-2 Ball .191 .243 .294 -.132 1-0 Strike .276 .325 .445 .005 -.100 1-0 Ball .290 .516 .493 .105 1-1 Strike .241 .352 .377 -.060 -.092 1-1 Ball .262 .398 .433 .032 1-2 Strike .177 .186 .260 -.183 -.108 1-2 Ball .207 .307 .333 -.075 2-0 Strike .288 .386 .493 .054 -.160 2-0 Ball .296 .725 .496 .214 2-1 Strike .258 .318 .421 -.014 -.156 2-1 Ball .281 .586 .475 .142 2-2 Strike .195 .200 .308 -.153 -.192 2-2 Ball .230 .470 .380 .040 3-0 Strike .296 .578 .496 .144 -.200 3-0 Ball .000 1.000 .000 .344 3-1 Strike .281 .418 .475 .060 -.284 3-1 Ball .000 1.000 .000 .344 3-2 Strike .230 .230 .380 -.098 -.443 3-2 Ball .000 1.000 .000 .344In this chart, the batting lines are self-explanatory. "LWR/PA" is the linear weights value (expressed in runs) per plate appearance, expressed as always by an amount above or below average. So where a pitcher throws a ball on 1-2, hitters will, on average, produce .075 runs less than average in that plate appearance. Likewise, where a pitcher throws a strike on 2-0, hitters will produce .054 runs above average in those plate appearances. Where a pitcher throws a ball on a 3-ball count, the result is a walk, which results in .344 runs above average for the hitters. These are custom linear weights values for the 2003 season.
The "Benefit" is the run value to the pitcher of throwing a strike instead of a ball on that count. So on a 1-1 count, throwing a strike instead of a ball has a net benefit to the pitcher of keeping .092 runs off the scoreboard (on average, of course). The comparable benefit of a 0-0 strike is only .069 runs. This seems to confirm a bit of baseball lore; we'll get to that in a minute.
One important thing to keep in mind is that the "value" of a strike on the 2-strike counts is not as high as is shown here. The reason for this is that foul strikes are not included (since they don't add a strike to the count) and so a lot of "strikes" on 2-strike counts don't end the at-bat. Instead, they end up with the same result as a ball on that count (because eventually the pitcher throws a ball instead of getting a strike that counts).
So as a result, the numbers for all the 2-strike counts are off, probably by quite a bit. If I had a good idea, even a rough approximation based on a few games' worth of data, how often 2-strike "strikes" were fouls, then I would be able to more accurately estimate the numbers. For now, treat these numbers as unreliably high.
Nevertheless, the numbers do tend to confirm some of the common sense about pitching to the count. The 0-1 and 0-2 counts are the ones where you can most get away with wasting a ball; the 3-2 count is the one where you always must, must, must be in the strike zone. For the nonexistent pitcher who can throw strikes like a metronome, 3-2 isn't a bad place to be; hitters bat just .230 there, and they have a touch less power than average. Much more power, though, than on the other 2-strike counts. But for the pitcher who doesn't hit the strike zone on 3-2, the hitter can take a walk, which is always a pretty positive result. The overall OBP on 3-2 counts is a whopping .470, because it's tough to throw 3-2 strikes over and over. 3-2 isn't where the pitcher wants to be.
In itself, though, that doesn't make 3-2 "the most important" count to throw strikes on. (Remember, as well, that these numbers are high-side estimates. The reason is that while 3-2 counts are pretty valuable counts, there are a lot less of them than 0-0 or 1-0 counts, for example. So while hitters faced 22,183 3-2 counts in 2003, they faced over 80,000 0-1 counts, and more than 75,000 1-0 counts. And since every one of the 175,638 plate appearances in the data included at least one pitch, it turns out that the first pitch has more impact than any other.
Now pitching coaches, and other baseball people, will often tell you that 1-1 is the most important count for a pitcher, because of the massive difference between 2-1 and 1-2. Mike Marshall, in a discussion with Steve ("CrashCourse") of NetShrine, recently said that
The highest batting average hitters achieve is in their one pitch at bats. Therefore, while strike one is very good, it has to be a pitch that does not allow hits. The key to pitching is one ball two strikes rather than two balls one strike. The first pitch is important, but the next two pitches are just as important.
Likewise, Greg Maddux considers the 1-1 pitch the most important pitch in baseball, because 1-2 and 2-1 are "different worlds". On the other hand, Reds pitching coach Don Gullett and Giants manager Felipe Alou - amongst others - say that the first-pitch strike is the most important.
The chart above will confirm that the impact of a strike versus a ball on a 1-1 count, is greater than that on a 0-0 count. The 1-1 strike is worth .092 linear weights runs to the pitcher, while the 0-0 strike is worth only .069 runs. But comparatively, the impact of 0-0 strikes is almost twice that of 1-1 strikes. This is because while there were 175,638 plate appearances that included 0-0 counts, there were just 68,748 that went to a 1-1 count. If you emphasize the importance of 1-1 strikes to a pitcher, but de-emphasize the importance of 0-0 strikes, he will not do as well. Here is the chart of all the counts and their relative "importance"... i.e. the amount of runs that would be saved by throwing all strikes in those counts, as opposed to throwing all balls.
Count Benefit Potential Runs Saved 0-0 -.069 12,106 runs 3-2 -.443 9,817 1-0 -.100 7,281 2-2 -.192 7,267 1-1 -.092 6,359 2-1 -.156 5,852 1-2 -.108 4,895 0-1 -.058 4,707 3-1 -.284 4,595 2-0 -.160 4,073 0-2 -.068 2,070 3-0 -.200 1,623Again, remember that the numbers for the 2-strike counts are only an estimate and are too high.
There is another way to look at the de-emphasis of the 0-0 strike and the emphasis on 1-1 and later counts that shows the wisdom of throwing early strikes. Starting pitchers in particular need quick at-bats. Starters who work hitters into deep counts do have more success in reducing batting averages as Mike Marshall points out. However, pitchers cannot consistently throw five, six, or seven pitches in an at-bat and still be successful, as Marshall himself points out in his online book, Coaching Pitchers:
1. Pitchers should throw all pitches for 66.7% strikes.
2. Pitchers should end 75% of at bats within three pitches.
3. Pitchers should end 100% of at bats within five pitches.
In order to end at-bats quickly, pitchers must maximize the number of strikes they throw early in the count, on the first three pitches. The base hits on early strikes - and they are going to happen - are a pitcher's occupational hazard. Pitchers shouldn't let them get in the way of pressuring hitters.
0-0 is the predominant count in baseball. The first strike is the soul of every pitcher's success, and pitchers who don't throw first-pitch strikes get killed. Consistently, even the most hittable pitchers in the majors give up base hits on fewer than 10% of their first-pitch strikes. Certainly, the old saw that the pitcher's most important pitch is a strike, rings true. But another popular piece of myth, that the 1-1 count is the big one, should probably be put out to pasture.
References and Resources
I'd like to thank Carlos Gomez, Tangotiger, Tom Tippett, CrashCourse from Netshrine, and the many readers who wrote me after Part One to share their ideas and insights.
Craig Burley can be contacted via e-mail.
<< Return to Article