I really enjoy watching the pitcher and batter battle one another during an individual plate appearance (PA) and then seeing the adjustments the two make as the game goes on. For me, its things like these individual battles within a battle and the hidden strategies which define them that give baseball its charm and substance.
So when Chris Welsh commented during Belisle’s last start that a key to Matt’s success rests on his ability to throw a first pitch strike, my ears naturally perked up. This statement certainly seems to be consistent with traditional baseball wisdom as there is an intuitive advantage to being ahead in the count. As it happens, I’ve been tracking several batting splits of all Reds positional players that were on both the ’06 and ’07 rosters in order to do a study on the effect of Brook Jacoby’s impact. I used the ’06 data set (3,768 PAs) to test the wisdom behind the importance of throwing a first pitch strike. When comparing count splits for Reds hitters in ’06, the OPS of batters was .251 lower in PAs that began with a 0-1 count (OPS=.684) versus those that began 1-0 (OPS= .935). Traditional baseball wisdom certainly seems to hold in this case.
However while on the surface it seems like a cut and dry case, this “first pitch strike” dogma has always struck me as not very useful for understanding a PA. At best it’s a blunt summation which only paints the broadest, most superficial picture of a given PA. Surely, throwing a first pitch strike doesn’t tell the whole story of what can often be a complex PA that many times lasts longer than 5 pitches. A lot can happen after that first pitch which can completely nullify its effect. A 0-1 count can quickly become a 3-1 count. Certainly there can be a big difference in first pitch strikes thrown from a guy with average stuff who is pitching to contact versus a guy with electric stuff. It’s fairly obvious that a fouled-off meatball probably doesn’t carry the same significance to the immediate fortunes of the batter as a well-located, nasty first pitch strike might. In the case of the pitcher who is “on his game”, it’s likely that he’s pretty effective on pitch two and three and so forth as well. If that’s the case, was it really the first pitch strike that did the batter in? Intuitively, there has to be more to winning an individual PA than simply throwing the first pitch in the strike zone. This led me to wonder what is the actual value of a strike anyway and what leverage does the pitcher really gain by throwing one. Is 0-1 really a pitcher’s count? Is 2-2 really an even count?
To try and answer some of these questions, I once again used the ’06 data set for Reds position players that were on both the ’06 and ’07 active rosters. I’ve broken down all possible pitch counts and examined the OPS for outcomes when the PA ended on that count. The data is sorted by balls and also by strikes to illustrate the effect of each on OPS.
A few points seem to be made clear when analyzing a PA on a situational basis:
1. First, the effect of the first strike doesn’t seem to be nearly as dramatic as the earlier more blunt splits suggested. An 0-1 count only reduced the batter’s OPS by a minuscule .009 versus hitting the first pitch (0-0, OPS=.912; 0-1, OPS=.903). However, throwing a ball is somewhat of a ding.
2. These splits also suggest that a batter gains no advantage by swinging at the first pitch. In fact, it probably is a bad idea as a philosophy given the advantage gained by hitting in an 1-0 count (0-0, OPS=.912; 1-0, OPS=1.054), the even greater advantages gained if a batter can successfully work the count deeper in his favor, and the fact that hitting in an 0-1 count places the batter at no real disadvantage (0-0, OPS=.912; 0-1, OPS=.903). In essence, here is a case where the batter can quite possibly gain by simply doing nothing.
3. Importantly, these data strikingly demonstrate that it is really strike two and ball three that are actually the true leverage pitches (dramatically swing the advantage in either the pitcher’s or batter’s favor). These data suggest that the first pitch in reality is not a high leverage pitch at all. Perhaps the correlation between first strike and OPS in the blunt split in the second paragraph above had more to do with the pitcher having command versus not having command rather than the first pitch strike directly impacting the outcome of the PA.
To further illustrate the significances of strike two or ball three, I averaged the differences in OPS resulting from each of the outcomes (i.e. the effect of strike one was determined by adding the differences between 0-0 and 0-1, 1-0 and 1-1 and so forth). These means (+/- standard deviations) are shown below:
This further demonstrates that it’s strike two and ball three that really are the true leverage pitches in any given PA. Also note that while the effect of strike one seems large, it’s effect was really inflated by the ding a batter takes by going from a 3-0 count to a 3-1 count. This really has more to do with the tremendous advantage that ball three gives the batter rather than the value of strike one per se.
Finally, when examining the distribution of OPS over the individual pitch counts, OPS neatly broke into three distinct groupings as demonstrated below:
In ’06 the national league average OPS for positional players was .793. Using league average as a baseline suggests several things that aren’t necessarily intuitive:
1) 3-2 is the only count that is truly an even one since it is the only one where the associated OPS approximated league average (suggesting neither the pitcher nor the batter had a distinct advantage).
2) Any count with 2 strikes and less than 3 balls is a pitcher’s count (i.e. the pitcher has leverage). This means that 2-2 really shouldn’t be thought of as an even count as there is a clear advantage to the pitcher.
3) Any other count should be considered a hitter’s count since based upon outcomes on those counts, the hitter’s performance tends to be better than league average. Getting back to the notion that throwing a first pitch strike is a key to pitching success, these data suggest, an 0-1 count is actually a hitter’s count.
In summary: while throwing a first pitch strike is better than throwing a first pitch ball, it’s strike two that is really the key to a pitcher’s success.
Chris Welsh would've been more accurate to say that the key to Belisle's success is command and the ability to get a second strike on hitters. Ultimately this suggests an ability to miss bats pays off big.