Swinging at 3-0 pitches: A high-risk decision.

Swinging at a 3-0 pitch is a high-risk decision. How often do batters do it? What factors influence the likelihood?

Baseball is a game full of rules, written and unwritten. Early into this spring training, the New York Mets already find themselves embroiled in the midst of an unwritten rule violation related to sandwich eating. Other examples of these rules are more interesting than sandwich eating because they relate to the strategic aspect of the game. For example, swinging at a 3-0 pitch. 3-0 is a tremendous hitter's count. The hitter is only one ball away from being awarded first base, so the pitcher needs to work in the strike zone with pitches that he is confident he can command. Typically, a 3-0 pitch is a less than full velocity fastball thrown through the middle of the zone, which is certain to be classified as a good pitch to hit. However, for the most part we see that the batter lets this pitch sail on by for a called strike. The risk of making an out, even on such a hittable pitch, is just too high considering a walk has almost been achieved. However, there are cases when a hitters swings at a 3-0 pitch. We can ask: how often does this happen? What are the typical results? In which game situations (e.g., inning, score) are we more likely to see a swing?

PITCHf/x data can help respond to these questions. Using the Baseball-Savant PITCHf/x search tool, I gathered every 3-0 pitch thrown in a major league game for the 2009-2014 seasons. Any pitch that was marked as an intentional ball or pitchout was removed, as those pitches do not fit with the focus of this work. Before responding to the questions stated above, it is worth determining the extent to which my statements about typical 3-0 pitches are true. At Baseball-Savant, the strike zone is divided into 13 sections as shown in the image below:

On 3-0 pitches, zone 5 is the 5^th most common zone for a pitch to be thrown through; zones 11-14 are more common. Overall, the pitcher throws a fastball (Fourseam, Twoseam, Cutter, Sinker) 94.7% of the time (95.91% when in zone 5). Hitters can be quite sure that on 3-0 they are going to see a (~90 mph) fastball and if the pitch is in the strike zone it is most likely to be in the middle of the zone. But how often do batters swing? A mere 7.6% of the time. Most surprisingly is that this swing rate only jumps up to 11.7% when the pitch is right down the middle. For this same time frame (2009-2014), league-wide swing rate (across all counts) is around 46%. So clearly the lure of the impending walk is affecting hitters' tendencies. It may be that teams have instituted rules about swinging at 3-0 pitches that limit what players would do on their own. Regardless, what happens when a player does swing?

Since 2009, there have been just over 3500 swings at 3-0 pitches. When batters have put the ball in play they have fared quite well with a .353 batting average on balls in play (BABIP). This rate is higher than the typically expected rate (~.300), so it might be taken as evidence that players should be swinging more at these 3-0 pitches. However, this is almost certainly a select sample of players that are swinging in these counts (read: power hitters), and as already outlined these are likely good pitches to hit. So it seems fair to expect that the BABIP calculated from these 3-0 swing data is somewhat higher than typical. Most successful contact goes for a single, but doubles, triples and home runs happen too (at rates of 6.9%, 0.9%, and 9.2% per ball in play, respectively).

So while hitters seem to do quite well when swinging at 3-0 pitches, their performance is not so extraordinary that we should advocate for a lot more swinging. Remember that a hitter in a 3-0 count is quite likely to get on-base via a walk. Pitchers have thrown a ball 33% of the time in a 3-0 count. And even if they throw a strike, a 3-1 count is still an advantage for the hitter. So taking that down the middle 90-mph fastball still leaves the hitter in a solid situation, and therefore swinging 3-0 is somewhat ill advised.

This gets us to looking at the game situations that lend themselves to more 3-0 swings. For example, early in the game might be more likely than late in the game, which is also likely dependent on the score. A swing on a 3-0 count in the 7^th inning is probably more likely if the game is a blowout than if it is tied. It may also depend on the number and location of runners on base. And all of that likely depends on the batter. As Baseball Prospectus' Sam Miller pointed out in the article linked above, Reds manager Bryan Price said of Joey Votto:

"In a tie game, he has the green light."

Price's comment came after a Votto 3-0 swing resulted in a game winning home run, so his perspective is certainly biased by a recent event. But the comment does lend some insight into how tolerant managers will be of their players swinging 3-0; a complex interaction of factors influences things.

The Baseball-Savant data I used above does not include a few factors of interest (e.g., score, runners on base) so I used Retrosheet for this information. My database is not yet updated with the 2014 data, but I expect the 2009-2013 rates are representative of how 2014 played out. But it is something to note. Some interesting trends emerge when looking at inning, score, and runners on base.

Inning appears to be a factor contributing to whether a batter will swing in a 3-0 count. The proportion of 3-0 swings has been higher in the first six innings (and roughly equal across innings 1-6), than in innings 7 through 9+. Stated differently, ~76% of 3-0 swings have happened in innings 1-6. If the proportion of swings were distributed evenly across innings we would expect ~66%. So we have some evidence that batters are changing their approach on a 3-0 pitch as a function of inning. They are swinging less often at these pitches in the later innings.

The score of the game also has an effect on batters swinging at 3-0 pitches. Because there are many possible run differences occurring within games at different frequencies I just binned scores at the time of the pitch according to whether the batting team was leading, tied, or trailing. If the score had no influence on whether a batter swings 3-0, we should expect roughly 1/3 of the swings to have happened in each bin, as the breakdown of leading, tied and trailing are roughly equally represented. However, that is not the case. Most 3-0 swings happen when the batter's team is leading. And they appear to be more reticent to swing when their team is trailing. Interestingly, inning did not change the trend of more swings at 3-0 pitches happening when leading. Batters may be more risk averse when they are losing because they do not want to make an out that could reduce the likelihood of a comeback; whereas when their team is already winning this relation is experienced differently and making an out seems less costly.

A final factor that I considered as mediating the rate at which batters swing at 3-0 pitches was the number and location of runners on base. I suspected that batters (and managers) would be more hesitant to swing at a 3-0 pitch with a runner on first, as the fear of hitting into a double play would be overwhelming. However, this did not turn out to be the case. If anything it appears as though batters swing more when there is a runner on first. However, as far as I can see there is not much of a distinct trend in swinging more often at 3-0 pitches as a function of base runners:

There are a few situations where swing% is higher than typical rates (i.e., 7.6%), but it is not clear what is driving the effect.

Conclusions

After many words, we now have some responses to the original questions posed in this article. First, unsurprisingly, 3-0 pitches are likely to be fastballs. When they are in the strike zone, they are most likely to be right down the middle. However, batters are do not swing very often. When they do and put the ball in play they fare quite well, but this result could just be a function of the type of hitter that is typically allowed by his manager to swing on 3-0. The inning and score each seem to influence the likelihood that a batter will swing 3-0, with more swings happening early in the game, and when the batter's team is leading. A batter's decision to avoid swinging at a 3-0 pitch despite the strong likelihood they know what is coming is part of baseball culture and in some ways one of the many unwritten rules that are part of the game.

. . .

All statistics courtesy of Baseball-Savant and Retrosheet.org.

Chris Teeter is a featured writer and editor at Beyond the Box Score. You can follow him on Twitter at @c_mcgeets.