In my last post, I looked at the variation in team infield shifts in current baseball and saw how these shifts affected the batting average on balls put in play. For this post, I thought it would be interesting to look more generally at batting averages on balls put in play (with home runs removed). Here are some things I’ll explore:
- Trends in BABIP. How have batting averages on balls in play changed in recent baseball season? Given all of the defensive shifting, how have BABIP values changed for different batted ball types such as ground ball or line drives?
- BABIP stars? Which players have excelled in BABIP in recent seasons?
- Relationship of BABIP and spray angle? We know that teams tend to shift for players who tend to pull the ball. Is there a relationship between a player’s BABIP and his tendency to pull the ball? Can we identify the players who don’t pull the ball and still maintain high BABIP values?
- Spray angle and launch velocity. Since players tend to pull their balls, do they tend to hit balls harder on the pull side than the opposite side?
- BABIP, spray angle and batted ball type. We’ll see that there is a relationship between a player’s BABIP and the fraction of balls that are pulled. Does this same type of relationship hold for all types of batted balls?
Batting Average on Balls in Play
Let’s start with looking at the batting average on balls put in play (BABIP) where home runs are removed, that is
BABIP = (H – HR) / (AB – SO – HR)
Using Lahman data, I’ve graphed the overall BABIP for the seasons 1969 through 2019. We see a gradual increase in BABIP from 1969 to 1992, a big increase in BABIP in the early 90’s, and BABIP has stayed relatively constant in the last 25 seasons.
In passing, here are the top 10 BABIP seasons in this period (minimum 400 balls in play). A couple of famous players, Ichiro Suzuki and Derek Jeter, occupy four of these top 10 seasons.
Let’s focus on the Statcast seasons from 2015 through 2019. Overall, we saw that BABIP hasn’t changed over this period. But one gets a different impression if the BABIP values are plotted against season for each of the four batted ball types. As one might expect, ground ball BABIP values have dropped from 0.282 to 0.270 in this five year period. But this decrease is offset by the increase in BABIP values in this period for line drives and fly balls.
I define a new variable Spray Angle that is equal to negative the launch direction for a left-handed hitter. A negative (positive) value for Spray Angle corresponds to a ball hit to the pull (opposite) side of the field for all hitters. We know that hitters tend to hit to the pull side. Also, batters tend to hit balls harder to the pull side. For example, suppose we define a hard hit ball as one hit at 100 mph or higher. Below is a graph of the hard hit fraction as a function of the spray angle — clearly it is more likely to see a hard-hit ball for spray angles between -30 and -10 degrees.
Pull Fraction and BABIP
Of course, batters vary in their tendencies to hit to the pull side and baseball teams are very aware of these tendencies given all of the infield shifting in current baseball. What is the relationship between a player’s pull percentage and his BABIP? (Remember we are not considering home runs.) I collected all of the hitters who had at least 1000 cumulative BIP over the period 2015 through 2019. Below I construct a plot of the pull fraction against the BABIP. There is a relatively strong relationship here — batters who pull a lot tend to have a small BABIP. There are six points (that I have boxed in the graph) that seem to stand out — these are players who have unusually small pull fractions and high BABIP values.
Six Special Hitters
Here are the six special players — all appear to spray their BIP evenly to pull and opposite sides. DJ LeMahieu really stands out as he hit about 54% of his balls in play to the opposite side. (I’ll mention Joe Mauer shortly.)
Increase in Launch Speed in Pull Direction
We saw earlier that the hardest hit balls tend to be in the pull direction. That raises the question — what do players lose in terms of launch speed by hitting to the opposite side compared to the pull side? For my group of players with at least 1000 BIP in the period 2015-2019, I plot the median launch speed (horizontal) against the increase in median launch speed in the pull direction (vertical). As one might expect, most of the players lose between 1 to 5 mpg (on average) by hitting to the opposite side. But there are a number of points below the horizontal line at 0 — these correspond to the hitters who actually hit harder to the opposite side, on average. The colored (red) points correspond to the six with the smallest pull percentages listed above — four of these players appear to hit harder on average to the opposite side.
We all know that Joe Mauer (who retired in 2018) was a great Minnesota Twins player who likely will be inducted in the Hall of Fame. What is fascinating about Mauer is how he distributed his hits to the pull and opposite sides. Below I have listed the number of hits H_0 and balls-in-play N_0 to the opposite side and hits H_1 and balls-in-play N_1 to the pull side for the last four seasons of his career. As we saw earlier, Mauer sprayed his balls equally to both sides of the infield — his success rates were a little higher to the opposite side.
Pull Fraction and BABIP for Different Types of Batted Balls
We earlier saw a negative relationship between pull fraction and BABIP for our group of players with at least 1000 BIP for our Statcast seasons. Suppose we construct this same type of graph for each batted ball type? Do we see the same pattern of relationship?
Here’s the graph. First, note that the location of the pull fractions depends on the type of batted ball. Ground balls tend to be pulled (average pull fraction around 0.72), followed by line drives (average pull fraction around 0.6), followed by fly balls (average pull fraction around 0.5), and pop ups (average pull fraction about 0.35). In terms of relationships, pull fraction is strongly negatively associated with BABIP for ground balls, and slightly negatively associated with BABIP for fly balls. I see little relationship between pull fraction and BABIP for line drives and pop ups.
As usual, this brief exploration raises more questions about BABIP. In particular, here are some questions that I encourage the interested reader to explore.
- Why did overall BABIP suddenly increase in the early 1990’s?
- I understand that BABIP on grounders would decrease recently due to all of the infield shifting. But what is the explanation for the increase in BABIP for line drives and fly balls? (Remember I am excluding home runs in this study.)
- It would be interesting to showcase the players who were unusually strong in BABIP. For example, Tony Gwynn was remarkable for his ability to spray his hits (like Joe Mauer) to all fields. Is this a dying talent in modern MLB players?
- Here I have explored how BABIP depends on pull fraction and batted ball type. I have not considered pitcher effects — is there evidence that some pitchers are specially good in limiting BABIP? Also the role of batter speed in BABIP would be an interesting topic to explore — which players benefit with infield base hits?