Yes, BABIP is simple when it comes to what it is. It is batting average on balls in play. The league average is around .300. In 2020, the average was .297. A batting average equivalent was .245. To a degree, the specific number is useless because we need to know the context meaning of the stat. Why is BABIP important to handicapping a baseball game?
The answer is in regression to the mean and defining BABIP in another way. Let’s define it so we can use it. A batted ball in most cases does not have an aim from a batter. There are occasions when batters are trying to directionally aim where they hit the ball, but most times the ball comes off the bat in any direction. A pitcher tries to affect where the batter hit the ball too. They pitch inside or outside, up or down intending to miss the large part (barrel) of the bat thus the batter makes weak contact. We see this in pitch sequences over and over. BABIP is an average on balls hit in play. It is random, where batters hit the ball, thus we can interestingly use this stat.
I used to think this was a stat for the quality of batters, however, it really is not. If a batter does not control where he hits the ball then he cannot control his BABIP either. This is where we can use it. All players will not have the same success as the rest when they hit the ball somewhere. Some will hit groundballs to fielders, or flyballs to fielders more often or less often. Over the course of a huge number of at-bats, the BABIP for all players would be very close. Over the course of shorter terms, we can see the variations in BABIP from player to player. These are the outliers we are looking for! Oh, yeah. We are going to use BABIP for pitcher performance too!
Pitchers have opposing hitter data like avg. or GB% etc., and they have a BABIP. A normal BABIP we identified was .297 or around .300 so we are using this number as the mean average. What if we see a pitcher with a BABIP of .385? Does this mean the pitcher is not good? What if his BABIP was .245? Does this mean the pitcher was excellent? The answers are exactly the opposite of what you might think. The better over the mean average like .385 from above means the pitcher is unlucky. The balls batted against him are not being hit to fielders. It does not mean the pitcher is a poor pitcher, and because he has been unlucky, we know that this temporary, and his BABIP will sink closer to the average. He should then see some games where balls are hit to fielders thus evening out his overall BABIP. Let’s say this another way. The pitcher has been unlucky to earn a .385 BABIP, thus he should be getting more outs above the mean average to compensate for the .385, therefore he will see better results and we can bet on that! It applies the same way when below the average. If a pitcher has a BABIP of .245, he has been very lucky having most of the balls put in play hit to fielders. His luck will change with time to get closer to the mean average of .300, which means he will see a worse performance. We can bet on that too! We have now found a way in which we can predict a spot where either good or bad performance can be expected. We need to use this with caution. This is not how we make wagers but we can use it as data to influence our thinking. It is a supporting data point to improve a narrative you have or disprove one, it is not a valid enough data point to stand alone to make a wager.
BABIP helps identify pitchers who will perform a certain positively or negatively. When combined with other data points, BABIP is a strong influencer. We need to look at the hard-hit rate as well. The higher the hard-hit rate is, there is the chance the BABIP will be above the mean too. If a ball travels more than 95mph off the bat, it is a hard hit. The fielders have less chance to get to (range) these balls as they were hard hit past them. These are scenarios where there is a logical reason for the poor BABIP performance and it is not an outlier. Anyway, the point is to understand the luck of where a batted ball is hit; whether at a fielder or not, and over time will even out creating opportunities in predictive outcomes.