So Pikachu and I have decided to write a short (maybe not) series of SABR "intro" FanPosts for the less stats-inclined out there. This will be the first of them, and hopefully it opens some eyes/helps people understand/is fun to read! A lot of this info is from the FanGraphs SABR Primer. The "BABIP gods" have come up quite a bit in comments lately, so I was hoping this post would clarify the religion a little bit.
BABIP stands for Batting Average on Balls in Play. It's calculated as (H - HR) / (AB - K - HR + SF)
As you can see, it's just a formula for figuring out how many balls in play (not homers, strikeouts, walks or IBBs, and including sac flies but not sac bunts) turn into hits. There are two sides to the BABIP stat: hitting and pitching, and we'll start with hitting.
Hitters, when they step to the plate, are obviously trying to "hit the ball where the fielders ain't" and get a hit. Batters do, in fact, have a reasonable amount of control over their BABIP, and good hitters tend to have substantially higher career BABIP than poor hitters. With the help of pitchF/X and hitF/X data, it's possible to easily determine how many types of each batted ball a player produces, of which there are four: line drives (the best), ground balls and fly balls (basically tied) and infield fly balls (terrible). Line drives have the best chance to fall in for a base hit of any of the four batted ball profiles (IIRC, around 70% turn into hits). Ground balls turn into base hits around 28% of the time, while fly balls turn into base hits around 22% of the time, but fly balls turn into extra base hits and homers (obviously) more frequently. Infield flies are actually almost as bad as striking out - IFFBs turn into outs around 95% of the time. More about hitter BABIP after the jump.
Now I think it's important to note that the term "luck" referring to BABIP isn't entirely accurate. It very much oversimplifies the observation, and I think is where some, maybe a lot, of non-stat-inclined fans get caught up. Players can and do, of course, head to the plate with a goal - maybe it's hit a fly ball to score a runner, get a base hit, or hit a ground ball to the right side to move the runner. These are absolutely 100% things players can try to do. However, in the long run, hitters just aren't very good at "aiming" their hits off Major League pitching between fielders, and as a result, random variation plays a great deal of importance in analyzing a player's future performance. Ground balls that find a hole between fielders, or a fly ball that misses a glove and turns into a triple doesn't really have a ton to do with a batter's skill, but the batter is trying their best to aim the ball there, so they do deserve some credit.
The best way for a hitter to improve his BABIP is to hit lots of line drives. "Lots" in the sense of liners is about 25% liners (LD/BIP), and it's extremely difficult for a player to hit more than that over the course of a season. Adam Lind right now is hitting just over 24% liners, and we can all see that he's being very successful. Ground balls also turn into hits quite a bit, but they come with the downside of having diminished power potential. Yunel Escobar is hitting over 60% ground balls this year - his AVG has gone up (he's also cut down on IFFBs) as a result. Fly balls are less likely to be hits than grounders at around 22%, but they can turn into homers (not affecting BABIP, but affecting production) and extra base hits. Jose Bautista, when he is on, hits a crazy number of fly balls and lots of them turn into homers. Lately it seems to me like he's hitting way too many grounders (whether or not he is trying to). Infield fly balls can absolutely devastate a player's offensive production. Aaron Hill has popped up around 13% of the time the past two years, and it really hurts his numbers - it's basically giving a free out on 1/8 of your fly balls. It is also very helpful when a player can consistently avoid IFFBs - Joey Votto hasn't hit one since 2009, meaning he gives himself as many good chances for hits as possible.
Another way for a player to improve his BABIP is to be fast - fast players are able to turn more grounders into hits, which can have sometimes have a pretty significant effect on that player's BABIP.
Just to show the BABIP difference between a good player and a bad one:
Votto has a career BABIP of .357. DeWayne Wise - a pretty fast player! - has a career BABIP of .251, so even speed can't have nearly as much of an impact on a player's BABIP as just being good at hitting. As you can see, there can be significant BABIP differences between hitters, but probably not as high as one might expect.
On to pitchers. BABIP for pitchers is a lot simpler - basically, they have almost no control over it. BABIP for pitchers is typically around .290 to .300, though it can change for individual pitchers for a number of reasons.
Reason 1: Defense. The obvious reason - if a pitcher has the Reds' (minus Jonny Gomes) or the Rays' defense behind him, the defense will make outs on a lot more hits than a team with Adam Dunn at every position. This will cause a pitcher's BABIP to go down, but it doesn't say much at all about the pitcher's individual talent. This is one of the main reasons DIPS, or Defense Independent Pitching Statistics, were invented, and I'll talk about two of the most popular, FIP and xFIP, in a coming FanPost.
Reason 2: Luck/Random variation. All the time we see hitters take ankle-high pitches out of the park, or pop up 90 MPH fastballs down Broadway. These things happen all the time, but a pitcher who is getting pop ups on every 90 MPH fastball down Broadway is highly unlikely to continue their success.
Reason 3: Talent level changes. Over the course of a season, every player makes adjustments that can negatively or positively (or not at all) affect their BABIPs. We all saw Ricky Romero toss more fastballs about a month into the season with great success, but that may have resulted in a higher or lower BABIP. At the same time, pitchers who throw too many 90 MPH fastballs down Broadway need to make an adjustment, and almost without fail, they do. This early adjustment period can often confuse hitters' approaches and cause weak hits (but of course it doesn't always).
Reason 4: Groundballer/flyballer. Groundballers (Romero), in general, have higher BABIP than flyballers (Brandon Morrow), but groundballers have the ability to get double plays and give up fewer extra base hits. In fact, groundballers give up fewer hits and fewer extra base hits on ground balls (getting more soft grounders/two hoppers) than do flyballers.
Another, perhaps surprising, thing to note regarding pitcher BABIP is that Major League pitchers give up line drives at pretty much a constant rate - while pitchers may get hit hard over small samples, pitchers that are good enough to stick in the Majors give up 19% line drives pretty much uniformly.
To conclude: any player with a tremendously high (Matt Joyce, Morrow) or tremendously low (Ubaldo Jimenez in the first half of 2010, Aaron Hill in 2010) BABIP should be expected to regress positively or negatively to their career average. Whether or not that actually happens is prone to many outside variables, but that should be the expectation.