Batting average on balls in play, better known as BABIP, is a relatively new baseball statistic that helps distinguish how much of a pitcher or hitter's performance is due to skill and how much is due to luck.
Simply defined, BABIP is the batting average posted by a batter (or yielded by a pitcher), on all balls that stay in the park and land on the field of play. This means that unlike normal batting average, BABIP does not include strikeouts, home runs, or sacrifice flies. The formula for BABIP is as follows:
BABIP = (H-HR)/(AB-K-HR+SF)
The theory behind BABIP is that although batters and pitchers have a great deal of control over how much they strikeout and how many home runs they hit or allow, neither batters nor pitchers have much control if any over what happens to the ball once it leaves the bat. Will the ball be hit hard, but right at a fielder for an easy out? Or will it be hit poorly, but trickle between two infielders for a hit? According to the theory behind BABIP, this is mostly decided by pure, dumb luck.
At first many people were skeptical of BABIP, as the theory behind it flew in the face of decades of baseball wisdom, which held that a good batter hit the ball harder and thus had a higher batting average, and a good pitcher induced softer hits and thus yielded lower averages. However, BABIP has been backed up overwhelmingly by statistical evidence.
It turns out that for most players in most years, their BABIP posted or yielded will be around .300. Once in a while a player will post a BABIP dramatically higher or lower than .300, with a corresponding spike or dip in normal batting average, but BABIP evinces tremendous mean regression, which is to say that any player who posts an extremely high or low BABIP will be statistically extremely likely to return to around .300 the following year, and very few players can consistently remain well above or below .300 year after year.
BABIP can be extremely useful in helping to determine whether a player with a breakout season or a off year really improved or declined, or simply got unusually lucky or unusually unlucky. When a player posts unusually high or low statistics, it is now easy to check under the hood and say why. If the player has an unusually high or low BABIP, it's safe to say they just got lucky or unlucky, and are very likely to return to "normal" performance in the near future. But if their BABIP is in line with career norms, than it is probable that they really did improve their underlying skillset or suffer some sort of real decline in ability.