Batting Average on Balls in Play — the rate at which a hitter's contacted balls (excluding home runs) become hits.
Batting average on balls in play is the most-cited regression flag in baseball analysis, and the most misused. The number is simple: hits minus home runs, divided by at-bats minus strikeouts minus home runs plus sacrifice flies. It tells you what fraction of a hitter's contact found a hole.
Voros McCracken's original insight was that pitcher BABIP is mostly noise — over long stretches, every pitcher's BABIP regresses toward the league average around .300. That research kicked off the entire defense-independent pitching movement and underpins FIP, xFIP, and SIERA. For pitchers, a season-long BABIP much higher or lower than .300 is, in most cases, a luck signal.
For hitters the story is more nuanced. Hitters who consistently make hard contact, hit line drives, or run well will sustain BABIPs above .320; weak-contact pull hitters can sit at .260 forever. The right reference point for a hitter is not the league average but his own career baseline, regressed for contact quality.
Modern analysis has largely replaced raw BABIP with expected statistics — xBA, xwOBA, expected hit probability per batted ball. Those metrics control for exit velocity and launch angle and answer the underlying question (was this contact actually hit-quality?) more directly. BABIP still earns its keep as a quick first pass: a pitcher with a 4.50 ERA and a .360 BABIP is, more often than not, a buy candidate even before you open the Statcast page.
We don't model BABIP directly anymore — xwOBA does the same job with less noise — but pitcher BABIP gaps remain a useful tiebreaker when we're choosing between two starters with similar projected lines.