Fielding Independent Pitching — a pitcher rate stat built only from outcomes the defense cannot influence.
Fielding Independent Pitching was designed to answer a simple question: how would a pitcher's ERA look if his defense and his batted-ball luck were exactly average? Voros McCracken's research in the early 2000s found that pitchers exert far less control over balls in play than scouts had assumed. Two pitchers with identical strikeout, walk, and home-run numbers will post very different ERAs depending on the defense behind them and the bunching of hits against them.
FIP isolates the events the pitcher fully owns. Strikeouts and walks are entirely on him. Home runs are mostly on him — once the ball clears the fence, no defender can do anything about it. Hit-by-pitches are on him too. Every other batted ball is folded into a league-average constant.
The output is scaled so that league-average FIP equals league-average ERA, which makes the two numbers visually comparable. A 3.50 FIP and a 3.50 ERA mean roughly the same thing about pitcher quality, even though they're built from different inputs.
FIP is more predictive than ERA over modest samples and is the standard pitcher rate stat on most public sites. Its weakness is that it treats every home run as the pitcher's fault, ignoring park effects and the run environment — a pitcher in Coors Field will look worse by FIP than a true-talent equivalent in San Diego. xFIP and SIERA address that by neutralizing the home-run rate.
Our pitcher prior blends FIP with xERA and recent form. FIP gives us the long-run skill estimate; xERA gives us the contact-quality update; the blend is what feeds into starter run-allowed projections.