Fantasy Baseball Trends · Buy-Low, Sell-High, Hot & Cold
HR/FB Regression Candidates
Pitchers running a home-run-per-fly-ball rate well away from the league mean of roughly 14 percent. HR/FB is one of the least sticky pitching outcomes, so a pitcher at twice league rate with an inflated ERA is usually a buy rather than a problem.
HR/FB Regression
HR-per-fly-ball is one of the least sticky pitching outcomes — it regresses hard toward the league mean (~14%). When a pitcher’s ERA and his contact-quality xERA disagree and HR/FB is the reason, the ERA is the number that moves.
Methodology — luck, not just the rate:
- xHR/FB — the HR/FB a pitcher deserves, modeled from his barrel rate (barrels are the batted balls that become homers). Self-calibrated to the league daily.
- Luck = HR/FB − xHR/FB. A big positive gap means homers are leaving the yard that his contact doesn’t justify → they regress → buy.
- BRL (barrel% allowed, league ~8%) is the tell: low barrels + high HR/FB = unlucky. A pitcher with high barrels is earning his homers — those are filtered out, not bought.
- ERA → xERA — the contact-quality ERA. The gap (ERA − xERA) is the runs that should come off as the HR luck normalizes.
HR/FB is well above what his contact quality justifies (low barrel rate) AND ERA runs above xERA. The homers are unearned and should dry up — buy the ERA correction. Pitchers who are genuinely getting barreled are filtered out (their HR/FB is real).
0 pitchers
Click any column to sort.xHR/FB is fit on barrel rate (R² is modest by design — HR/FB is mostly noise, which is the point). The ERA correction is a regression signal, not a guarantee — park, sequencing, and a genuine fly-ball profile all matter. The BRL column is the guardrail: a pitcher getting barreled (flagged earned) is allowing real damage, not bad luck.