What is LOB% (Strand Rate)? Definition, Formula, and Example
LOB% (left-on-base percentage, also called strand rate) measures the share of baserunners a pitcher prevents from scoring; league average sits around 72-73%.
What is LOB%?
LOB% — left-on-base percentage, often called strand rate — is the percentage of baserunners a pitcher allows who do not come around to score. If a pitcher gives up 20 hits and walks across a stretch of starts but only 5 of those runners eventually score, his strand rate is 75%. The metric isolates a pitcher's ability (and luck) at preventing runs once traffic is already on the bases. League-average LOB% across MLB sits steadily near 72-73% every season, which means the typical pitcher strands roughly seven of every ten runners he allows. Strand rate is one of the four "luck indicators" sabermetric analysts use alongside BABIP, HR/FB rate, and xERA to identify pitchers whose ERA is likely to regress.
How LOB% is Calculated
The FanGraphs/Baseball Reference formula uses ER (earned runs) rather than counting baserunners directly:
LOB% = (H + BB + HBP − R) / (H + BB + HBP − 1.4 × HR)
The numerator counts the total baserunners minus runs scored (i.e., runners stranded). The denominator counts baserunners minus a 1.4-weighted home run term — this adjusts for the fact that home runs always score the batter regardless of skill, so they shouldn't penalize a pitcher's strand-rate component. The 1.4 multiplier roughly accounts for the average number of runners on base when a homer is hit (the batter himself plus ~0.4 baserunners).
Anything above ~80% over a full season is almost always unsustainable and signals positive regression coming. Below 65% suggests bad luck or genuinely poor sequencing.
Worked Example
In Blake Snell's 2023 Cy Young campaign with the Padres, he posted a 2.25 ERA across 180 innings — extraordinary on its surface, but his LOB% was 86.7%, far above his career norm of ~77% and well above the league baseline. He also walked 99 batters, the most in MLB. Snell's FIP that year was 3.44 and his xERA was around 3.70 — both pointing to ERA inflation in 2024. Sure enough, his 2024 ERA jumped to 3.12 across an injury-shortened season as his strand rate normalized to 78%.
Compare that to a pitcher with the opposite problem: a starter running a 4.50 ERA but a 65% LOB% and a 3.60 FIP is leaking baserunners to bad sequencing or bullpen-assisted inherited-runner damage. He's likely to improve.
Why LOB% Matters
LOB% drives forecasting decisions across the industry. Fantasy managers sell high on pitchers running 82%+ strand rates and buy low on those stuck below 68%. Front offices use it during trade-deadline pitcher evaluations — a 4.20 ERA starter with a 3.40 FIP and a 65% LOB% is a buy candidate, while a 2.90 ERA starter sitting on 84% is a sell candidate. Bullpen leverage decisions also lean on it: managers track how individual relievers strand inherited runners, which is a related but distinct measurement (Inherited Runners Scored %).
For betting markets, sportsbook over/under lines for starting-pitcher props (strikeouts, runs allowed) account for strand-rate regression — a pitcher whose ERA is propped up by stranding everyone is priced more aggressively than the surface ERA suggests.
Limitations and Misconceptions
LOB% is not purely luck. Pitchers with elite stuff and high strikeout rates genuinely strand more runners because they can punch out the bases-loaded threat. Career strand rates for elite arms (Clayton Kershaw, Justin Verlander) sit around 77-78%, meaningfully above league average. But year-over-year variance for any individual pitcher is large, and single-season spikes above 80% almost always regress.
LOB% also doesn't account for inherited runners — those are credited to the previous pitcher. A reliever can have a great LOB% while still letting every inherited runner score.
Related Terms
In Legends Deck: Strand rate feeds our pitcher card "Composure" rating — pitchers with sustained above-average LOB% over multiple seasons get a small bonus to runner-on simulation outcomes, while one-year outliers are regressed toward the mean before card ratings are locked.