Double steals in Major League Baseball

Stealing a base in baseball involves risk but can create scoring opportunities. We evaluate the effectiveness of double steal attempts–in which two runners attempt to steal bases simultaneously–using observational play-by-play data from Major League Baseball spanning 1980 to 2022, plus the 2023 season following significant rule changes. Our outcome is run value, which measures the change in expected runs resulting from a play. To estimate the causal effect of a double steal attempt across different base-runner configurations, we apply propensity score methods adapted from the personalized medicine literature, specifically inverse probability of treatment weighting. We find that double steals with runners on 1st base and 2nd base are associated with increased run value, especially in 2023. In contrast, double steal attempts with runners on 1st base and 3rd base show no consistent benefit. These findings suggest that, although the double steal may be underutilized in Major League Baseball, it may be a valuable strategy when applied in the right context.