Should You Bring The Infield in With Runners on 2nd and 3rd and Less Than 2 Outs?

Photo Credit: Jim McIsaac/Getty Images

MLB teams have a strong understanding of when to bring the infield in with a runner on third and fewer than two outs. They also understand when to play the infield in versus setting up for a double play with runners on first and third and one out, or with the bases loaded and one out. In this analysis, I’ll focus on situations in which there are runners on 2nd and 3rd with less than two outs: Should you bring the infield in to cut off a run, but at the same time increase the risk of giving up a big inning?

My dataset includes roughly 26,000 pitches from these situations over the past 11 seasons in the Statcast era, making this a sub-1% edge case that teams may be able to further optimize. I focused on all 2nd and 3rd situations with less than 2 outs, an absolute score difference of 2 or fewer, and situations in the 5th inning or earlier. I chose these filters to reduce bias from lopsided games, where additional run scoring has little impact on win probability, and to limit late-inning bias, when teams may adjust behavior in response to anticipated run scarcity as the game approaches its end.

Table 1: Infield in analysis in 2nd and 3rd with less than 2 outs

From Table 1, teams bring the infield in 29% of the time with zero outs and 43% of the time with one out. Playing the infield back is effective at reducing expected runs, as mean_delta_run_exp is lower in those cases. However, in both situations, mean_delta_batting_team_win_exp is lower when the infield is brought in. This highlights the familiar trade-off between run expectancy and win probability—and suggests that MLB teams are not making the win-maximizing decision. In most situations with runners on second and third and fewer than two outs, teams should instead bring the infield in to maximize the probability of allowing zero additional runs in the inning and, in turn, maximize overall win probability.