After baseball's rule changes, the stolen base is officially back
Major League Baseball's decades-long embrace of analytics altered the game in countless ways, perhaps most notably by encouraging an infatuation with the home run at the expense of small ball tactics, including the humble stolen base. The numbers made it clear that the risk of losing a base runner while attempting a steal did not justify the potential reward, relegating the once-beloved tactic to a second-tier strategy employed only in select situations. Stolen base attempts plummeted from 4,365 in 2012 - a year after the release of the film "Moneyball," based on the Oakland Athletics' analytical approach to the game - to 3,700 in 2013. Teams averaged just 3,111 stolen base attempts from 2019 to 2022, excluding the pandemic-shortened 2020 season.
And now, suddenly, the stolen base is back. The batch of new rules introduced this season - including the pitch clock, the limit of two "disengagements" per batter (when the pitcher attempts a pickoff or steps off the rubber) and larger bases (that cut the distance between bags by 4 ½ inches) - have revived the efficiency of the stolen base as an offensive tactic. So much so that for the first time in decades, the reward for attempting to steal second base significantly outweighs the risk.
"I don't know all the advanced metrics," Trey Mancini told CBS Sports before the season, "but one thing that I do know is that stolen bases are going to go up for sure."
Mancini was correct. Through Wednesday, there have been 314 attempts to steal this season, or 0.85 per team per game. At a similar point last season - roughly 10 games into the season - teams had attempted just 0.47 stolen bases per game. You would have to go back to 2011 to find a comparable rate of stolen base attempts per game this early in the season. More importantly, the new rules helped boost the success rate to better than 81 percent in the early going for the first time since stolen base opportunities were regularly recorded in 1920.
Former MLB All-Star Bret Boone made a direct connection between the pitch clock and the increase in stolen base attempts on his podcast last week.
"I think you're going to see a lot more stolen bases this year, because not only is it a clock, but when the clock gets down to three, two, one, that pitcher's not worried about throwing over," he said. "He's worried about getting this ball to home plate before a clock violation."
To understand the value proposition, you have to start at the beginning. The point of a stolen base attempt is to advance a runner, thereby increasing the chances he scores a run. Assuming the number of outs is equal, a team is more likely to score at least one run - and perhaps more - with a runner on second than on first. That's the reward. The risk is getting caught and costing your team an out. Thanks to the millions of data points available in Major League Baseball, and using a win expectancy finder like Greg Stoll's, we know how many runs a team should score with a man on first and no outs (0.89 from 2015 to 2022), a man on second and no outs (1.14) and no men on with one out (0.27). With that data, we can estimate the risk-reward proposition of attempting to steal second under the new rules, which is considerably sunnier than it had been. (This discussion will be limited to attempts to steal second base because the sample sizes, and risk-reward calculus, are significantly different for stealing third base and home plate.)
The risk-reward calculation for stealing a base is relatively easy to evaluate. You simply multiply the reward (expected runs added, multiplied by base stealing success rate) and subtract from that the risk (expected runs lost, multiplied by the base stealing failure rate). For example, the reward for stealing second base with no outs to start this season - based on last year's run expectancy numbers - has been .20 expected runs (a gain of 0.25 expected runs by advancing to second multiplied by the early success rate of 81.3 percent), while the risk was 0.11 expected runs (a loss of 0.62 expected runs by getting thrown out, multiplied by the early caught-stealing rate of 18.7 percent). Since gaining 0.20 expected runs is greater than losing 0.12 expected runs, the risk has been worth the reward.
Last season, the reward for a successful stolen base attempt of second base was 0.19 expected runs gained, compared to a risk of 0.15 expected runs lost - still positive, but a less desirable trade-off than we are seeing this season.
While the stolen base is back, every team isn't running at the same rate. The Chicago White Sox have seen their stolen base rate skyrocket this season. After attempting a steal of second base once every three games last year, the White Sox were averaging one attempt per contest through Wednesday. Their overall attempts as a ratio of opportunities to start the season was up 171 percent year over year, the biggest increase in baseball. The Baltimore Orioles (+153 percent) and New York Yankees (+117 percent) have also seen a huge spike in stolen base attempts of second base compared to opportunities to do so (situations with a man on first). The Yankees are also revitalizing a base-stealing technique from 50 years ago, according to Ken Rosenthal of the Athletic. It involves taking a short primary lead with a lateral hop during the pitcher's delivery, giving the runner more momentum toward the bag.
At the other end of the spectrum are the Texas Rangers, Los Angeles Angels and Colorado Rockies. Each of those teams have seen their attempts per opportunity decline by about 50 percent this season.
"We're not really constructed to steal bases and move guys over," Angels Manager Phil Nevin told the Athletic. "We will do that. We'll play winning baseball. But we'd like to hit them in the seats."
But those teams have been the exception, and the combination of new rules and new techniques could actually lead to a new record. The most stolen bases, including third base and home plate, in any MLB season came in 1987, when 3,585 bags were swiped. Vince Coleman led the league with 109, 49 more than Harold Reynolds. (Coleman's total was also more than 22 teams recorded last season.) It doesn't appear any one player will come close to those marks; projections at FanGraphs estimate that Oakland's Esteury Ruiz, Atlanta's Ronald Acuña Jr. or Baltimore's Cedric Mullins or Jorge Mateo will lead the league with around 34 stolen bases. Yet if baseball as a whole maintains this level of running and success for the whole season, the 30 teams together would finish with approximately 3,350 stolen bases, good enough for the third-most over the past 100 years, behind 1987 and 1999 (3,421).
Related Content
Texas is home. The Briggle family and their trans child are fighting for it.
China's struggles with lab safety carry danger of another pandemic
How progressive Denmark became the face of the anti-migration left
