An Exceptional Season in the Annals of Stealing Home

In the first game of the 2021 ALDS between the Rays and Red Sox, Randy Arozarena stole home. With two outs, two strikes, and a dangerous left-handed hitter at the plate, the defense wasn’t worried about the runner at third. Arozarena took advantage, sprinting for home as southpaw Josh Taylor began his methodical windup. The speedy rookie timed it perfectly, taking off as soon as Taylor turned around and launching into his slide by the time the ball left the pitcher’s hand.

It was the first straight steal of home in the playoffs since Jackie Robinson accomplished the feat in 1955. Arozarena also made history by becoming the first player to ever homer and steal home in the same postseason game. But his stolen base wasn’t just historic in and of itself, and his jump wasn’t the only thing he timed perfectly. By stealing home that October, he tied a beautiful bow atop one of the most impressive league-wide seasons in the history of stealing home plate.

Stealing home is hard. In the last 50 years, 3,228 baserunners have tried; only 864, or about 27%, have crossed the plate safely. Efforts to steal home account for just 1.6% of all stolen base attempts, and successful steals of home account for just 0.6% of all stolen bases:

But something funny happened in 2021: Runners who tried to steal home were succeeding more than usual. In fact, they were succeeding more often than not. During the regular season, 17 baserunners were triumphant in their quest to steal a run; only 14 were stopped in their tracks. That’s a success rate of nearly 55%, more than twice as high as the overall success rate of the last fifty years.

As you might have guessed, that’s rather unusual. But just how unusual? Well, since 1973 (the first year that play-by-play records are complete on Baseball Reference), the league-wide stolen base success rate had never been higher than 50% — until 2021. To make things even more interesting, the only year (since ’73) the league has ever stolen home at exactly a 50% success rate was 2020, when runners were safe on six of twelve attempts. Thus, in each of the 47 years from 1973 to 2019, runners were caught stealing the plate more often than they were safe. The league-wide success rate varied widely, from 14.4% to 42.4%, but it had never reached 50%. Then, all of a sudden, the league hit that number two years in a row. Runners stole home a total of 23 times in 43 attempts, an unprecedented success rate of 53.5%. Previously, the success rate over two consecutive seasons had never been higher than 41.3%:

Even if you include the incomplete data Baseball Reference keeps from 1915 to ’73, the 2021 season still holds the second-highest stealing-home success rate in recorded history, trailing only the 1939 campaign. Meanwhile, the 2020 and 2021 seasons continue to hold the record for the highest-ever combined success rate in two consecutive seasons. The only other pairs of consecutive years to top 50% were the 1939 and 1940 seasons and the 1923 and 1924 seasons:

So was this the start of a new trend in stealing home? Perhaps runners would continue to steal the plate at more and more efficient rates; alternatively, might they be inclined to break for home more often to maximize the benefits of their newfound efficiency? Unfortunately not. As it turns out, things basically went right back to normal. After two ground-breaking years in home-stealing efficiency, baserunners simply returned to their old ways in 2022. Runners attempted 32 steals of the plate — one more than the year before — but were safe only 10 times. That’s seven fewer successful attempts and eight additional outs at the plate:

Considering how quickly things reverted to the status quo, I became even more curious about what happened in 2021. One explanation I considered was the unusually high workload for left-handed pitchers that season. Southpaws were on the mound for 29.8% of plate appearances; that’s tied for the highest percentage of batters faced by lefties in a season since 2002 (when our splits data begins). Runners can take a bigger lead off third base with a lefty pitching, so it stands to reason they might steal home more often against left-handed pitchers.

However, if the abundance of left-handed pitchers were encouraging runners to steal home, you’d expect to see an increase in attempts, not just successes. Yet attempts to steal home were actually quite low that year — the third-lowest total ever in a full season. What’s more, the vast majority of steals of home in 2021 (14 out of 17) came with a righty on the bump. None of the 17 stolen bases were straight steals either; they were almost all part of a double steal, which means the handedness of the pitcher was less relevant.

With that theory disproven, I turned to another: the low number of attempts. By and large, attempts to steal home are less frequent than they used to be:

However, from 2014 to ’19, attempts to steal home began to rise again, both in raw numbers and as a percentage of total stolen base attempts. The trend peaked in 2019, when runners made 53 attempts to steal home, accounting for 1.7% of all attempted stolen bags. Stealing home hadn’t been so popular since the mid-90s. The next year, however, the number of attempts dropped right back down:

In 2020, runners began picking their spots more carefully. That practice continued in 2021. More selectivity could certainly explain why prospective home-stealers were more successful. That being said, it’s not as if runners became selective to an unprecedented degree. Efforts to steal home accounted for 1% of all stolen base attempts in 2020 and 2021. That’s the same rate as in 2017, and it’s more than in every year from ’13 to ’16. What’s more, this doesn’t explain why the success rate dropped so steeply in 2022; runners were just as selective that year.

With another theory debunked, I returned to the film room. After watching and re-watching every steal of home from the 2021 season, I’m ultimately not convinced there’s any singular explanation other than season-to-season variance. If anything, I’m more inclined to blame poor defense than to credit smart baserunning. That’s not to say the runners deserve no praise, but plenty of those steals would have been outs with a slightly better throw:

While others might not have happened at all were the fielders paying closer attention:

It may seem hard to believe that random variation can explain such a statistical outlier, especially since the 50% success threshold is so rarely crossed and the average success rate is so much lower. But runners have been getting significantly more efficient at stealing home in recent years:

Over the last ten seasons, the league-wide success rate for stealing home is 36.7%. Even if you exclude the numbers from 2020 and ’21, it’s still 34.2%, which is significantly higher than it used to be. Throughout the previous four decades (1973–2012), the success rate was only 25.6%. In that time, the year-to-year success rate could be as low as the mid-teens or as high as the low-40s. Therefore, if the average success rate is now in the mid-30s, 50% doesn’t seem like nearly as much of an aberration. It’s still unusual — historic, even — but hardly unbelievable.

Finally, it’s worth considering that a 50% success rate isn’t necessarily a trend teams want to maintain. The vast majority of attempts to steal home come with two outs. With two outs, the breakeven point for stealing home is only about 30–35%. Moreover, most of these attempts also come as part of a double steal. In that situation, the breakeven point is even lower, since the worst-case scenario doesn’t change (the runner is out and the inning is over) but the best-case scenario is even better (a run scores and another runner moves up into scoring position). In other words, if runners are regularly stealing home with a 50% success rate, it’s not a sign of smart baserunning; it’s an indication they aren’t taking enough chances.

Therefore, while such a high success rate for stealing home is a fun statistical anomaly, a league-wide success rate of around 30% is ultimately where we want to be. That’s the best way to maximize run scoring, and it improves the fan experience, too. I don’t know about you, but I’d like to live in a world where runners are stealing home as much as they reasonably should.