clock menu more-arrow no yes

Filed under:

How base stealing made a (mini) comeback in 2020

New, comments
MLB: Game One-St. Louis Cardinals at Chicago Cubs Kamil Krzaczynski-USA TODAY Sports

I did what I often do when looking for article ideas, which is digging through the leaderboards over at FanGraphs to see if anything catches my eye. I noticed that the number of steals (adjusted for season length) slightly increased in 2020.

It shouldn’t come as much surprise, then, to hear that stolen base attempts were also up last year.

Granted, teams were still considerably more hesitant to steal than they were at the start of the 2010s. To some extent, the home run surge in the back-half of the decade likely contributed to this trend. The higher the home run rate, the greater the relative cost of being thrown out stealing. That stolen bases waned during a power-driven era isn’t overly-shocking. What is intriguing to me, however, is the rise in stolen base success rate.

And if you want all of these graphs consolidated into one table:

Stolen Bases Data Table, 2011-2020

Season Games Attempts SB CS Success Rate SB/2430 Attempts/2430
Season Games Attempts SB CS Success Rate SB/2430 Attempts/2430
2011 2429 4540 3279 1261 72.2% 3280 4542
2012 2430 4365 3229 1136 74.0% 3229 4365
2013 2431 3700 2693 1007 72.8% 2692 3698
2014 2430 3799 2764 1035 72.8% 2764 3799
2015 2429 3569 2505 1064 70.2% 2506 3570
2016 2427 3538 2537 1001 71.7% 2540 3542
2017 2430 3461 2527 934 73.0% 2527 3461
2018 2431 3432 2474 958 72.1% 2473 3431
2019 2429 3112 2280 832 73.3% 2281 3113
2020 898 1177 885 292 75.2% 2395 3185

Fewer runners tested opposing defenses, but the ones that did were more likely to get away with their thievery.

What can explain this? I don’t think catchers were especially rusty in the unorthodox 2020 season. Their collective fielding percentage, .993, was the exact same as it’s been in seven of the past eight years. I would love to look at Statcast fielding data, namely, pop time, but those catcher metrics don’t appear to be publicly available for 2020. What we can do as a rough proxy is compare 2015 and 2019, since those two seasons have the largest stolen base success rate gap in the Statcast era.

Catcher Pop Times, 2015 vs. 2019

Season Pop Time on 2B Attempts Pop Time on 3B Attempts
Season Pop Time on 2B Attempts Pop Time on 3B Attempts
2015 2.02 1.60
2019 2.01 1.61

The average pop times for attempts at second and third base are...virtually identical. That would indicate the key factor here isn’t catchers. Using Statcast’s data, we can turn to the runners themselves and take a look at what percentage of steal attempts were at second base vs. third.

Stolen Base Attempts, 2B vs. 3B

Season 2B Share of Total Attempts Total Success Rate
Season 2B Share of Total Attempts Total Success Rate
2015 91.7% 70.2%
2016 89.0% 71.7%
2017 88.6% 73.0%
2018 90.0% 72.1%
2019 90.7% 73.3%

There doesn’t appear to be much here. The lowest success rate pairs up with the highest share of second base attempts, but the highest success rate corresponds to the second-highest 2B%.

Another possible explanation is that the success rate rose because better base thieves started comprising a greater share of attempts. This is reflected in the data...sort of. I took the top 10 and 25 leaders in stolen bases attempts and calculated what percentage of the MLB total they made up for that year.

Stolen Base Distribution, 2015-2020

Season Top 10 Share of Total Attempts Top 25 Share of Total Attempts Success Rate
Season Top 10 Share of Total Attempts Top 25 Share of Total Attempts Success Rate
2015 13.4% 26.1% 70.2%
2016 14.4% 28.2% 71.7%
2017 13.5% 26.1% 73.0%
2018 13.2% 26.8% 72.1%
2019 13.3% 26.6% 73.3%
2020 14.9% 28.0% 75.2%

It’s worth noting the top 10 made up a greater share of all attempts in 2020 than in any of the five preceding seasons. At the same time, though, 2016, the second-lowest success rate, had the second-highest top 10 share and the highest top 25 share.

In my mind, that leaves two main theories in play as to why the success rate surpassed 75% last year:

  1. Individual season-to-season data still suffers from small sample size shenanigans. In reality, there’s nothing major to see.
  2. Teams have gotten smarter with regards to when they steal, which pitchers they steal against, etc.

For now, I guess we can add this to my list of weird things I’ll be watching for in 2021.