Going Deep: The Intentionality of Stealing Bases

Austin Bristow II takes an in-depth look at stolen bases, including a new statistic to measure stolen base intention.

(Photo by Carlos Herrera/Icon Sportswire)

Chasing stolen bases has been an exercise in patience and clairvoyance since the dawn of fantasy baseball. There are some players that we can routinely count on to carry our teams in steals. Yet not even they are wholly reliable.

Billy Hamilton is possibly the most prolific base-stealer of the past decade, stealing over 100 bags in the minors twice and never totaling fewer than 54 steals once he saw regular playing time in 2014. According to Baseball Savant, Hamilton is the third fastest man in MLB, topping out at 30.0 ft/sec, or 20.45 mph for some perspective.

His speed, along with his talent of knowing when to strike, has enabled him to achieve an 85.7% success rate on the basepaths. And yet, Hamilton is only on pace for 31 steals through his first 50 games of the 2018 season.

Speed and base-stealing ability are only two factors in predicting a player’s stolen base total. I see two others, one fairly obvious and the other a bit more subtle.

The third and most important component of base-stealing is getting on base. In 2018, Billy Hamilton is posting a career-low .287 OBP to go along with his .200 batting average. He simply has not given himself opportunities to steal second, let alone third base. As the old baseball adage goes, you can’t steal first base.

However, Cesar Hernandez, who ranks as the fastest second basemen at 29.1 ft/sec, sports a .384 OBP, yet only has 9 steals in 11 attempts, the exact same figure as Hamilton. This wide margin of difference between OBP and steals can be explained by the fourth component of stealing bases: intention.

Once a player reaches first base, it is in his hands whether or not he decides to run to second. In this way, intentionality plays a major role in whether or not a player steals a base. A fantastic example of this is Manny Machado. In 2015 the Orioles star stole 20 bases in 28 attempts while posting a .359 OBP. In the following season, Machado posted a .343 OBP, but only attempted three steals, failing to reach in all three attempts. Within a single season, Machado dropped his steal attempts from 28 to 3, despite only a small drop in his on-base percentage and sprint speed (27.5 in 2015 to 26.1 in 2016).

In order to better quantify a player’s intentionality in stealing bases, I’ve developed a statistic that should do just that. SBot, or stolen base opportunities taken, looks at how many times a player gets on base and decides to attempt to steal a base. The equation, (SB+CS)/(PA*OBP), favors players that attempt a lot of steals while maintaining a low on-base percentage. The top 10 players in SBot between 2015 and 2017 are shown below.

Billy Hamilton 0.440 174 29 0.298
Dee Gordon 0.332 148 43 0.34
Rajai Davis 0.299 90 21 0.302
Jonathan Villar 0.285 92 28 0.339
Starling Marte 0.239 98 26 0.345
Eduardo Nunez 0.217 72 21 0.332
Delino DeShields 0.214 62 19 0.333
Leonys Martin 0.207 45 15 0.283
Ben Revere 0.205 66 18 0.311
Hernan Perez 0.193 52 12 0.286

From what I’ve gathered from my initial research, Billy Hamilton’s .440 SBot is a massive figure, possibly historically high. Looking at some of the most prolific base-stealers of all-time, you’ll find that Rickey Henderson (1406 SB, .401 OBP) posted a .325 SBot in his career and Lou Brock (938 SB, .343 OBP) posted a similar .323 SBot despite his significantly lower on-base percentage.

In his six seasons as a major league player, Billy Hamilton has put up a .439 SBot, mainly due to his abysmal .297 OBP. From what I can tell thus far, a SBot over .200 is very high, indicating a strong willingness to attempt a steal whenever on the basepaths. You’ll notice in the table above that none of the players with a top 10 SBot over the past three years posted an OBP over .350 in that timespan. SBot definitely favors low OBP players, simply by mathematical definition. Nonetheless, Jose Altuve (100 SB, .386 OBP) comes in at #17 in SBot over the past three years with a .162 mark, the only player besides Hamilton and Dee Gordon to post triple-digit steals in that span.

So, what does SBot (pronounced extremely similarly to spot) tell us; what can we do with it? We can use SBot to look at how often a player will run whenever they have the opportunity. A simple Google search can pull up articles from spring when players talk about wanting to be more active on the basepaths, but SBot is a quantifiable metric that shows a player’s results of that intent. Assuming there is no significant change to a player’s running intention or on-base percentage, their rate of stolen base attempts ought to remain fairly similar.

Using this perspective, we might conclude that Billy Hamilton’s .220 SBot in 2018, when coupled with his decrease in OBP, likely reflects a change of intention on when to run. In accordance with a drop in OBP, Ender Inciarte’s SBot is at a career-high .341 mark, meaning that he may continue to run at this impressive clip until instructed otherwise. What may be the most interesting takeaway from the 2018 SBot numbers is Jose Altuve’s .061 mark. Despite a notable decrease in OBP, Altuve’s SBot also dropped to a career low in 2018, possibly indicating he is less willing to steal than he was in previous years.

Here’s a quick look at the 2018 SBot leaders:

Ender Inciarte 0.341 18 5 0.312
Dee Gordon 0.278 16 2 0.325
Mallex Smith 0.269 11 5 0.369
Michael A. Taylor 0.254 11 1 0.254
Billy Hamilton 0.220 9 2 0.287
Tim Anderson 0.213 11 1 0.301
Starling Marte 0.203 10 3 0.366
A.J. Pollock 0.190 9 2 0.349
Jean Segura 0.187 12 2 0.338
Tommy Pham 0.183 8 5 0.394

I do want to acknowledge a few shortcomings of the stat. Changes in on-base percentage ought to be heavily considered when looking at changes in SBot, as the two rate stats correlate extremely highly. Additionally, SBot does not account for the situation. There may be times when a player gets on base, yet has a teammate on second base in front of them. Other times, a manager may tell his baserunner not to attempt a steal, such as tight games in the late innings. I do plan on continuing my research into this new statistic, looking for ways to implement situational data into SBot findings. I hope to be able to account for the score of the game, the inning, how many outs, who is pitching, and who is catching, along with other situational aspects that would affect a runner’s intention.

Despite these limitations, stolen base opportunities taken is a step towards recognizing that there is more to stealing bases than being fast and getting a good jump. At the end of the day, as with most things in life, you have to want it.

Austin Bristow II

Raised as an Atlanta Braves fan in central Illinois, Austin Bristow II attended Eureka College for undergrad and Purdue University for his master's degree in Higher Education Administration. Since co-founding his home league at age 16, Austin has been obsessed with fantasy baseball. Austin serves as the Staff Manager for Pitcher List.

3 responses to “Going Deep: The Intentionality of Stealing Bases”

  1. Brett says:

    Great stuff Austin, looking forward to seeing how the statistic continues to evolve.

    On that note, have you done any preliminary research to its predictive capabilities? How strong is the year over year correlation?

  2. Harper Wallbanger says:

    Maybe it’s kept intentionally simple, but one recommendation to improve the stat would be to use 1B+BB+HBP for the denominator instead of PA*OBP. This eliminates homers, triples, and doubles, cases where an SB is impossible or unlikely.

  3. Jeremy says:

    Is there a way to filter out times when a player gets on base with someone in front of them who prevents them from stealing? That seems like an important part of SBot.

Leave a Reply

Your email address will not be published. Required fields are marked *

Account / Login