Who is Leveraging, or Squandering, their Sprint Speed

The premise of this blog post can be summarized in a couple questions: (1) what is the relationship between player sprint speed and player value on the base paths and (2) which players are the greatest exceptions to whatever that relationship might suggest?

Fortunately, those questions are addressable due to FanGraphs’ BsR metric, which aggregates player actions on the base paths, and Baseball Savant’s Sprint Speed metric, which measures the fastest single-second sprint speed of players. With each of those data sources readily available, player data could then be joined so as to compare both metrics.

Wishing to have a more substantial set of base running data, I have opted to look at qualifying hitters from 2021 versus to this point in 2022, of which there were 132. Below is a scatterplot juxtaposing FanGraphs’ BsR metric with Sprint Speed.

It might come as little surprise that there appears to be a positive, if fairly noisy, relationship between any given player’s speed and the value they create while on base. The correlation between these two metrics is 0.62 and a simple linear regression (represented by the turquoise line above) results in an R-Squared value of .381. Without being particularly rigorous, it at least appears that player sprint speed influences their value as a runner.

Next, using a simple linear regression model, one can create a (very) rudimentary estimate for BsR using sprint speed as the solitary explanatory variable. In other words, the sprint speed of any particular player can be at least somewhat predictive of how valuable they are as a runner (e.g. if this player’s sprint speed is X, then the best guess for their base running value is Y).

Below is a table featuring those players from 2021 whose BsR metric most drastically underperformed relative to what sprint speed predicted.

According to Statcast, average sprint speed is 27 ft/sec, so a handful of the players in the table above in fact have above average speed, while performing below average on the base paths. Randal Grichuk, for instance, represented substantially negative value running all the while being more than one foot/second faster than his peers on average.

In fact, 6 of 10 players in the table above have average speed and still failed to create positive value while running on base.

On the other end of the spectrum are those that outperformed relative to what their sprint speed alone might suggest.

Nearly all of the players listed here have above average speed, yet still provided additional value while running relative to what speed alone predicted. Only one player, Elvis Andrus, made this Top 10 list while being slower than average.

One clear distinction between these two lists is simply the total of stolen bases across groups. Together, those who outperformed their predicted BsR value stole 266 bases in 2021, or nearly 27 apiece on average. Those players who most underperformed their predicted BsR value stole just 51 bases. BsR incorporates much more than SBs alone, but clearly being able to steal bases relatively frequently (likely too with an efficient rate of success) goes a long way in providing value on the bases.

It is clear that speed relates to value on the base paths, but that relationship is not so strong as to preclude obvious exceptions to the rule. Looking at the Top 10 tables above, it appears as though faster players still run (a pun, yikes) the risk of presenting below average value on base. Meanwhile, even a below average runner like Elvis Andrus can crack a Top 10 list for over-performers.