Catching Up with the Oft-Overlooked Double

In the context of 2020s baseball, doubles in some ways have an uphill battle to capture our attention. After all, home runs are generally hit farther! And harder! Home runs are crucially less subject to chance: they are home runs without the benefit of anyone’s first baseman masquerading as a range-limited left fielder, for instance. Meanwhile, doubles can come to be as pop-ups lost in the sun, an outfielder’s failure to charge a chopper through the infield sufficiently, or a grounder that glances off third base.

The point being, doubles aren’t so useful in making predictions as home runs (or any of the three true outcomes). And baseball, ever increasingly it seems, is all about making predictions. Doubles don’t correlate particularly strongly with launch angle or exit velocity and can be rather fickle. However, specifically given that relative lack of attention, what follows will be about doubles.

To offer some necessary context around doubles, the chart below tracks all extra base hits, by season, dating back to 2000.

Well, for one thing, doubles (and triples for that matter) don’t benefit from the dramatic dips and spikes that home runs have of late. From 2013 on, really, there hav been fairly regular 500 hit season-over-season fluctuations in raw home run figures. In the five season period from 2014 to 2019 alone, raw home run figures jumped 61.9%. Hence, the ongoing discussion about the baseball.

Meanwhile, doubles are both more common than home runs and less volatile from season to season – facts that don’t lend themselves to evolving discussions. They are also conceivably less subject to fluctuations based on changes to baseballs. Still, doubles have had a clear development of their own over the last decade or so: they are in decline. This is pure unbridled conjecture, but that trend might be a result of two factors: first, players adjusting their launch angles to hit a few more home runs at the cost of some doubles and, second, outfield defense (and the shift) becoming more sophisticated over the last handful of seasons.

In 2000, the ratio of doubles to home runs was 1.56:1. In 2010, that ratio was 1.84. In the 2019 season (excluding 2020 as it was played exclusively during a more home run-prone time of year) that ratio had dipped to 1.26:1, a ratio fairly similar to that of 2021 to this point.

Speaking of 2021, even with a handful of games remaining, extra base hits in general look to end up considerably down relative to recent years; both home run and doubles totals lag behind 2019 (which, to be fair, was the high-water mark for XBHs) by roughly 1,000 hits.

Despite the fairly consistent season-over-season totals of doubles relative to home runs, doubles themselves take on far more shapes and forms than home runs, by definition, can. To very generally illustrate this point, the histogram below visualizes the distribution of exit velocities for all doubles hit in 2021 to this point.

The distribution of exit velocity of doubles has been distinctly left-skewed, demonstrating the relative frequency of softly struck doubles. To frame this distribution, the season’s minimum and maximum home run exit velocities are represented by the turquoise blue lines. No home run was hit softer than Abraham Toro‘s July home run which left his bat at just 88.7 mph; Manny Machado, on the other end of the spectrum, hit a home run on August 20th that left his bat at 119.6 mph.

As for that long left-side tail, 1,209 doubles to this point in 2021 had exit velocities of less than 88.9 mph. Given there have been 7,581 doubles that year with just a handful of games remaining, nearly 16% (15.95%), or one-sixth, of doubles were hit softer than the softest home run in 2021. Shohei Ohtani, somehow, hit a double this season with an EV of just 29 mph – an all season low. Conversely, not a single double recorded by Statcast came off the bat faster than Machado’s 119.6 mph home run.

Similar to exit velocities, the launch angles of doubles in 2021 are considerably more varied.

The distribution of doubles’ launch angles appears to be more normal than in the case of exit velocities, but there remain substantial tails on either side of the distribution. 498 doubles came off the bat at less than 1 degree in 2021; 142 doubles had a launch angle of >40 degrees. Home runs, meanwhile, adhere to a much narrower distribution. Eric Haase struck the lowest launched home run in 2021 according to Statcast, coming off the bat at just 13 degrees. Xander Bogaerts, on the other hand, hit a home run at 50 degrees (!) off the bat; unsurprisingly, that ball went just over the Green Monster.

Finally, a quick look at where those 2021 doubles have been hit.

Clearly outfielders, to some extent, are out there doing their jobs. A visualization like this also seems to validate and justify the increasingly common four-man outfield defensive setups employed against those batters disinclined to hitting pitches on the ground the opposite way.

While this data from Statcast comes via doubles in all stadiums, which generates some fuzziness overall, it does well in illustrating the vast variety in how doubles come about. In baseball today, where prediction is currency and randomness is sought to be minimized, it’s an interesting view into one of the more frequent, yet haphazard, outcomes of an at bat.