Do Speedy Players Really Put Pressure on a Defense?

Some clichés are unique to the hometown announcing team — The White Sox’s TV voice, Hawk Harrelson, has one for every possible occasion, though some announcing teams are a little more judicious with their use — but there are also truisms that are far more widely used throughout the game. Some have been around so long, they don’t even seem like clichés anymore, they’re just part of the vocabulary of baseball.

By and large, as long as the repetition doesn’t bother viewers too much, these common turns of phrase are relatively harmless, though many of them assert truths that may not be quite right. Routine groundballs are seldom exactly that and may give viewers an unkind view of fielders who have trouble corralling them even if the particular grounder in question was anything but routine, for example. That particular turn of phrase is virtually impossible to test on any sort of larger scale.

One phrase that a number of broadcast teams have used – likely all of them, though I cannot assert that with certainty – occurs when one of the team’s faster runners reaches on an error by one of the opposing infielders. “They know he’s so quick coming out of that box” the color man might quip. “His speed forced the defense into that error!”

If that our hypothetical broadcaster is right and faster runners do reach base via errors more frequently than their leaden-shoed counterparts, then it would have two major impacts. First, since it would be something of a skill, we should think about including reaching on error (hereafter, ROE) within on-base percentage or at least including it for fantasy calculations due to the extra stolen base chances. Second, and perhaps more importantly, there should be a relatively high – and testable – correlation between stolen bases and ROE.

Rather than having to explain away why some basestealers are more effective than others and how that affects the numbers, I’ve chosen to look at total stolen base attempts (SBA) or CS+SB. There isn’t a perfect correlation between speed and stolen base chances, but in lieu of 40-yard dash times, it’s probably the best proxy variable available.

To the first possibility, the impact of a few extra times on base might not seen all that important, but Jose Reyes reached base on an error some 13 times last season, which isn’t trivial. If he turned even a quarter of those extra times on base into a successful steal – either directly stealing second or stealing third later in the same inning – that’s an extra four steals that he would have otherwise lost. It’s not quite the difference between batting average and OBP of course, but just like savvy fantasy owners in traditional 5X5 leagues know to look at a runner’s walk rate instead of just their batting average, reaching on error could theoretically be another way of explaining how someone like Rajai Davis can steal 40 or more bases despite struggling to get his batting average over .260.

Looking at the five seasons from 2008 to 2012, the correlation between SBA and ROE never rises above .32 as this table shows:

Year	Correlation Coefficient
2008	0.316
2009	0.314
2010	0.303
2011	0.310
2012	0.296

For something with a relatively low correlation in an absolute sense, the numbers over the last five years have been remarkably similar. It would be incorrect to say that speed has no influence on whether or not a player frequently reaches via defensive incompetence, but such a low correlation tells us that, at the very least, there is a big difference in the skills it takes to steal a base and to reach on an error. That is, if one can call reaching on error a skill at all.

Among the noted speedsters who were tied for the league lead in ROE over the last five seasons are players like Nelson Cruz, Torii Hunter, Derek Lee, and Corey Hart. While a few slower players at the top of the leaderboards are no proof in and of themselves, they do illustrate the point nicely: Errors are the makings of the defense more or less irrespective of who put the ball in play. At the other end of the speed spectrum, the aforementioned Davis stole 46 bases and was thrown out 13 times in 2012, giving him 59 SBA, the most of anyone with 300 PAs or more. He reached on via an error naught but twice.

Lest this turn into an argument over the difference between hitters who ran a lot and hitters who actually ran effectively, I ran the numbers again and in all five years the correlation was almost identical, and in four of the five years the number was actually just a hair lower than the correlation for total attempts.

So where does this leave the two previously hypothesized outcomes?

First, adding ROE to OBP for fantasy purposes isn’t going to be particularly instructive. The extra times on base don’t reliably turn into extra stolen base attempts and they aren’t something owners will be able to count on from week to week let alone from season to season. For some players, they may be able to grab an extra bag or two when the get on base but don’t see a benefit in their OBP, but take it as little more than nice bonus.

As far as the announcer’s friend is concerned, it’s hard to make a definitive truth judgment one way or another. A young third baseman who sees Ben Revere or Mike Trout throw on the afterburners as they leave the box on a slow roller up the line well may hurry his throw and in doing so make an error, which would prove the old adage right, except that it’s impossible to say whether a sufficiently fleet of foot runner would have beaten the throw anyway. That said, the vast majority of the time the runner’s speed has very little to do with whether or not he’ll reach via an error.

It isn’t worth peppering the announcers who make themselves available on Twitter with this information, patterns of speech are notoriously difficult to change in adulthood and anyway it hardly matters to the casual viewer. But for fantasy owners, don’t let their figures of speech color otherwise sound judgment: While speed may consistently lead to stolen bases and in rare occasions contribute to defensive miscues, the two have very little to do directly with one another.