The job responsibilities of catching position can be very nuanced and many of the things that make a good backstop are attributes that rarely get noticed by fans. As an example, a recent ‘Fancy Stats‘ article by Neil Greenberg discussed the effect that Nationals catcher Jose Lobaton has on getting his pitchers extra strikes due to his pitch framing ability, a very subtle skill that is near intangible in contrast to abilities like hitting prowess or handling an opponent’s running game with your throwing arm.
A similar skill that can also often go unnoticed from a pitcher’s perspective is pace—how quickly you are able to make a pitch, collect yourself, get the sign, and throw the next pitch. Given the effects of timing on the ultimate success of an at bat for a hitter and the need for a pitcher to disrupt this timing in order to get outs, pace can play an unheralded role in a pitcher’s performance.
Pace goes beyond a pitcher’s internal clock, with many factors based on the rapport a pitcher and catcher have with one another playing a role in the outcome and whether a pitcher’s pace is quick or slow; ultimately, there is a particular level of comfort that a pitcher has with a catcher with respect to pitch calling that can affect pace.
With this in mind, let’s take a look at how the Nats starting rotation’s pace stats look, with the trio of catchers used so far in 2014—Sandy Leon, Jose Lobaton, and Wilson Ramos—taken into consideration. First, let me briefly discuss the data. PITCHf/x data from Nats games through May 5th was collected to calculate pace between pitches, with careful curation of the data done in order to remove outliers.
Ultimately, curation involved removing data points that were longer than 60 seconds and less than 10 seconds. This was done to remove first pitches of an inning, pitches after a home run (in order to counter the various lengths of time it took for hitters to jog around the bases), pitches where replay was involved, and other data that was felt to be physically impossible, with the hope that this pruning would give us the best picture possible of the effects of catcher on pitching pace. With these considerations in mind, let’s look at some pace results:
|Gio Gonzalez||Jose Lobaton||25.068|
|Gio Gonzalez||Sandy Leon||24.174|
|Jordan Zimmermann||Jose Lobaton||26.310|
|Jordan Zimmermann||Sandy Leon||25.927|
|Stephen Strasburg||Jose Lobaton||26.349|
|Stephen Strasburg||Sandy Leon||25.975|
|Stephen Strasburg||Wilson Ramos||27.075|
|Tanner Roark||Jose Lobaton||25.621|
|Tanner Roark||Sandy Leon||24.483|
|Taylor Jordan||Jose Lobaton||27.286|
|Taylor Jordan||Sandy Leon||26.603|
For reference, here are each player’s average pace—note that these averages were calculated using the aforementioned criteria, for those who use FanGraphs’ pace statistic and find a roughly four second shift in the pitcher’s averages:
Across the board, pitchers are a little quicker when Sandy Leon is behind the dish. With the pitchers, Taylor Jordan appears to be the slow poke, even slowing down Leon’s typically quicker pace with the staff by roughly a second. Overall, we do see some effects of the catcher on a pitcher’s pace.
Is this a significant effect? Let’s run an analysis of variance (ANOVA) to see if it is—for those numbers averse, feel free to skip to the pretty picture further down the page.
Using pace as our dependent variable and pitcher and catcher as our independent variables, the ANOVA results are as follows:
Cutting to the chase, we find that catcher does not have a significant effect on pace, but (no surprise here) the pitcher toeing the rubber does (p=0.022). Briefly, a Tukey’s test to look at the average differences between catchers:
|Sandy Leon-Jose Lobaton||-0.176||-1.262||0.908||0.923|
|Wilson Ramos-Jose Lobaton||1.234||-1.588||4.056||0.561|
|Wilson Ramos-Sandy Leon||1.411||-1.457||4.278||0.481|
Regarding the statistically significant results between pitchers, this stat was driven by the differences in pace between Gio Gonzlaez and Taylro Jordan, the quickest and slowest members of the rotation, with a difference of roughly two seconds in average notching a p-value of 0.04, which is just satisfies the criteria for significance of a p-value at or below 0.05. Additional ANOVA modeling including pitch type and inning did not show any statistically significant differences in average pace.
For the numbers averse crowd, welcome back! Overall, we did not find any statistically significant effects of catcher on average pace (or inning or pitch type), but did with pitcher. For those who a little more visual, the scatterplots below show show pace across inning, broken down by both pitcher and catcher, confirming the first table of results showing Leon getting pitchers to work quicker than Lobaton or Ramos:
While we don’t see any statistically significant results, pace is nonetheless an important aspect of the pitcher-catcher battery, and while again not a significant result, the quicker a starter works, the more success he tends to have, using RE24 as our marker of success:
While statistically these results aren’t terribly robust, the effects of pace (and the catcher) on the game are innately important, not only in its potential to disrupt hitter timing and rhythm, but also on a pitcher’s teammates. The longer a pitcher takes to decide what to throw, the longer his defense sits in their crouches, awaiting the ball to be put in play. The longer they wait, the greater potential to lose focus on the game and become distracted.
Pace also plays a role in length of game. In a recent interview, Boston Red Sox manager John Farrell discussed how starting pitcher pace can negatively affect game length. Like many things related to the position, the catcher’s role on pitcher pace will remain a potentially critical piece in a game’s outcome, despite its statistically small effects.