The validity of “pitching to the score” isn’t really a sabermetric debate. After all, it pits the importance of pitcher wins versus ERA, two statistics that have been tracked since the advent of the box score. Nonetheless, the discussion usually breaks down along party lines, with number crunchers often insisting the concept is a myth, while old school advocates extol it as a virtue.
So, does pitching to the score really exist, and how many hurlers actually do it? The easiest way to answer that question would be to poll the pitchers themselves, but there doesn’t seem to be a consensus among the fraternity. For every Cliff Lee, who seems to think some pitchers have an innate ability to win games, there is a David Cone, another former Cy Young award winner who is fond of saying, “It’s now how you pitch, it’s when you pitch.”
There’s value to guys who can go out there and just get wins. It don’t matter if you give up four runs or no runs. Just so long as you don’t let a team score more runs than your team scores. There’s definitely something to be said about that.” – Phillies’ ace Cliff Lee, quoted by Philly.com, July 17, 2013
Because there are so many variables involved in the equation, it’s impossible to definitively settle the “pitching to the score” debate. Two seasons ago, I took a look at Justin Verlander’s performance on a granular level and was unable to uncover evidence that he was making concessions to the score. This time around, I’ve decided to take a more broad approach. Where there’s smoke, there’s fire, so if pitching to the score is as commonplace as some suggest, there should be at least some anecdotal statistical evidence among a larger sample. In order to find it, a five-year comparison was made between the ERA and run support rates for the 22 pitchers who have had a qualified season in each of those years. The results are provided in the chart below.
Note: Sample includes all pitchers with a qualified season in each year from 2008 to 2012. RS/9 is the sum of all runs scored while a pitcher is still in the game divided by the number of innings he pitched and then multiplied by 9. The correlation coefficient (r) ranges from -1.0 to +1.0. The closer r is to +1 or -1, the more closely the two variables are related.
Not surprisingly, the range of correlation between ERA and run support is wide. At one end are Ryan Dempster and Ubaldo Jimenez, whose ERAs have been lower in seasons when they received more run support. At the other end of the spectrum are CC Sabathia and Edwin Jackson, two pitchers whose statistics suggest they just might be pitching to the score. In between these extremes, the rest of the sample has exhibited a varying relationship between getting and giving up runs. However, only six pitchers, or approximately one-quarter of the sample, have shown a positive correlation of at least 0.5 between the two statistics.
Sabathia’s position near the top of the list is interesting because the big lefty fits the “pitch to the score” profile. As a workhorse on a typically high scoring team, Sabathia would be expected to give up a few extra runs, but, in reality, the Yankees’ ace has actually had an inverse relationship between ERA and run support while in pinstripes. As the chart above shows, Sabathia’s ERA did move along with run support earlier in his career, but when taken as whole, it doesn’t appear as if the big lefty has made a habit of pitching to the score.
Just as correlation doesn’t imply causation, the lack of a connection doesn’t preclude one from existing. Pitchers may very well be adjusting their approach based on the score, but other factors could be muting the impact. Similarly, the limited sample size in the snapshot above, both in terms of the number of pitchers and seasons included, prevents the analysis from being conclusive. However, failure to prove a negative is not how theories are confirmed. Besides, it doesn’t really matter whether pitcher are adapting to the score. The only thing that counts is whether their adjustments are impacting performance. So far, “pitching to the score” advocates have failed to illustrate that relationship.