Pitching to the Score

Jim McIsaac

MLB Network analyst Harold Reynolds recently claimed that CC Sabathia "pitches to the score". Do the numbers back up Reynold's statement?

Baseball personalities, like MLB Network's Harold Reynolds, often talk about "pitching to the score", the idea that elite pitchers perform at a higher level when the game is on the line. Last week in a segment on MLB Now, Reynolds claimed that Yankees ace CC Sabathia is an example of a pitcher who "pitches to the score". Dave Cameron of FanGraphs examined the data to find out if there is any indication that CC Sabathia performs better than usual in pressure situations:

In that segment, Harold Reynolds cites CC Sabathia as an example of a pitcher who pitches to the score, noting that he performs differently when the game is on the line than when he’s just trying to get outs and has some runs to give up. While one will never be able to definitively prove or disprove the intent of a pitcher, given that we are left to only measure what they do rather than what they are thinking, Reynolds’ claim is testable. If Sabathia pitched dramatically better in close games than with a big lead, it would show up in the data.

It does not.

FanGraphs keeps track of how players perform in low, medium, and high leverage situations, so Cameron began by taking a look at how CC Sabathia performed as leverage increased. The data showed that Sabathia's BB% increased in higher leverage situations, while his K% and HR/9 decreased.

I don't believe in "pitching to the score" when applied in the way Harold Reynolds used it to describe CC Sabathia, but the changes in BB%, K%, HR/9 lead me to believe that while the results do not differ much with a change in leverage, CC Sabathia may alter his approach depending on the pressure of the situation.

Baseball Reference tracks how pitchers perform by scoring margin. That is a useful resource in this case, since the amount of leverage a situation is assigned depends in part on baserunners, scoring margin splits would seem to be a better way of determining whether Sabathia truly "pitches to the score." Cameron noted that the numbers didn't suggest Sabathia performed differently when looking at the scoring margin. I took a look at the scoring margins splits and noticed a few things that unsettled me a bit.

The first issue is that the results for a "Tied Game" are skewed by the large percentage of PA that occur early in a game, when pitchers are making their first trip through the opponent's lineup and generally perform better. Another problem I encountered is that instead of listing results for when a pitcher's team was, for example, four runs ahead and four runs behind, Baseball Reference chose to put those results together, which isn't very useful.

While I agree that there is no indication that a pitcher performs better than normal in high leverage situations, the data does indicate to me that pitchers alter their approaches in some way according to the importance of the situation.

Analysts like Harold Reynolds often mention "pitching to the score" when defending their belief that we should put more stock into wins when evaluating a pitcher because they may give up more hits and runs when they are pitching with a lead. Even before I was aware of sabermetrics that type of assertion bothered me because it did not track logically. Reynolds is implying (whether he realizes it or not) that pitchers put less effort in when their team has a lead. It's the same lapse in reasoning that occurs when you hear someone suggest that closers are incapable of pitching well in non save situations. I could go on, but Keith Law did a great job of destroying that line of the thinking in an old episode of the Baseball Today podcast, so I'll include that below instead.

An Attempt to Further the Discussion

The answer to the question "Does CC Sabathia perform better in high leverage situations?" is simply: "No, but he does perform differently."

The most interesting question, in my opinion, is this: In what way does a pitcher alter his approach in high leverage situations? Here is a short list of things we would need to know to adequately answer that question:

  • Does the number of pitches per PA vary significantly depending on the leverage of the situation? And what would constitute a significant variation in pitchers per PA?
  • Does the pitcher's selection of pitches vary depending on the leverage of the situation? If it does, in what way?
  • Does the pitcher's velocity increase or decrease depending on the leverage of the situation? This would be a bit harder to determine because a pitcher's velocity can vary significantly from start to start or even within a game. (Anecdotally, Justin Verlander has been noted to "reach back for something extra" in certain situations.)
  • Does a pitcher's (and to a certain degree, a catcher's) choice of location vary depending on the leverage of the situation? If it does, in what way?
  • Note: Admittedly, this list is not exhaustive so feel free to leave your own thoughts or questions in the comments below.

    Btbs-twitter-insert_medium

    X
    Log In Sign Up

    forgot?
    Log In Sign Up

    Forgot password?

    We'll email you a reset link.

    If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

    Forgot password?

    Try another email?

    Almost done,

    Join Beyond the Box Score

    You must be a member of Beyond the Box Score to participate.

    We have our own Community Guidelines at Beyond the Box Score. You should read them.

    Join Beyond the Box Score

    You must be a member of Beyond the Box Score to participate.

    We have our own Community Guidelines at Beyond the Box Score. You should read them.

    Spinner.vc97ec6e

    Authenticating

    Great!

    Choose an available username to complete sign up.

    In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

    tracking_pixel_9351_tracker