FanGraphs Logo

Forums | FanGraphs Baseball » Baseball Statistics

Measuring SP Consistency

(1 post)
  1. Newcomer
    Member

    Talk about starting pitchers often includes descriptions of consistency as a positive value to contrast with upside. "He's no ace, but he's a steady 4," "He'll keep you in the game when he's on the mound, and that's all you can ask of a 5th starter," "He knows how to pitch even when his stuff's not there," etc.. That got me to thinking about how we could quantify that, and here's what I've got.

    You can use ERA, FIP, tRA, or whatever you want as a base. But the idea is you get two numbers.

    Part 1: ERA is an average. It tells you how many runs a pitcher a pitcher allows per nine innings pitched. It tells how the pitcher will be performing, on average.

    Part 2: You take a look at the pitcher's performance in each GS, and you want to determine the level of production your starter matched or exceeded in, say, 80% of his starts (an arbitrary number). We'll call this ERAc, c for consistency. This is what the manager can feel pretty confident he'll get, even if the pitcher isn't really on. 80% might be demanding a lot of consistency, so maybe something like 65% (a good majority of his starts) would be a better bet.

    Here's where it gets fun. Take the difference, ERAc-ERA, and you have the consistency factor, CF, which tells you how close to his average performance a pitcher performs in at least x% of his starts.

    I imagine this would take a large sample to become stable, but we're usually talking about long-time veterans when consistency comes up. I'm not ready to put in the work just yet, but I'd like to see what kind of results we get. How wide is the spread between, on the one hand, pitchers who perform about the same every time out, and on the other hand, pitchers who are either really on or really off? It would be nice to see this expressed in a number, or maybe a set of numbers based on different percentages. People often look at how long a pitcher stays in the game, or how often they got pulled before 5 innings. This could provide another measurement of consistency.

    Thoughts? Suggestions? Has this been done somewhere? Is there some obvious reason I'm missing why it won't work?

    Posted 3 weeks ago #

RSS feed for this topic

Reply

You must log in to post.

Player Linker - Contact Us - Terms of Service - Privacy Policy