Using Coefficient of Variation to Look at Consistency in 2013

Chris Moran penned an excellent look at the streakiness of Rays hitters Evan Longoria and Wil Myers in 2013 this morning. This reminded me of some really excellent work from the peerless Bill Petti that looks into consistency for hitters. His formula is easy enough to copy and paste into further work provided you don’t mind the grunt work of compiling the data for 2013. First, I wanted to take a look at the Rays hitters:

Avg. is the average wOBA for each game for each batter. You will notice that this is not the same as their wOBA on the year which uses Plate Appearances as the denominator instead of games, but that’s the point for this exercise. VOL is the measure of consistency that we’re looking into. It might be beneficial to see this as a sort of continuum:

Virtually every fan has proclaimed at one time or another that they wish their team or Player X was more consistent, but what of course they mean is that they want that subject to be consistently GOOD so it’s important to keep both goals in mind here. The Rays featured a mix, as you would expect. Sam Fuld was consistently terrible, while Longoria was consistently very good and he’s not alone with Wil Myers, Desmond Jennings and Ben Zobrist solidly living in the same zone. James Loney is a guy that’s a little less consistent, but still very good, while Yunel Escobar is a consistent guy that doesn’t quite bring the production. Living with constraints means that it is very difficult to field an entire lineup of very good hitters that bring that same excellent level of aptitude on a daily basis. We’ve looked at the Rays hitters, but let’s take this idea to the team level starting with hitting:

As a team, the Rays were in the running for most consistent team in the American League, but as you can see with the White Sox, consistency itself is not the end all, be all. In fact, the best offense in the league was also the least consistent and they’re joined at the bottom by a couple of other pretty good teams. In all, I found a correlation of .22 between Avg. and VOL indicating that there is some value to being consistent, though the weak relationship suggests that talent is the ultimate divider. Let’s switch over to pitching:

Here we see that the Rays were in the running for the LEAST consistent team when looking at pitcher wOBA on a game-to-game basis. This is despite the fact that they basically tied for the best average game-to-game wOBA with the Tigers who are not far off in the consistency game. On the whole, I found a correlation of -.33 here which would seem to indicate that consistency is less desirable on the pitching-side.

There’s a ton of questions this stuff brings up and I hope the reader takes the reins and runs with this stuff, because I don’t have a ton of interest in going much further. Thoughts for future studies would look at the relationship between pitching and hitting VOL and it would be nice to see some ranges of expected production for both sides. This is what happened and I look forward to reading what others put together on what it all means.

Advertisements

About Jason Hanselman

Rays fan.
This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to Using Coefficient of Variation to Look at Consistency in 2013

  1. Drew Pressoir says:

    Correct me if I’m wrong but the higher the VOL the less consistent, right? So doesn’t that make Fuld the least consistent player?

    • Jason Hanselman says:

      Thanks, Drew, you’re right. I flipped the labels on the vertical axis, but have since corrected. Appreciate you bringing that to my attention.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s