Umpire graphs

In a game where everything is analyzed and analyzed again, I've always found it odd that there has been very little work done (that I know of) with umpire's statistics. After all, umpires are subjective in how they call balls and strikes. I like it that way, though, I'm not about to suggest we expand QuesTec's power.

So if we know that umpires are subjective, shouldn't we be keeping track of how they're calling games?

Thankfully, Baseball Prospectus does that, although it's not well publicized. I'm not even sure if it's a stat they offer for free, since I subscribe.

The point of this diary, though, isn't to ramble on about that. It's to show you all what I've done with the stats BP provides, and ask some questions of you.

I've graphed out a couple key sets of data that I think are telling as to how an umpire is performing.

The first one is simply a scatterplot of BB/9 and K/9 with red lines to indicate MLB average, shown here:

I believe I'm right to assume that the umpires in the lower right hand corner favor pitchers, and those in the opposite corner favor the battter.

I also graphed out the percentage (of all pitches seen) of balls and strikes.

This time, as before, the lower right hand corner favors pitchers, and the upper left favors batters.

Some of the umpires remain in the same category, while others (such as John Hirschbeck and Randy Marsh) appear to switch who they favor.

I'd assume that the graph of pitch type is more authoritative, but what's causing the disparity? Could it have something to do with the pitchers themselves, since K/9 and BB/9 aren't completely controlled by the umpires, as ball and strike % are?

What do you think?

[there's more info and graphs, including more explanation, on my site, The Pastime]