Improved and refined umpire graphs
In the latest chapter of my ongoing saga with umpire statistics, I’ve refined the graphs, and added new features like averages and standard deviation.
I also pared the data down to looking at just six umpires this time around. I picked them at random from the outliers on the scatterplot I did earlier.
The umpires are John Hirschbeck, Deryl Cousins, Lance Barksdale, Tim McClelland, Brian Gorman and Angel Hernandez.
UPDATE: I added Doug Eddings to the chart. He’s by far the biggest outlier among MLB umpires on the scatterplot, I had to see how he compared. Not well, as it turns out.
—–
Deryl Cousins has been at the edge or outside the upper standard deviation for a long time, only coming within range in 2006. Doug Eddings is skiiing down the slope of BB/9. He really doesn’t like to walk people. Does he have a huge zone? Let’s look at the next graph to see if we can determine that.
—–
And the companion graph, SO/9:
Eddings is near the top of the standard deviation, but he isn’t as strikeout-prone as he is walk-phobic.
Also, I’d be willing to bet that there aren’t two strike zones more dis-similar than the ones called by Deryl Cousins and John Hirschbeck.
—–
And finally, the graph that I think is most telling; balls to strikes ratio:
Doug Eddings again shows up well beyond the lower standard deviation. He just doesn’t call many balls at all.
I think this is the most insightful graph, for a couple reasons. One, there’s an issue with foul ball strikes counting against the umpire, though he has little to do with it. Two, the only thing that an umpire calls without anyone else being able to interfere is a ball. The more balls he calls, against the league average, the smaller a strike zone he has. I think that makes sense, don’t you?
March 13th, 2007 at 10:23 am
My subjective impression of umpires is that the same Ump may call breaking balls differently than fastball/changeup/splitters. It’s likely that, should that be true, umpires are likely to have a different effect on the results of different pitchers. For example, an otherwise generous ump who is stingy on curveballs is likely to hurt a guy like Zito a lot more than someone like Felix.