Swivel graphs
Up until now, I’ve been using OpenOffice’s built in graphing feature to produce the many graphs I’ve posted on The Pastime. It’s kind of a hassle, but one that I’m used to.
I got an email earlier today from the guy who runs Swivel.com, telling me that he’d borrowed some of the data I’d used for the unified projections to show me what his graphing utility could do. I’d have to say I’m impressed by the potential. Swivel will take a data set you’ve uploaded and create a bunch of graphs with it. It takes a little tweaking to get them to look like you want, and there are still a few features that I’d like to see added. All in all, it’s a nice tool.
I particularly like the speed at which data can be turned into various graphs. I’m sure as it develops, you’ll see swivel graphics on here more often.
Here’s one representation that I knocked together quickly. I fed the umpire ball and strike percentage data to swivel, and this is what it came up with.
Very similar to my original graph, though I wish it would display names on scatterplots. I’m still playing with it, though.
January 17th, 2007 at 3:48 pm
Me again. Sorry, but I don’t feel this graph is necessary. You have to ask yourself what this scatterplot is showing. It shows that umpires who call more balls call less strikes. It also shows that umpires who call less balls also call more strikes. That should not be surprising.
As you suggest in your post, I think what you are probably most interested in finding out is who calls the most balls. If so, a simple ranked list would suffice.
January 17th, 2007 at 4:37 pm
There are problems with this graph, certainly. For one, it’s hard to see the scale. There’s no option for standard deviation to be displayed.
However, a ranked list of “who calls the most balls” wouldn’t work, since you have to compare the number of ball called to the total pitched thrown, and the league average, to get an idea of who has a smaller strike zone. If you look at some of the other work I’ve done under the “UMPIRES” tag, I think you might see what I’m getting at.