As I noted yesterday, something about the movement on Cueto’s pitch plot didn’t look right at all, so I sent out some emails and Mike Fast who has a blog about the Pitch F/X system and also just started writing for The Hardball Times got back to me this morning. Now, here is what I saw when I plotted his pitches yesterday:
If you don’t know why that looks funny, go here to see where pitches generally fall in terms of where they are plotted by pitch type. Now, as you can see, either the camera is off slightly, or Johnny Cueto gets different movement on his pitches than any pitcher in baseball.
After reading what Mike Fast had to say, it would appear that the camera was off ~4 inches in the horizontal plane if we assume that Doug Davis is throwing about the same as he did last year. If we adjust for that, then the plot would look like this:
That plot looks a whole lot better and syncs up quite well with what his plot should look like based on what he throws. By the looks of it, it would appear that Cueto was throwing a 4 seam fastball a lot, a 2 seam fastball once, while mixing in 15-20 sliders and changeups each.
2 responses so far ↓
1 Alan Nathan // Apr 5, 2008 at 3:35 am
Several comments:
The column sv_id is a data and time stamp, with the least significant digits being the time in seconds.
One needs to be very careful about making arbitrary corrections to the PITCHf/x data. For example, if the horizontal break is off by 4 inches, it does not follow that “the camera was off ~4 inches in the horizontal plane.” Life is almost surely more complicated than that. I generally calculate my own pfx_x and pfx_z values from the 9 parameter fit to the trajectory. I would try that first before trying to make some arbitrary correction. If the camera calibration is off, then it will affect much more than just the pfx_x values.
2 Doug Gray // Apr 5, 2008 at 5:49 am
Alan,
Right now in my data I am just marking it as ‘odd’ but not really adjusting it because I can honestly say up front that I don’t know enough to do it. Basically though I just took the simple correction that Mike suggested in the email and plotted it accordingly, just to see. Once that was made though, the movement on everyones pitches seemed to match up for their stuff last year. Obviously as we gather more data we will begin to see if the camera is off at GABP, but so far it seems like it is. Pitchers are either getting different movement than last year and are throwing slower…. or the camera isn’t working correctly at GABP.
Leave a Comment