MLB Gameday BIP Location

Status: Finished. Some UI fixes/polish are still needed, but it is fully functional right now.

This project was inspired by something Tom Tango said on his Inside the Book blog. He offhandedly mentioned being able to superimpose the balls in play from one park onto another park. This does absolutely zero atmosphere or park normalization. I would like to be able to normalize for fence differences, but this would require HitF/X and quite a bit of math to render.

Note: A browser supporting the canvas element is required. As of this writing, this is Chrome 2+, Firefox 3+, Opera 9+, and Safari 4+.

Pitcher Cards

Status: Beta. Fetching and generation of cards is complete, but no normalization is done.

After Josh Kalk was hired, he took down all of his pitcher cards. Although other people have since offered their own PitchF/X graphs, everything has simply used the raw data. This is unfortunate because, while the accuracy of the data has improved over 2008 and especially 2007, there still exists pitches that are wrong. For example, a left-handed pitcher throwing a pitch right-handed or release points well above the pitcher's height. Or, MLBAM using a y0 different than the standard 50 feet for several games. Also, because MLBAM needs to do realtime pitch classification, the algorithm lacks the retrospective accuracy that others have.

Because of this, something that actually corrects the data and displays it graphically would be useful. However, there would still be the problem of getting hired by a team and being required to take down the cards. So, I intend to open source the entire program to download the data, correct it, and graph it.