Posted 04 March 2010 - 08:59 AM
One needs to exercise some caution when using data from this site. For example, when I looked at Varitek's chart I could only find eight of the ten home runs he hit at Fenway Park, so I checked pitcher-by-pitcher and found that two pairs of home runs were superimposed--Shields/Meche and Cecil/Mitre. Knowing this, I could see that two of the dots were slightly elongated. I followed up by checking the home runs on Hit Tracker. Here I found that horizontal angles and distance were close enough for the Cecil/Mitre pair to account for them appearing superimposed but while the Sheilds/Meche pair down the right-field line were close to the same angle, the distances were quite different--73' in standard distance and 57' in true distance. Superimposing this pair of home runs on another park could result in an erroneous picture.
One other thing that should be noted is that balls in park are plotted at the position of the fielder when he gets the ball, not where the ball struck, so you cannot differentiate between balls hitting the wall first and ones that don't reach the wall. At Fenway, this could have a huge impact on evaluating balls hit to left field and left-center. For example, a ball hit off the LCF wall that the LF misses on a jump may end up plotted as a double to medium CF.
I still like the idea and I am going to suggest to the author that expanding the filtering options to include off RHP or LHP might be useful.