THE BOOK--Playing The Percentages In Baseball

Are line drives based on the angle off the bat mainly?  My feeling, is that any ball that is “barreled” cleanly is a line drive.  Some line drives go for home runs, some a straight liners through the middle (excluding ground balls and clear lofted fly balls).  I would say that speed off the bat would be the best indication of what a line drive is.  Not sure how possible that is without hit f/x data.  Once we got the sample sizes up, we could have a database for each players type of hits in regards to barrel explosion, and probably be able to tell what part of the bat the ball was struck at.  From my playing experience, line drives are cleanly struck balls that jump off the bat...is this how they are classified now?

We had other threads on the matter.  You know where else bias exists?  If the fielder caught the ball or not!  So, if a guy catches alot of line drives, they’ll be marked as regular fly balls.

wow, that’s absolutely ridiculous.  That doesn’t make any sense at all.

You know where else bias exists?  If the fielder caught the ball or not!  So, if a guy catches alot of line drives, they’ll be marked as regular fly balls.

I and others have suggested the possibility that whether the ball is caught or not might be a source of bias but I am not sure that a definitive study has been done to prove this.  I think it is great that Colin has explored another possible source of bias.  But as Colin himself admits it is a difficult area to study and there not even a clear idea whether misclassification of hit balls is really creating a problem with our fielding evaluations or not.  I know that Colin tried to obtain data from STATS and possibly BIS to compare for his study and I assume since it was not included that he went ahead and published what information he had and left that comparison for the future.  But it may be that the total of subjective misclassifications will end end up being relatively small.  And as Dave Smith pointed out in his post at Retrolist the major portion of what mosclassifications that do occur may be due to a genuine difference of opinion between observers rather than actual bias.

For my presentation at the 2009 Pitch f/x Summit on how Hit f/x might improve fielding metrics I originally hypothesized that being able to objectively define a hit ball by its characteristics of speed off the bat and vertical angle might be significant improvement over subjective classifications such as fly and liner.  But I am less sure of that after having looked at the data.  There was less overlap of classifications for similarly hit balls than I would have predicted (http://baseball.sportvision.com/summit/2009) and, as Dave Smith points out in his post fielder positioning is a legitimate explanation of why similarly hit balls can be classified differently.  So is the fielder to which a ball is hit, and also the effect of spin on the hang time of balls that come off the bat with otherwise identical speed and vertical angle.  A stop watch would solve the problem IF we had accurate hit ball distances, but we don’t.  However, I did suggest to Ross Paul at MLBAM that time to distance be included in the overhaul of the stringer data collection procedures that he plans for this off season.  I also suggested some other improvements that would result in more accurate distance information from Gameday.  We will see if they are implemented.  Both time and accurate distance will be included in Field f/x when it is implemented ( now tentatively scheduled for 2011) but when and if independent baseball analysts will have access to that information remains to be seen.

Does this explain Chip Caray’s inability to call a competent game?