Quote:
Originally Posted by NoOne
back on topic:
averages are a start... but distribution and volatility are required to give a real comparison... and definitely not based on small samples, like the ones above.
batted ball stats are a %$#@. good luck with that stuff. one website i go to only has info for the last 10-12 years or so. then, you have the problem of conflicting definitions of what a line drive / fly ball is, etc... at this moment in time the data is a bit murky and difficult to use in a precise way. i.e. you can get 2 data sets from the same period of time in the MLB and get 2 completely different results. so, what kind of integrity is there at the moment? not much.
|
I understand your point about the variable definitions for fly balls, ground balls and line drives. That is the reason that I included data on Ground ball outs and Air ball outs. Those definitions are pretty concrete.
With regard to sample size, I would agree that the data from 2014, 2015 and 2016 (YTD) doesn't provide a huge sample size (certainly not enough to provide the kind of statisical analysis I'd prefer that would include distribution and standard deviation analysis), but the consistency of the data between the 3 sets of data is significant enough to support the idea that there shouldn't be as significant a deviation from prior years data as what is show by the OP's data. I gradual change indicating a shift in the way the game is played (such as a greater emphasis on hitting the ball in the air), but a rapid switch like that detailed in the OP is most likely the result of an outside influence (in this case a change in coding/setting/etc). If this type of statistical shift in the span of a single season were seen in real life, one would expect some outside influence to be the cause (i.e. lowering the mound, raising the strike zone, etc).