Tuesday, June 9, 2009

As the thunder crashes down

Half of the car alarms on my street are going off in response to lightning striking the trees near my apartment building, and at this point sleep isn't a solid option. So, I'll take my first shot at the easiest of the Hit f/x analysis. Using just the sortable stats available on fangraphs.com for teams batted ball data, running correlations against the April average speed off bat provided by Matt Carruth, here are the correlations for your reading pleasure.

(For the non pocket protector carrying members of our audience--correlations range between 1 and -1, with 1 meaning that they are 100% positively correlated, -1 meaning they are 100% negatively correlated, and 0 meaning they are not correlated at all.)

BABIP: 0.26
GB/FB: -0.56
LD%: 0.07
GB%: -0.53
FB%: 0.54
IFFB%: -0.29
HR/FB: 0.60
IFH: 0.22
IFH%: 0.29
BUH: -0.52

(BABIP: Batting average on balls in play; GB/FB: Ground balls per fly ball; LD%: Line drive percentage; GB%: Ground ball percentage; FB%: Fly ball percentage; IFFB: Infield fly ball percentage; HR/FB: Homeruns per fly ball; IFH: Infield hits; IFH%: Infield hit percentage; BUH: Bunt Hits)

So, what can we take from this? First off, and very surprisingly, LD% is almost entirely uncorrelated to speed of the ball off the bat. This is somewhat counterintuitive. For years you've heard analysts and bloggers alike refer to line drives as hard hit balls, not least of all at this blog. This data runs counter to that theory. This bears watching as the above data is for April 2009 only, and we're going to see all of these shifting around over time (but not THAT much).

While LD% is mostly unconnected to off-the-bat speed, fly ball % and ground ball % are clearly very connected--with higher speed off the bat leading to more fly balls, and lower speed off the bat leading to more ground balls--and at the same rate in opposite directions: 0.54 and -0.53 respectively.

In other, unsurprising news, higher speed off the bat correlates very strongly to how often fly balls convert into home runs (0.60). I'll be very interested to see the data for our friend Big Papi, for instance, and also for Mariano Rivera. Ortiz's problem stems from his extraordinarily low HR/FB%, and Mo's troubles have come from precisely the opposite.

Surprisingly, while the correlation is somewhat low, harder hit balls lead to more infield hits. I can't tell if this is because the defensive players are placed farther back against teams that hit the ball harder, or if it's because of the Baltimore Chop phenomenon, where batters slam balls directly down in front of the plate, hoping to get across 1B while the fielders are waiting for the ball to come down. Sure doesn't appear to have anything to say about well placed dribblers, which are the slowest kind of ground balls, right?

And on the BABIP note, while we've been saying for years that the best way to improve BABIP is to hit the ball hard, I guess that was back when we thought that hitting the ball hard meant hitting line drives. This data pretty much deep sixes that theory, right? Given that LD% (on which average BABIP floats around .700) is mostly unconnected to average speed off the bat, neither will BABIP be.

As far as it concerns the Yankees, it seems to me that given the correlations to FB% and HR/FB% shown above, the teams that have the highest speed off the bats will be very worth watching in homer friendly Yankee Stadium.

Oh, and as Matthew suggested, the more bunts, the lower the average speed off the bat, unsurprisingly.

2 comments:

  1. Good stuff. Found this site via the link on Fangraphs and I'll be following via RSS. Hope to se more hit f/x and pitch f/x work from you.

    ReplyDelete
  2. wow this is way too technical for me. but I am impressed.

    ReplyDelete