Magnus Metz

Normalizing PITCHfx

In this article the idea of forcing release points to a central point in the animation is realized - to ease the visual comparison of flight paths across pitch types. The idea was implemented by just using average release points; but that can be deceptive, since it the alters the entire path of each pitch.

The (current) proposal is what we will call ‘normalized’ (or regressed towards the mean) PITCHf/x animations. The idea is to get an idea of a ‘typical’ flight path for each pitch type. It is assumed that averaging over the relevant PITCHf/x parameters is the best way to go about this! Anyway, this is a starting point. Note that the same Darvish data frame from my most recent post is used.

library(knitr)
library(pitchRx)
library(animation)
opts_knit$set(animation.fun = hook_r2swf)
dat <- scrape(start = "2013-04-24", end = "2013-04-24")
atbats <- subset(dat$atbat, pitcher_name == "Yu Darvish")
Darvish <- plyr::join(atbats, dat$pitch, by = c("num", "url"), type = "inner")

library(plyr)
library(reshape2)
idx <- c("x0", "y0", "z0", "vx0", "vy0", "vz0", "ax", "ay", "az")
for (i in idx) Darvish[, i] <- as.numeric(Darvish[, i])
agDarvish <- ddply(Darvish, c("pitch_type", "stand"), summarize, x0 = mean(x0), 
  y0 = 55, z0 = mean(z0), vx0 = mean(vx0), vy0 = mean(vy0), vz0 = mean(vz0), 
  ax = mean(ax), ay = mean(ay), az = mean(az))
agDarvish$b_height <- "6-1"  #estimate average batter height (for strikezones)
animateFX(agDarvish, layer = facet_wrap(~stand))

References

Published

16 May 2014

Tags