Elsewhere, I’ve looked at batted ball data to see if we could quantify a Marine Air effect at Safeco Field. There didn’t seem to enough evidence to posit that average battled ball distance was really that different in Seattle. With this evidence, I was interested in whether or not we can see a difference in the batted ball rates at the two most extreme offensive parks in baseball, Coors Field and the Ballpark in Arlington. So first I looked at the average batted ball distances on non-bunts according to Baseball Heat Maps for players that have played on both the Rockies and another team since 2007 (the start of the Pitch F/X era). I didn’t use a hard cutoff, but I usually didn’t include hitters that had less than 200 plate appearances as both a Rockie and a non-Rockie, and usually used about 50 innings as a minimum for pitchers. Then, I did the same for Rangers players.
|Player||As a Rockie||Before Rockie||After Rockie|
Hitters hit the ball a few feet farther when they played for the Rockies than when they played with other teams. The difference between pitchers was even more dramatic, as pitchers were much worse (when it came to batted ball distance) when they played for the Rockies.
Just as a case study, I looked at Dexter Fowler’s 2012 and broke down his batted ball distance by series, and denoted whether or not the series was at home or the road. The method is a little crude, but it should be helpful. Fowler is interesting because while others teams have expressed interest in him, he has extremely large home/road splits. This exercise may help us see if the ball is travelling farther at home, and thus inflating up his numbers.
Fowler’s average batted ball distance is 258.96 feet at home, and 248.45 feet on the road. Of course, positive home/road splits can be explained by the hitter just feeling more comfortable at home (that is, philosophically, he could just be hitting the ball a little harder at home), but with the Coors effect seemingly real, it is hard to not take note of this and speculate that Fowler would be a much weaker hitter if he didn’t get to play 81 games at Coors field (when you compare him with other hitters, you notice that 248 feet per batted ball is not very good).
|Name||As a Ranger||Before Texas||After Texas|
The Case Study for the Ballpark in Arlington I used was Derek Holland’s 2012. Holland has extremely large home/road splits, and is much worse at home than he is on the road.
|Date||Park||FIP||xFIP||Average Batted Ball Distance|
We see that Holland’s average batted ball distance at home in Texas is 265 feet, and on the road it is 255.85, which is about as dramatic as we saw with Dexter Fowler. This explains why his FIP (which weights home runs) has a bigger difference between home and road than his xFIP (which does not weight home runs literally, but only on fly-ball rates). It seems that this evidence makes it clear that while dimensions may play a role in Texas, the “jet stream affect” is real, and does help hitters. However, when we looked at all the players without breaking it down into Home/Road splits, we didn’t see the effect we expected. This seems to be because of the success of Joe Nathan, Koji Uehara, and Mike Adams with Texas. When we look at hitters, we see more of what we would expect, with a relatively (2 to 6 feet per batted ball) large difference in the average amount of feet per batted ball.
So we kind of saw what I expected. Coors Field seemed to show a legitimate difference in batted ball difference, something that we could probably attribute to the altitude allowing the ball to fly more (just like football kickers are usually better in Denver). The difference was real, especially when we broke down Fowler’s season. Derek Holland’s difference was just as real, but the overall data was less conclusive than Colorado’s. Since Denver is a more extreme environment, this was to be expected.