Thursday, April 23, 2009

2009 Offense and Home Runs

One of the biggest stories in the early going this season (for better or worse) has been the number of home runs hit at Yankee Stadium. Brian Cashman commented in an story today that the Yankees are "monitoring" the situation. More interesting, and not really touched upon anywhere that I've seen so far, is this throwaway line:

But Cashman also said home runs are traveling about eight feet farther so far this year compared to last season. "It's possible that the ballpark is a home run-type park," Cashman said in an interview before his talk. "We'll see. The ball is going farther in every park, not just ours."

Is this true? Are home runs really traveling about 2% farther (8 feet/400 feet) than a year ago? I'm going to try and do some research today because this sounds like an excellent project for the blogosphere.

Update: I am NOT seeing the increase that Cashman is talking about. Heading over to Hit Tracker, I downloaded all the home runs hit in 2009 (through April 20) and all the home runs hits from April 6, 2008, to April 20, 2008 (to match the data set with the 2009; I also removed one bad data point). Here is what I found:

2008: 364 home runs, average distance = 396.15 feet, SD = 25.21, 95% confidence interval = 396.15 +/- 2.59 feet.
2009: 414 home runs, average distance = 397.7 feet, SD = 26.71

H0: 2009 HR Distance = 2008 HR Distance
Ha: 2009 HR Distance > 2008 HR Distance

With 95% confidence can we say that 2009 home runs have travelled farther than 2008 home runs? Using a one-sided hypothesis test, we cannot reject the null hypothesis. Our test statistic, 1.18 [(397.7 - 396.15)/(26.71/sqrt(414)], is less than 1.645, the z-value of 95% cumulative probability. Home runs in 2009 cannot be said to have travelled farther than home runs in 2008, unless there is better data (of which I am unaware) than what Hit Tracker can provide. It's possible Cashman and the Yankees have their own numbers, but I find it difficult to believe that they would vary so much from the publicly available data set.

Now, since 188 games have been played in 2009 and 414 home runs have been hit, I am comfortable stating (without running the numbers) that MORE home runs have been hit than in 2008 (205 games, 365 home runs), I am just not comfortable stating that the home runs have travelled farther.

If you have any questions about my methodology or data, I'd be happy to provide answers. It is entirely possible that I am looking at this the wrong way and I would appreciate feedback.

Update II: Inside The Book has a big discussion of this issue; looks like Cashman was quoting a number that is about a week old.

Also, I think there could be one biasing factor in the data: if there are truly more home runs being hit this year, it's also possible that there are more home runs just barely clearing the fences instead of falling as outs on the warning track. Thus, while the averages may not be changing, that number could be masking a significant increase in the distance batted balls are traveling. Inside the Book does a bit of looking at the distance travelled by the Top 100 homers.

No comments: