Temperature and League HR Splits

Recently, there has been some discussion about the disparity in run-scoring between the AL and the NL. For example, here are two articles on the subject by William Burke and Joe Sheehan, and David Pinto. Because of the designated hitter, AL teams tend to score more runs and hit more home runs than NL teams. This year, this has not been the case. In my investigation of the impact of temperature on home-run hitting I looked at the differences in early-season temperatures between the leagues.

Here are April home run and temperature data from 2000-2008. The temperature data exclude indoor games.

	AL			NL	
Year	HR/G	Outdoor Temp	HR/G	Outdoor Temp
2000	2.60	62.03		2.57	63.90
2001	2.32	60.19		2.35	65.48
2002	2.12	60.31		1.74	64.64
2003	2.13	58.35		2.08	62.79
2004	2.20	60.82		2.13	66.40
2005	1.96	61.82		1.85	64.68
2006	2.42	62.21		2.21	66.13
2007	2.04	57.85		1.69	62.49
2008	1.73	61.53		1.83	65.00
00-07	2.22	60.45		2.08	64.56
08 Diff	-0.49	1.09		-0.25	0.43

The data reveal some useful information. AL teams typically play in colder whether than NL teams. But, even though April 2008 was colder in the AL than in the NL, AL baseball was over a full degree warmer than it had been in the previous eight seasons, while NL baseball was under half-a-degree warmer. Home runs in the AL were down by half-a-HR per game (22%) and NL homers were down by a quarter-a-HR per game (12%) in 2008. This is not looking good for the temperature hypothesis.

Let’s take a closer look at the league differences in home runs and temperature by year.

	HR Gap	Temp Gap	
Year	HR/G	Temp	Temp Impact
2000	0.03	1.87	-0.03
2001	-0.04	5.29	-0.08
2002	0.38	4.33	-0.06
2003	0.05	4.43	-0.07
2004	0.06	5.57	-0.08
2005	0.11	2.86	-0.04
2006	0.21	3.92	-0.06
2007	0.36	4.65	-0.07
2008	-0.10	3.46	-0.05

00-07	0.14	4.11	-0.06
08 Diff	-0.24	-0.65	0.01

In Aprils from 2000–2007 AL teams averaged 0.14 more home runs per game than NL teams did, while NL games were played in parks that were 4.11 degrees warmer than AL parks. In 2008 the temperature gap between leagues actually shrank, as AL teams played in warmer conditions than NL teams relative to the past; therefore, the homer gap should have increased rather than decreased and reversed.

The final column of the table above measures the impact of the temperature gap on HR/G differences between leagues, using the 0.015 impact that each degree contribute to home runs that I discussed in my previous post. This captures the impact that temperatures have on narrowing the home-run gap between the leagues. From 2000–2007, higher temperatures in the NL have kept the homer gap 0.05 home runs per game less that it would be under the same temperature conditions. This means that if the AL and NL played in the exact same temperature environments, the homer gap would be 0.05 home runs per game higher than it has been. Based on temperature, the AL homer gap over the NL should have increased by 0.01 home runs per game in 2008. Therefore, it seems that temperature differences do not explain the change in home-run rates between leagues.

Again, I reiterate what I said about the fluctuation of home runs in my previous post. It’s very difficult to pin the decline in home runs this season on something other than random variation. As Bob Nightengale points out in the USA Today, home run rates are not all that different this season than they were a few years ago.

Comments are closed