American League teams scored 4.7 runs per game and National League teams are scoring 4.4 runs per game during the first month of the 2007 season. Those averages are down from last year’s averages of 5.0 and 4.8 runs per game respectively. And despite Alex Rodriguez‘s record-tying 14 home run effort, home runs were relatively rare during the first few weeks of the 2007 season. American League lineups hit 1.12 home runs per game in 2006, but launched less than one last month. National League teams were even less powerful, hitting 0.86 home runs per game in April after hitting 1.10 last year. Why was run scoring down in April? And why do lineups seem to sputter during the early part of every season?
My previous research suggests cool air temperatures have a negative relationship with important batted ball outcomes such as hits and home runs. This is probably caused by dense air slowing down airborne batted balls. Is the cause of the last month’s low-scoring games? First, let’s take a look at the average game-time temperature in open air stadiums during the first month of the season over the past three years:
Week Temperature (F) 2005 2006 2007 Week 1 65.0 57.9 52.2 Week 2 63.1 64.6 54.0 Week 3 65.6 64.6 59.9 Week 4 56.4 61.4 64.3
The average temperature for major league baseball games, 58.2 degrees Farenheit, was over four degrees cooler than the average during the early part of the previous two seasons. This relationship seems relevant to understanding why home runs, and consequently run totals, are down this season. However, an alternate explanation for the decline in run scoring could contend that pitchers are “ahead” of hitters at the beginning of the season. This bit of folk wisdom is supported by pitchers’ earlier arrival at spring training and the belief that many hitters need more game experience to refine their timing.
These competing explanations can be tested fairly easily. To study the factors influencing whether or not a batted ball becomes a home run, I submitted 60,000 plate appearances occurring during the first two months of the 2005 and 2006 seasons to a binary logistic regression model that accounted for game-time temperature and also the day of the season. If the time of year is a significant predictor of home run rates after accounting for temperatures, then there might be something to hitters or pitchers improving with experience. Perhaps hitters getting their timing back would allow them to launch more fly balls that turn into home runs. Or perhaps pitchers will get more comfortable and make fewer mistake pitches with more work.
The results? Game-time temperature was a significant predictor (at the p<.001 level) of whether or not batted balls left the ballpark, but the day of the season was not statistically significant. A batted ball has a 4.0% chance of leaving the park during a game played in 70 degree conditions, but only a 3.5% chance of becoming a home run in a game played in 50 degree conditions. This relationship exists regardless of whether or not the game is being played during the first week of the season or in the middle of May. In summary, it’s true that hitters gain an advantage in hitting home runs as the season progresses, but this advantage can be explained entirely by accounting for air temperature changes.