Monday, March 9, 2009

Do Some Teams Really Live and Die by the Home Run?

Sometimes you hear it said about teams that they "live and die by the home run." The implication is that teams who rely on power to generate their offense are subject to feast or famine type games - sometimes pounding out tons of runs and otherwise being stymied for very few. You never hear the phrase "live and die by the single", so people tend to particularly associate homers with these kinds of all or nothing performances.

This post intends to determine - is this true, and if so, to what extent? To do this I once again rely on a simulation. Plugging in the statistics for two types of teams and running thousands or games, we can compare the standard deviation of runs for the two teams. If the adage is true, we expect the team that relies on power to have a higher SD - reflecting their feast of famine style offensive games.

In the 2008 National League the Atlanta Braves and Philadelphia Phillies represented those two extremes. Both were good offensive clubs, but the Phillies relied on a league-best 214 HR's and a poor .255 BAV while the Braves played small-ball and hit only 130 HR's but with a .270 BAV.

The simulator, admittedly a bit crude, had the Braves ballplayers hitting homers at the lower rate. The simulator, playing 100,000 games had the Braves scoring 4.67 runs per game (close to their actual average of 4.65 runs per game), but with a standard deviation of 3.16 runs.

How about the power-laden Phillies - did they really "live and die by the homer" any more than the Braves did? The simulator, now running with the Phillies high rate of homers, had the Phillies scoring 4.74 runs per game (slightly lower than their actual 4.93), but with a standard deviation of 3.13 - almost no different (and actually slightly lower) than the Braves SD of 3.16!

The myth appears busted. Teams who rely on power are no more prone to offensive feasts and famines than teams who rely on other facets of the game. In the last post we talked about individual players and their standard deviation of batting runs. We found that individual power hitters had a slightly higher SD than punch and judy hitters. That may be true, but in the context of an entire teams, the difference in variance between power hitting teams and contact hitting teams are so small that the differences are negligible. So the next time you hear an announcer (probably Joe Morgan) spout off that a team is bound to be streaky because they live and die by the home run - take it with a grain of salt.