Yesterday, I posted a graph that looked at the trend of percentage of extra inning games per year. Crashburn Alley brought up this thought in the comments: "I wonder if lower scoring games increase the probability of extra inning games." It's a great thought, since during the high-scoring mid 90's, extra inning games decreased. They also appear to be increasing again as scoring begins to decrease. So this is what I found, data come from Retrosheet:
This is that same graph with runs per game included for each season. It definitely looks like there is something to the idea. Here is that same graph with only the five year moving averages for each line (H/T: thehemogoblin)
So then I looked simply at runs per game versus percent of extra inning games per year:
Of course everyone knows that correlation doesn't lead to causation, but there does appear to be some correlation there.