The increase in games might explain an increase in the absolute number of perfect games, but it does not, by itself, explain the increasing percentage of perfect games.
According to Baseball Reference, there has been a total of 391,362 games played in the Major Leagues since 1876. We know that there has been a total of 19 perfect games, giving us, over the history of the Major Leagues, one PG for every 20,598 games.
But, in the period since the start of 1991, we have 7 PGs. According to my calculations, there have been 44,845 games played since the start of the 1991 season (including post-season games). So, over these last two decades, the rate for PGs is one every 6406 games.
Therefore, in the period 1876-1990, there were 12 PGs in 346,517 games, giving a rate of one every 28,876 games.
Based on all this, we can see that, in the period 1991-2010, the rate of PGs (on a total games per PG basis) is more than 4 times the rate for the period 1876-1990.
So, there’s little doubt that John T’s observation about the rate of PGs increasing is true. The question, though, is whether we can offer any meaningful explanation for this shift. John T suggests steroids, but not only does that fail to take into account the arguments i offered in my previous post, it also neglects the fact that batters have also been taking steroids, and the last 20 years has seen some of the biggest offensive seasons in baseball history.
I’m not really convinced by this argument either.
I guess you could make a case for or against it by looking closely at figures like overall OBP, strikeout rates, BABIP rates, park factors, etc., etc., and seeing how they have changed over the years, but i don’t have the time or the technical skill to factor in all these variables.
While the increase in the number of PGs over the past two decades is noticeable, i’m still not convinced that 7 Perfect Games out of 45,000 (actually, out of 90,000 if we remember that there are two starting pitchers per game) is enough to be statistically significant.