Bob Timmermann mentioned the other day that "Sunday's Dodgers-Mets game was a dual complete game for Brad Penny and Ramon Martinez" when I took a look at the D-Rays' inability to complete a single game the entire season. That started me thinking about pitchers duels in general and double complete games (or as I prefer to call them, Bi-Complete Games or Bi-CG, for short) specifically.

Given that the number of complete games per season have been dropping steadily for decades, has the Bi-CG disappeared?

Well, investigating that question is complicated. Baseball does not record when both pitchers have thrown a complete game. All right, it does, but as a complete game for each pitcher. There's no way to tell that both accomplished the feat, not unless one looks up every box score and verifies it.

I thought it might be possible to predict the number of Bi-CGs from the statistics we have. I took the total number of complete game divided it by total games and squared the whole lot in an attempt to capture the probability that any given season for the given league would end in both pitches tossing a complete game. Then I took the probability and multiplied it by the number of actual games in the league.

Here are the figures for the last ten years (2005 data through last night):

That's quite a change, from about nine or ten in 1996 to about two last year. Also, this year in the NL, one would expect just one bi-CG but from Bob we already know that there was at least one.

Here are the averages per decade.

That's quite a constant evolution over decades. So, the next time you hear Joe Morgan complaining about how you never see a pitchers duel today, keep in mind that the same could have been said by old timers in the days of the Big Red Machine.