O-O-O-Phobia?

In my last entry I mentioned the chess party I’m going to this afternoon, which I am supposed to take two games to (my best loss, and a game where I carried one plan to its conclusion). I noticed that both games had something in common — I did not castle queenside when (at least arguably) I should have. This led me to wonder whether I have O-O-O-phobia, the condition of being reluctant to castle queenside.

So I got out my chess notebooks and made a table, counting the number of games I castled kingside, queenside, and neither. Just for fun, I did the same thing for my opponents. Here are the results.

For comparison, here are the figures for my opponents over the same period.

So in recent years, my opponents have been much more reluctant to castle queenside than I have! From 2004 to 2009, I castled kingside in 94 games, queenside in 31, and didn’t castle at all in 33. For all practical purposes, you can say I castle kingside 60 percent of the time, queenside 20 percent, and don’t castle 20 percent of the time. Meanwhile, my opponents castled kingside 113 times, queenside 13 (!) and neither 32. Percentagewise, that’s 72 percent, 8 percent, and 20 percent.

If you go back before 2004, the figures are quite different. 2003 was a bizarre year, when only half of my opponents castled kingside and nearly a third castled queenside. Before that, in 2002, I had a very kingside-oriented year, opting for that direction in more than 80 percent of my games. And in 2000 and 2001, there was basically no difference between me and my opponents.

All in all, it’s hard to know what to make of this. If I ever had O-O-O-phobia, it was in 2002 and 2003, but I think there is a lot of random fluctuation. I obviously didn’t do anything special in 2003 to provoke so many opponents to castle long; it was just the luck of the draw.

Certainly these frequencies are heavily affected by one’s choice of openings. For example, a Sicilian player, especially someone who plays the Dragon, will have a lot of opponents who castle queenside. A player who regularly opens 1. d4 will probably not castle queenside very often, because there are almost no main-line 1. d4 openings where White normally castles that way. (This makes sense, because White usually plays d4 and c4, so he would need to move his king all the way to b1, and possibly even to a1, to be as safe as he would be on the kingside.) So it’s worth noticing that since 2000 I have been exclusively a 1. e4 player, which has probably increased my frequency of queenside castling compared to my opponents. This year I’m going back to a mix of 1. e4 and 1. d4, so my queenside castling percentage should go down. I also think that there may be a long-term trend to more people playing 1. d4, at least in the upper rating levels here in California.

Out of curiosity, I tried to see if I could find statistics like the ones above on the Internet. It’s easy to find percentages of games that begin 1. e4, 1. d4, etc., along with win percentages. But I couldn’t find any data on percentages of queenside versus kingside castling. Does anyone know a site with that information?

During my Google search, I not surprisingly stumbled upon a very nice chess blog: Castling Queen Side, by Polly Wright. The title refers to the blog’s theme, the chess experiences of a female player. Not really relevant to what I was searching for, but I recommend the site, and I will put it on my blogroll next time I get around to updating it.