Breaking the Berlin Wall

Judit Polgar's defeat of Sergey Karjakin nearly one year ago from the drawish Berlin Wall excites me. I was a fan of Vladimir Kramnik's successful use of this defense in his title match with Garry Kasparov in 2000, when Kasparov failed four times to defeat it. I am a fan of games decided in the endgame, especially when grandmaster games are decided by a superior pawn endgame. In this case, especially, I am pleased to see Judit winning with a final move that disrupts the structure of Karjakin's pawns.

Black to move

White's king will gobble up all the black pawns, leaving White at least two pawns ahead. Judit's older sister, Susan, mentioned in Breaking Through: How the Polgar Sisters Changed the Game of Chess (2005) that "pawn wars" was part of the girls' early training. It seems reasonable to believe that Judit's early training, too, included mastery of key pawn endgames. Such training would have made clear her plan in this game to push her passed a-pawn down to draw the black king to the queenside, away from the decisive kingside pawn endgame.

After my initial enthusiasm for the endgame waned, I thought to investigate the opening through some database research. Given the drawing reputation of the Berlin Wall, how often has a player rated 2700 or higher lost since the World Championship match between Kasparov and Kramnik?

I have downloaded the latest updates for Chess Base Big Datatbase 2011, so my database is reasonably complete. A search reveals that when Black is 2700+, there have been 200 games in the years 2000-2011. 42 White wins, 37 Black wins, 121 draws. Black does quite well. All of the White wins are decided through endgame technique. Some of these games conclude with textbook examples that are worthy studies for the improving class player, such as Michael Adam's demonstration of how to create a Lucena Position.