I’ll be playing again in the US Chess League this fall, and partly because I want to avoid the debacle that was my play last year, and partly because I have some time now, I cracked open a chess book again. (It’s really mostly because I want to try and do better, but having time also helps.)
I had bought How Karpov Wins a few years back from Moe’s Books in Berkeley for $6. Towards the end of my time on the fellowship, I began studying the games in this book, but I didn’t make it too far.
Now, about 3 years since I last seriously studied, I decided to try again with the book. The games aren’t annotated too deeply, eschewing most variations for pretty general remarks, many of which add nothing in my view. As an example, here’s a game between Estevez and Karpov from 1972 (full game is here, if you want):
White didn’t do too much in the opening and Mednis lets the position go by without any remarks. But playing through the game, I was wondering, can’t Black just play 13…d5 here? I thought maybe there’s something between the Rc1 and Qc7, but there are simply too many pieces in the way. The computer agrees as well, giving …d5 as objectively best with clear equality.
Instead, Karpov played 13…Bd7 and maneuvered behind his pawns for a while. It’s not hard to note …d5 as an option, and Mednis could even have worked it into his pre-built narrative because Karpov avoided straight equalizers on a couple occasions to keep more pieces on the board. Objectively, that shouldn’t have worked out as by move 25 or so, he was much worse, but his weaker opponent didn’t quite handle the position properly and he escaped to rough equality.