Math and Life

A Good Knight for Chess Records

b4 Thursday, I haven’t had much of a reason to be excited about this world championship. But Friday had it’s surprises, and now I’m hooked again. I’ll leave the actual chess commentary to the experts, and take up some of the less pressing analysis.

Friday’s game was crazy. The first 20 ply (10 game moves) featured 15 knight moves, 10 of which were consecutive.

Is that some kind of record?
— Jerry (ChessNetwork)

Let’s find out.

On this website, I downloaded pgn files of all the world championship matches since 1886, consolidated them into one file, and went to work.1

Here are the results:

First, were there any other world championship games with more knight moves in the first 20 ply?

Knight moves in first 20 ply.PNG

Not even close. The previous record was a 10-way tie, with 9 knight moves apice. The only modestly interesting thing I found here is that these two games between Kamsky and Karpov were identical up to the 14th move.2

Ok that was boring. What else can we do with this? Well, we know knights are fun and hop around everywhere. How do other pieces compare?

Moves in first 20 ply.PNG

No real surprises here. Kings move the least in the opening, followed by rooks and queens. The fun comes in when we dig a little deeper into the games.

Those 3 king moves happened in this glorious game. White is checked on h4, moves his king to e2, is checked again, moves to f2, is checked again, moves to g2, then WINS THE GAME 10 moves later. You really should take a look.

Raw number of moves wasn’t the only thing that made Friday’s match special. We also had 10 consecutive knight moves. How do the other pieces compare?

Consecutive moves in first 20 ply.PNG

Rather poorly, it turns out. The only piece that is moderately close are the pawns, and there are 16 of them. If you need a game to look at, I’ll point you towards the 8-pawn mover.

But what if we look at the entire game instead of just the first 10 moves?

Consecutive moves.PNG

Most of these games are draws by repetition. They go something like this:

White: how bout a 3-move repetition?

Black: ehh. I mean maybe

White: what if we do it on different squares?

Black: split the difference. We’ll do it on another square.

White: sounds like a plan

See the 8th game of the Lasker v Schlechter 1910 championship (starting on move 32).

Similarly, here are the maximum number of moves made by each piece over an entire game.

moves total.PNG

Unsurprisingly, this we end up with really long games (red bar). I’m going to recommend is the 45-pawn-mover between Bobby Fischer and Spassky. In the course of the game, 4 (4!!) pawns reach the 7th rank, with plenty of tactics in between. Also highly recommended.

If anyone wants some other oddball world championship stats, comment below. I'm on break and need something to do.

1After about an hour, I figured out that 445th game in this massive text file had an extra line of space between two games, which offset everything after it. This was incredibly frustrating

2I initially thought there were duplicates in the database, so I spent another 20 minutes trying to work that out before realizing they were distinct games.

Noah Caplingerchess