Find the best move! Once upon a time, I sincerely believed that was my overriding goal during a game of chess. Naive, but nowadays I know better. The truth is that dodging banana skins is more fruitful, so to speak, than the pursuit of golden apples.

In part, this is a simple story about experience and humility. After making enough bad moves, one comes to realise that there are always more lurking around the corner. But really, it’s not about me. The past decade or so has seen a fundamental shift in the way that games of chess are perceived, for which the near-omniscient chess computer has been the driving force.

The moves of the world’s best players used to be held in awe. They made mistakes, of course, but identifying them could take a considerable amount of legwork. These days, a chess engine will take just seconds to point out the errors in a game. For every position, the engine shows its preferred move and the resulting evaluation in hundredths of a pawn (e.g. +1.47 signifies a healthy advantage to White). That acts as a benchmark for all the other moves.

The perceptual shift is significant. In chess analysis, thought used to precede truth, and a move that appeared to exhibit a truth would be awarded a ‘!’ signifying ‘good move’. But with the computer by our side, truth precedes thinking, and our moves merely manifest varying degrees of falsity.

But surely one still just has to find the best move? Well, sort of. If those statistics have taught me anything, it’s that ‘satisficing’ wins chess games. Consistently finding suboptimal but decent moves seems more effective than being a perfectionist who blunders intermittently.

Both the big internet chess servers, chess.com and lichess.org, will produce an ‘accuracy’ measure (out of 100) to show how well a game was played by each player overall. It tends to make me feel more chastened than uplifted, but just once in a while, it offers a pleasant surprise. I was thrilled to win a bracing tactical battle in the German Bundesliga last month, but it’s a habit to check for all the warts after the game. My joy was all the greater when I saw my accuracy rating: 99.7 per cent! A red letter day.

Luke McShane (SV Werder Bremen)-Valentin Baidetskyi (MSA Zugzwang)

Schachbundesliga, Viernheim, February 2024

1 e4 e6 2 d4 d5 3 Nd2 c5 4 Ngf3 cxd4 5 exd5 Qxd5 6 Bc4 Qd6 7 O-O Nf6 8 Re1 Nc6 9 Ne4 Qb4 10 b3 Be7 11 Bb2 b5 White leads in development, and will soon recover the pawn on d4, so castling was safer. 12 Bd3 Bb7 13 a4 bxa4 14 Rxa4 Qb6 15 Nxd4 Rd8 16 Qa1 Nb4 (see diagram) 17 Nf5 The best way forward! The main justification is that 17…exf5 18 Nxf6+ gxf6 19 Rxe7+ Kxe7 20.Rxb4 Qxb4 21 Ba3 wins the queen, with a huge advantage. Or Black can immediately capture on d3 or e4, which are all met by Nxg7+ followed by recaptures. Nxe4 18 Nxg7+ Kf8 19 Bxe4 Rd2 20 Rxb4 My original intention, 20 Bxb7, is met by 20…Qxf2+ 21 Kh1 Nxc2 with an annoying fork. Forcibly removing the knight is worth the sacrifice. 20…Bxb4 21 Bxb7 Rg8 21…Qxb7 22 Ba3! prepares to activate the Qa1. e.g. 22…Bxa3 23 Nxe6+ fxe6 24 Qxh8+ Kf7 25 Qxh7 wins the queen on b7. 22 Bf3 Rxc2 22…Qxf2+ 23 Kh1 Rxc2 24 Rf1! wins 23 Bd4 Qa5 Allowing a neat finish. 24 Bc5+ Qxc5 24…Rxc5 25 Nxe6+ Ke7 26 Nxc5+ Bxe1 27 Qe5+ wins quickly. 25 Nxe6+ fxe6 26 Qf6+ Ke8 27 Qxe6+ Kd8 27…Kf8 28 Qf6 is mate 28 Qxg8+ Black resigns