In Chess, Variability in Play Is Not Dumb Luck

In Chess, Variability in Play Is Not Dumb Luck

We often confuse the ups and downs of human performance in games with luck.

By Christopher Chabris in the Wall Street Journal

In a tense game in last year’s world chess-championship match, Magnus Carlsen made a huge mistake—an error so obvious to grandmasters that as soon as his hand left the piece, he saw it himself and paused nervously before writing down his move. Incredibly, his opponent, Viswanathan Anand, failed to make the winning reply, allowing Mr. Carlsen to escape. It was a rare “double blunder” among the chess elite—but was luck involved?

Games are often categorized as being ruled primarily by skill or by chance. In some children’s favorites, like War or Chutes and Ladders, every move is dictated by the roll of dice or the shuffle of cards, so luck alone determines who wins. In poker, Monopoly and most modern tabletop and computer games, there are both mechanisms for generating random events and opportunities to make meaningful choices, and the outcome depends on both skill and chance.

As a rule, the greater the number of decisions a player must make, the larger the role of skill in a game. But managing randomness demands certain skills too, such as understanding simple probabilities and mastering one’s emotional reaction to bad rolls or draws.

At the other end of the spectrum are games like chess or Go with no randomness, in which “pure skill” is being tested. Suppose Mr. Carlsen were able to see every possibility 100 moves into the future of each game and to evaluate them all objectively. He wouldn’t have blundered against Mr. Anand; indeed, he would easily defeat every single opponent.

But Mr. Carlsen is not that kind of computer: His human brain generates performances that naturally vary around an average level that reflects his true, underlying chess skill. Even the world champion will have better and worse games and stretches of play.

It is this kind of variability—the ups and downs of human performance in single games or even in individual moves—that we often confuse for luck. From Mr. Carlsen’s point of view, it was a stroke of luck that Mr. Anand missed something a grandmaster would normally see, but that doesn’t mean there is luck in the game of chess, any more than there is luck in tying one’s shoes.

In brief encounters, variability will have a greater effect than it will in longer stretches of play, so a single event will be a poorer measure of the competitors’ skill. By contrast, when players or teams compete in a long series of games, as in a chess tournament or a basketball season, it makes sense to see skill as the primary factor in success.

The irony is that we are drawn to the excitement of variability. We may concede, for instance, that the New England Patriots are more likely than any other NFL team to win the big prize at the end of the season, but as fans and spectators, we like it when the outcome of a particular game is in doubt. Superlative skill is marvelous to witness, but it needs the possibility of error and defeat to generate drama.

As students of a game grow more sophisticated, they are able to appreciate ever-higher levels of skill. Chess grandmasters love to see a fighting draw, and knowledgeable soccer fans can revel in a well-played 1–0 game. But for games to appeal to experts and amateurs alike, they need the right tension between skill and built-in randomness.