 "Chess (#1068)")
Most of the money is on Magnus Carlsen in the World Chess Championship in Chennai next month. Apart from being younger and more consistent, the world number 1 also has charisma - that undefinable quality which means he is noticed even by those people who don't play the game.
In many sports, the concept of an upset is in itself subjective, with experts weighing in on who they think is the favourite. Chess predictions however, tend to be extremely number-driven. Almost every prediction system and every underlying metric favours Carlsen.
There have been multiple attempts, including the Elo rating system, to compare playing strengths for players in the same era, players across eras, and players at different times in their own respective careers. These systems look at variables like simple win-loss-draw data and more complex ones like error-free play.
A strong player in a strong era may have less impressive win-loss stats than a strong player in a weaker era. However, ratings are correct for precisely this reason since the average ratings are higher in a strong era. After adjusting for rating inflation, Carlsen has the highest-ever rating and the best stats in an era, which is the strongest-ever.
Computer-based analytical checks for error-free play are deceptive. A player with a complex risk-taking style will have higher error rates than a simpler positional player. Carlsen is among the most error-free players of all time.
Match dynamics can alter normal equations, but looking back historically, there haven't been many major upsets in the 130-year history of the world championship. In the pre-ratings era, the big upsets would include Alexander Alekhine-Jose Capablanca in 1927, Max Euwe- Alekhine in 1935, perhaps Carl Schlechter-Emanuel Lasker in 1900 (match drawn) and Lasker-Wilhelm Steinitz in 1894. After ratings were introduced in 1972, only Vladimir Kramnik-Gary Kasparov in 2000 bucked a major rating difference to win, though of course, lower-rated players have done well quite a few times in matches.
The diagram, BLACK TO PLAY (Adams Vs Carlsen, London 2012) is shown from black's side and shows Carlsen's precision. The normal 53...Ke6? 54.Kf3 Kf6 55.Ke4 Kg5 56.Kxe5 Kxg4 57.Kd5 Kxh5 58.Kxc5 g5 59.Kd5 g4 60.c5 g3 61.c6 g2 62.c7 g1Q 63.c8Q is a draw. Play revolves around mutual zugzwang. If white has Ke3, black has Ke5, black wins only if white is on move.
Black won with 53...e4! 54.Kf4 e3 55.Kf3 The alternative 55.Kxe3 is met by 55...Ke5! 56.Kf3 Kd4 ( or56.Kd3 Kf4-+) 57.Kf4 Kxc4 58.Kf5 Kd5! 59.Kg6 Ke6 60.Kxg7 c4 61.g5 c3 62.g6 c2 63.Kh8 c1Q 64.g7 Qg5 65.g8Q+ Qxg8+ 66.Kxg8 Kf6! Play continued 55...Ke6! 56.Ke2 Kf6! 57.Kf3 Kg5 58.Kxe3 Kxg4 59.Ke4 Kxh5 60.Kd5 g5 61.Kxc5 g4! 62.Kd4 g3 63.Ke3 Kg4 (0-1) The end is 64.c5 Kh3 65.c6 g2 66.c7 g1Q+.
In many sports, the concept of an upset is in itself subjective, with experts weighing in on who they think is the favourite. Chess predictions however, tend to be extremely number-driven. Almost every prediction system and every underlying metric favours Carlsen.
There have been multiple attempts, including the Elo rating system, to compare playing strengths for players in the same era, players across eras, and players at different times in their own respective careers. These systems look at variables like simple win-loss-draw data and more complex ones like error-free play.
A strong player in a strong era may have less impressive win-loss stats than a strong player in a weaker era. However, ratings are correct for precisely this reason since the average ratings are higher in a strong era. After adjusting for rating inflation, Carlsen has the highest-ever rating and the best stats in an era, which is the strongest-ever.
Computer-based analytical checks for error-free play are deceptive. A player with a complex risk-taking style will have higher error rates than a simpler positional player. Carlsen is among the most error-free players of all time.
Match dynamics can alter normal equations, but looking back historically, there haven't been many major upsets in the 130-year history of the world championship. In the pre-ratings era, the big upsets would include Alexander Alekhine-Jose Capablanca in 1927, Max Euwe- Alekhine in 1935, perhaps Carl Schlechter-Emanuel Lasker in 1900 (match drawn) and Lasker-Wilhelm Steinitz in 1894. After ratings were introduced in 1972, only Vladimir Kramnik-Gary Kasparov in 2000 bucked a major rating difference to win, though of course, lower-rated players have done well quite a few times in matches.
The diagram, BLACK TO PLAY (Adams Vs Carlsen, London 2012) is shown from black's side and shows Carlsen's precision. The normal 53...Ke6? 54.Kf3 Kf6 55.Ke4 Kg5 56.Kxe5 Kxg4 57.Kd5 Kxh5 58.Kxc5 g5 59.Kd5 g4 60.c5 g3 61.c6 g2 62.c7 g1Q 63.c8Q is a draw. Play revolves around mutual zugzwang. If white has Ke3, black has Ke5, black wins only if white is on move.
Black won with 53...e4! 54.Kf4 e3 55.Kf3 The alternative 55.Kxe3 is met by 55...Ke5! 56.Kf3 Kd4 ( or56.Kd3 Kf4-+) 57.Kf4 Kxc4 58.Kf5 Kd5! 59.Kg6 Ke6 60.Kxg7 c4 61.g5 c3 62.g6 c2 63.Kh8 c1Q 64.g7 Qg5 65.g8Q+ Qxg8+ 66.Kxg8 Kf6! Play continued 55...Ke6! 56.Ke2 Kf6! 57.Kf3 Kg5 58.Kxe3 Kxg4 59.Ke4 Kxh5 60.Kd5 g5 61.Kxc5 g4! 62.Kd4 g3 63.Ke3 Kg4 (0-1) The end is 64.c5 Kh3 65.c6 g2 66.c7 g1Q+.
Devangshu Datta is an internationally rated chess and correspondence chess player