This fascinating story began when a Dr Wolfgang Eisenbeiss and an acquaintance came up with the idea of a chess match played between 2 masters: one alive and the other no longer in this world. Dr Eisenbeiss knew a medium, Robert Rollans, who was trustworthy and knew nothing about chess, so wouldn’t be able to influence the results.
A list of deceased Grandmasters was drawn up and Rollans attempted to contact one and persuade them to play a game. In the meantime, Dr Eisenbeiss searched for a living GM willing to play against a ghost. One man agreed, the great Viktor Korchnoi – who was ranked 3rd in the world. On the 15th June, 1985, the challenge was accepted by a spirit claiming to be Geza Maroczy (3 March 1870 – 29 May 1951). Maroczy came close to playing a match for the World Championship against Lasker in 1906 and is immortalised with the pawn formation known as the Maroczy Bind in the Sicilian (pawns on c4 and e4, controlling d5). At his peak, he had been ranked 3rd in the world.
Maroczy (communicating through Rollans) was given the White pieces. The moves were communicated through automatic writing, which Rollans passed to Eisenbeiss who then forwarded them to Korchnoi. When Korchnoi had decided on his response, he would pass the move back to Eisenbeiss who passed it to Rollans who would write down the move on a scoresheet and make the move on a chessboard. (Dr Eisenbeiss who was an amateur chess player, had trained Rollans in the understanding of notation and the moves of the pieces).
The communication of the moves typically took around 10 days but Korchnoi was often unavailable because of travelling (being a globe-trotting chess professional) and so the game took 7 years and 8 months with Maroczy resigning after 47 moves. In a strange coincidence, Rollans died 3 months later.
When asked about the play of his opponent, Korchnoi said, “During the opening phase, Maroczy showed weaknesses. His play is old-fashioned. But… I am not sure I will win.He has compensated the faults of the opening with a strong endgame. In the endgame, the ability of a player shows up and my opponent plays it very well.” Helmut Metz, a well known chess commentator, observed that Korchnoi’s opponent “controlled the endgame like the old masters from the first half of the century”.
Playing well enough to make Korchnoi unsure if he would win was not easily accomplished. The chess computers of the day were nowhere near as strong as today. To play in an “old-fashioned” style would be even more difficult to achieve.
As well as the game, Maroczy was asked a series of questions about his life and chess career during the 7 year period. These included basic details like where Geza was born, to positions he finished in tournaments to more difficult to discover information such as the chess playing strength of his children and the job he took after finishing school.
Dr Eisenbeiss was given the task of validating the answers and enlisted the help of a chess historian. Laszlo Sebestyen. Sebestyen was not told about the experiment, nor did he meet Rollans or Korchnoi, he believed he was researching for an article on Maroczy. He consulted numerous specialized libraries and interviewed Maroczy’s two surviving children and a cousin and was able to answer all but seven of the questions. Of the answers, only three differed from those obtained from Maroczy via Rollans.
One of the questions concerned a tournament in San Remo, Italy in which Maroczy played a surprising move (41. Qh5!!) that saved a game that had looked lost. The move had thrilled spectators and Dr Eisenbeiss thought it was likely Maroczy would remember this game even though it was played against a relatively unknown Italian player called Romi.
Maroczy’s response was also surprising. He said that he did not know a Romi but had had a friend called Romih (with the ‘h’ at the end) and this is the person he had beaten in San Remo. So Eisenbeiss asked Sebestyen to look into the correct spelling of the name. The historian found a German and a Russian book that spelled the player’s name Romi and a Hungarian one that spelled it Romih. Eisenbeiss looked into it himself and found two more references that spelled it Romi and was about to conclude that Maroczy’s claim was incorrect when he found the official San Remo 1930 Tournament programme where the player’s name was mention several times and always spelled Romih. The reason for the discrepancy was later discovered when a chess expert from Italy recalled that Max Romih was of Slavonic origin and had emigrated to Italy in 1918. He hadn’t dropped the ‘h’ from his name until after the San Remo tournament.