BY WAY OF APOLOGY
In many collections of Science Fiction stories, the author will add a note explaining how he came to write a certain story. It might have been that the landlord had told him the sheriff was coming in two days to kick him out if he didn't pay his rent. Fermov's Last TN, unfortunately, needs more explanation than that. At the 1980 Toronto Open, Rob Morrison showed a group of us his analysis of 27. (or 29.) Re7!! I was impressed. In July 1980 I composed the first draft and made up three diagrams to go with the story. I sent a photocopy, including the diagrams, to Rob Morrison to verify the analysis. Rob later wrote: "When I got around to sitting down and working on the line, I found that I had forgotten most of the ideas involved, and wasn't prepared at the time to reinvest several hours to refigure things out." In late 1981 I ran across the manuscript again and decided I'd better publish the story, no matter whether the analysis was checked or not. Wait much longer and the science would cease to be fiction. Then I got the idea of splitting it into two parts. To save effort I reused the diagrams from the first draft. Surely I could get Rob to verify his analysis under the pressure of somebody else writing the end to the story! Only a few days before our May-June deadline, Rob phoned with great regret to say that, despite hours of study, his move 27. Re7 could be refuted by 27...Bc5! He sent the analysis anyway. I checked it against the position published in CCE #53, p. 38 and found to my horror that he was right. Mikhalchishin was correct in taking the draw. A new ending was composed for the story. As I keyed it in, I noticed that the original manuscript also considered 27... Bc5! A few moments and a cold chill ran up my spine. Aaaaaaaaargh! The diagram printed in CCE #53, p.38, the same one I made up a year and a half ago, has the White K/g1 instead of /h1. No wonder Rob couldn't verify the analysis! No wonder only two people wrote in to suggest a winning line for White. One made a mistake in analysis; the other apparently analysed from the game position - without realizing that the diagram was different. I can only offer an apology. Here's Rob's analysis. With the K/g1, Black can draw after: 29.Re7 Bc5 30.Bxc5 Qc8. I think 29...Qc8 right away even wins for Black. In the game. Black could have avoided the whole analysis with 26... (or 28...) Kf8. Then White has nothing better than 29.Rh6 Kg8. So if you really need the point, don't play this variation, especially if your opponent reads CCE. The best defence we at CCE were able to find was 29.Re7 Bf4 30.Bh7+ Kf8 31.Bg6 Kg8 32.Bxf7+ Bxf7 33.Rh4 Bxg5 34.Rg4 Rc8 35.Rxg5+ Kf8 36.Rg7 Qxe7 37.fxe7+ Kxe7 38.c3. which strikes us as won for White despite the opposite Bs. Last, there is the line that was also suggested by Mike Bateman: 29.Re7 Qc8 30.Rh6+ Kg8 31.Bh7+ Rxh7 32.Rexh7 Bf8 33.f7+ Bxf7 34.Rh8++. For that Mike wins our prize, a small collection of books. Rob Morrison also wins a prize, and a double apology. Imagine, sending him the wrong position to analyse - twice!
Fermov's Last TN - Commentary and Discrepancies
Commentary
I was editor of CFC Bulletin which for various reasons had to become Chess Canada Échecs from issues 34 (1979) through 61 (1983). The early magazine issues were typed with an IBM Selectric, so no digital version ever existed. By the time of issues 53 and 54, I had shifted to digital typesetting using a Quasar Data Products S-100 CP/M computer with 64K RAM and two DSDD 8" floppy drives, outputting to a daisy wheel printer. This article was left over and it was time to get it in print!
Chess Discrepancies
Diagrams
Diagram 1 is correct
Diagram 2 and Diagram 3, as noted late in the text, are incorrect. The Kings at g1 should be on h1.
Diagram 3 may be incorrect on another count, the positioning of Black's bishops. The game score shows 25...Be8, as appears in Chess Informant 27, page 150, game 420. That was Rob Morrison's source for the game. However, two other sources give the move as 25...Bc5. Chess Assistant's Hugebase could easily be dismissed, as databases are more likely to contain errors than the original sources. And what is the original source? In Chess Informant, the game is listed as being played in SSSR, 1978, but Chess Assistant gives more precision: Daugavpils, Soviet Championship, round 1. Looking up the Soviet Championships in the Soviet Chess Encyclopedia, we find that 1978 was the year of the 46th USSR Championship, for which Daugavpils, the capital of Lithuania, was a 64-player Otborochny or qualification tournament most famous for the ascension of 15-year-old Garry Kasparov to the Soviet Finals (Higher League) for the first time. His score of 9-4 was matched by Uzbekistan champion Igor Ivanov, who was relegated to the Soviet First League by virtue of Kasparov's better Buchholz tiebreak. Ivanov later defected to Canada. The Soviet Encyclopedia conveniently gives the dates of Daugavpils as 27.06 to 16.07, and looking in the appropriate issues of "64" (never throw anything away) we find in issue 28, pages 8-9, the exact game, with brief notes by Gufeld and Nesis, and the move in question being 25...Bc5. So it's two-to-one against the Informant version being correct. Well, in handwriting, e can look like c, and an 8 is like a 5 with an extra stroke. So my theory is that White wrote out his annotations for Chess Informant longhand. Democracy has no place in chess, what are the objective values of the two moves? According to a few minutes with Fritz 4, both moves should lead to a draw. One way we might find out is by asking the players, who in the FIDE rating list are:
14602385 Mihalcisin, Adrian g SLO 2522 13 18.11.54
14100347 Taborov, Boris UKR 2446 0 05.06.60
Diagram 4 is correct, taking into account the discussion of Black's 25th move in the paragraph above. Its K correctly stands on h1. However, the caption of White to move is trivially incorrect, as Black is in check!
Diagram 1 is correct
Diagram 2 and Diagram 3, as noted late in the text, are incorrect. The Kings at g1 should be on h1.
Diagram 3 may be incorrect on another count, the positioning of Black's bishops. The game score shows 25...Be8, as appears in Chess Informant 27, page 150, game 420. That was Rob Morrison's source for the game. However, two other sources give the move as 25...Bc5. Chess Assistant's Hugebase could easily be dismissed, as databases are more likely to contain errors than the original sources. And what is the original source? In Chess Informant, the game is listed as being played in SSSR, 1978, but Chess Assistant gives more precision: Daugavpils, Soviet Championship, round 1. Looking up the Soviet Championships in the Soviet Chess Encyclopedia, we find that 1978 was the year of the 46th USSR Championship, for which Daugavpils, the capital of Lithuania, was a 64-player Otborochny or qualification tournament most famous for the ascension of 15-year-old Garry Kasparov to the Soviet Finals (Higher League) for the first time. His score of 9-4 was matched by Uzbekistan champion Igor Ivanov, who was relegated to the Soviet First League by virtue of Kasparov's better Buchholz tiebreak. Ivanov later defected to Canada. The Soviet Encyclopedia conveniently gives the dates of Daugavpils as 27.06 to 16.07, and looking in the appropriate issues of "64" (never throw anything away) we find in issue 28, pages 8-9, the exact game, with brief notes by Gufeld and Nesis, and the move in question being 25...Bc5. So it's two-to-one against the Informant version being correct. Well, in handwriting, e can look like c, and an 8 is like a 5 with an extra stroke. So my theory is that White wrote out his annotations for Chess Informant longhand. Democracy has no place in chess, what are the objective values of the two moves? According to a few minutes with Fritz 4, both moves should lead to a draw. One way we might find out is by asking the players, who in the FIDE rating list are:
14602385 Mihalcisin, Adrian g SLO 2522 13 18.11.54
14100347 Taborov, Boris UKR 2446 0 05.06.60
Diagram 4 is correct, taking into account the discussion of Black's 25th move in the paragraph above. Its K correctly stands on h1. However, the caption of White to move is trivially incorrect, as Black is in check!
Analysis
The most important analytical point is that after Rob Morrison's 27. or 29.Re7, Black can force White to go for a draw with the ever-so-calm 27...Qa5! 28.Re3 Bd7 29.Rh6+. The incursion of the Q at e1 keeps White just enough off balance so that he cannot complete the slow manœuvre of a mating attack. Thanks to Fritz 4 for this insight.
The most important analytical point is that after Rob Morrison's 27. or 29.Re7, Black can force White to go for a draw with the ever-so-calm 27...Qa5! 28.Re3 Bd7 29.Rh6+. The incursion of the Q at e1 keeps White just enough off balance so that he cannot complete the slow manœuvre of a mating attack. Thanks to Fritz 4 for this insight.
Author's note, March 2005 This story was published in two installments in Chess Canada Échecs ISSN 0225-7351 1982, #53 and #54. It is presented here complete with errors. I'm writing another page with commentary. I invented Internet Shopping and a whole bunch of things that never happened, but most of the discrepancies to be pointed out are chess-related! The first part, CCE #53, March-April 1982, page 19:
Ekleme: Düzeltilmiş diyagram 2 ve düzeltilmiş diyagram 3 (corrected diagrams 2 and 3, resp.):
