In his fascinating and idiosyncratic book Kings, Commoners and Knaves (Russell Enterprises, 1999), chess historian Edward Winter expresses his concern that "historical ignorance of the openings is rampant, with writers regularly analyzing from scratch positions already meticulously examined in the past." The object of openings research, of course, is to provide a basis for good chess, not to preserve a historical account of chess theory. Much of the theory of the past has been forgotten for the very good reason that it has essentially nothing to contribute to the winning of anyone's games. Notions of how chess games are won have undergone so much change, since openings variations started to be written down, that much old analysis is useful only to illustrate the historical development of chess understanding. Vast and elaborate branches of past theory have withered because they can only be reached by suboptimal play, sometimes by both players, as in 1. e4 e5 2. f4 exf4 3. Nf3 g5 4. Bc4 g4?! 5. Ne5? Qh4+ 6. Kf1 f6?!, a line treated at length in chess books of yore.
But old analysis is not entirely without current relevance, and Winter is correct in his claim that forgotten theory would sometimes benefit modern practioners as much as the latest Informant. This is particularly true in some of White's older systems in the open game (that is, 1 e4 e5, and not to be confused with open positions, which may arise from almost any system). The King's Gambit and the various subsystems of the Italian Game, notably the Evans Gambit, once constituted the bulk of the theory manuals. While chess theory in absolute always expands, the relative importance of the open game has greatly diminished over time, and so has its claim to precious printed space. And even within the open
game, the importance of the systems prevalent in the 19th century has diminished
as the Spanish Game has ascended in prominence. Accordingly, the editors of the manuals have had to prune, prune, and prune again the old theory, and they have not always been successful in preserving its outlines, even if they could discern them in what they inherited from previous editors.
One notable case arose in my own practice when the game Clark-Morss began 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d4 exd4 5. e5 d5 6. Bb5 Ne4 7. Nxd4 Bc5 8. Nxc6!? White's move had been considered highly dubious since before World War I, based upon extensive analysis of 8...Bxf7+ 9. Kf1 Qh4! by a certain Dr. Hermann Kaidanz.
Position after 9...Qh4
But when the players of the 1990s turned their curiosity toward 8. Nxc6, even the skeletal outlines of Kaidanz' analysis had been pruned out of the theory books, and all that remained in ECO was the stub of a single variation that seemed easy to question. Indeed, Dutch IM Rini Kuif published a favorable appraisal of 8. Nxc6 in New in Chess Yearbook 20, questioning 9...Qh4 and suggesting instead 9...bxc6, but with good chances for White.
Fortunately, after some casting about for sources, I discovered that I posessed copies of the relevant pages of the greatest and best, 1916, edition of the Handbuch des Schachspiels, the one edited by Karl Schlecter. There in glorious detail were extensive lines that Schlecter characterized as the "main" ones of Kaidanz' analysis. God bless your soul, departed Karl! Ironically, my opponent, a candidate master from Austin, Texas, played a brilliant novelty that Kaidanz had failed to consider. The rest of the story can be read in the game linked at the bottom of this page
My present evaluation of the position in the diagram, after much analysis of my own, is that it is equal. This does not sit well with me, since Black should be better if he is to have to familiarize himself with so many variations as produced by Kaidanz. I share the details in the notes to the game.
I usually offer more than one game here, but the reader will see that the game in question is annotated in very great depth, and even quotes the entire Handbuch analysis. The reader will also note that I have satisfied my intention, advertized in last month's column, to treat the Modern Two Knights: 4. d4 exd4 5. e5. I will deal further with this subject next month.
I now examine at some length a second example of a lost but nevertheless important variation (for the reader's benefit, I include this analysis, in both PGN and CBH formats, in the same file as the score of the Clark-Morss game).
1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. O-O d6
It is odd that this move is baldly ignored in two recent major treatments of the Italian. In his very interesting book A Startling Chess Opening Repertoire, Cadogan 1998, English IM Chris Baker adopts 4. O-O as his keystone move, yet has nothing whatever to say about 5…d6. Nor do Gufeld and Stetsko mention the slightest thing about 5…d6 in the 4. O-O chapter of their The Giuoco Piano, Henry Holt 1996. Talk about a lost variation!
White can, of course, play 5.d3 or 5. Nc3, transposing into quieter lines of the Italian, but the move given is certainly the most ambitious.
Position after 5...Bg4
Strangely enough, this highly consistent move, developing a piece and battling for the d4 square, is not given at all in ECO-3 or in Nunn's Chess Openings, and is barely treated elsewhere. Yet it was discussed at length in the theoretical literature of the 19th and early 20th centuries, and was contested in high-level games up until World War I.
A. Equally in the spirit of 19th century chess is 6. Qb3 Bxf3! 7. Bxf7+ Kf8 8.gxf3, but this encounters 8...Qf6! (Staunton analyzed 8... Qg5+ 9. Kh1 Qf6 10. Bh5 Bb6 11. Qd1 g6 12. Bg4 h5 13. Bh3 g5 and claimed that Black has adequate compensation for his lost pawn.) 9. Bxg8 Rxg8 10. d3 g5! 11. Qxb7 Re8 12. Qxc6 g4 13. f4 g3 14. Be3 Szekely-Steiner, Temesvar 1912. Black went on to win, but here he missed mate in four commencing with 14... gxh2+ 15. Kh1 Qg6.
B. Most in accord with
modern conceptions of how to play the White side of the Italian (where an early
...Bg4 is generally considered suboptimal) is 6. b4. This move was chosen by
Levitsky in two games of his match against the young Alekhine at St. Petersburg in 1912. 6... Bb6 7. a4 a6 8. d3 Nf6 and now:
Position after 8...Nf6
B1. 9. Be3 d5 10. exd5 Nxd5 11. Qb3 Bxf3 12. gxf3 Nxe3 13. Bxf7+ Kf8 14. fxe3 Qxd3 and White's king was in much more trouble than Black's in the third Levitsky-Alekhine match game.
B2. 9. h3 Bh5 10. Re1 h6 (highly ambitious; simply 10... O-O looks good) 11. Nbd2 g5 12. Nf1 g4 13. hxg4 Nxg4 14. Be3 Nxe3 15. Nxe3 Bxe3 16. fxe3 Ne7 (16... Rg8 looks good for Black) 17. Rf1 c6 18. Qe1 Bxf3 19. Rxf3 d5 20. exd5 cxd5 21. Bb3 and White had pressure down the f-file, which he converted to a win in the seventh Levitsky-Alekine match game.
B3. 9. a5 Ba7 10. Be3 O-O 11. Bxa7 Nxa7 12. Nbd2 was played in one of those exceedingly numerous games between German candidate masters that are found in ChessBase's "Megabase," and Black got into serious trouble after 12...d5? 13. exd5 Nxd5 14. Qb3. But he should have preferred 12...Qe7 13. Qe2 (13. Qc2 Nc6 14. h3 Bd7 is also even) 13... Nh5 14. Qe3 Nf4 15. d4 Nc6 16. h3 Bd7 with a perfectly adequate game.
6... exd4 7. Qb3
Unpleasant for White is 7. b4 Bb3 8. Qb3 Bxf3 9. Bxf7+ Kf8 10. gxf3 Ne5 11. Bxg8 Nxf3+ 12. Kg2 Nh4+ 13. Kh1 Rxg8, as given in the 1916 Handbuch.
To deal with alternatives at this point, it is necessary to digress at length from the main variation:
A. 7... Bh5 8. cxd4! (8. Qxb7 Nge7 9. Bd5 O-O 10. Bxc6 Rb8 11. Qa6 dxc3 12. b4 Bxb4 13. Ba4 Bxf3 14. gxf3 Ng6 and Black's kingside counterplay was sufficient for the draw in Brieger-Martin, Fort Worth 1951) 8... Bb6 (8... Bxd4 9. Qxb7 Nge7 10. Bb5 with a big advantage for White) 9. d5 Na5 10. Qa4+ Qd7 11. Qxd7+ Kxd7 12. Bd3 and the endgame is favorable for White.
B. The Berlin master Caro upheld 7... Qd7 8. Bxf7+ Qxf7 9. Qxb7 Kd7 10. Qxa8 Bxf3 11. gxf3, where Black's ambition is perpetual check:
Position after 11. gxf3
B1. 11... Ne5 was treated by Keres in Theorie der Schacheroffnungen, Sportverlag, Berlin 1952 (a translation of Theoria Shakmatny Debiutov).
B1a. 12. Nd2
Nxf3+ 13. Nxf3 Qxf3 14. Qd5! with a draw.
B1b. But it seems to me that White can get a significant advantage with 12. cxd4! and further:
B1b-1. 12...Qxf3 13. Nd2 Qg4+ 14. Kh1 Nf3 15. e5
B1b-2. 12... Nxf3+ 13. Kh1 Bxd4 14. Qd5
B1b-3. 12...Bxd4 13. Qd5! (13. Rd1 Nxf3+ 14. Kh1 Qh5 15. Bf4 Nxh2 and 13. f4 Nf3+ 14. Kg2 Qg6+ 15. Kxf3 Nf6 are extremely good for Black) 13... Qh5 14. Nd2 Nxf3+ 15.Kg2 Qg4+ 16. Kh1 Nf6 17. Qb3 and Black's kingside counterplay appears to be at an end.
B2. Caro himself continued 11...dxc3 12. bxc3?! (Keres pointed out 12. Nxc3! Qxf3, with a draw) 12…Qxf3 13. Nd2 Qh3 14. Qf8 and now:
B2a. 14. e5 Nxe5 (14... Nh6? 15. e6+!) 15. Qg2 Qxc3 16. Rb1 Ne7 is unclear, but Black seems to stand well enough.
B2b. 14. Qf8 Nf6!
B2b-1. 15. Qxg7+ Ne7 16. Qxh8 Ng4 17. Nf3 Qxf3 and Black's attack is too strong.
B2b-2. 15. Qxh8! Ng4 16. Qxg7+ Ne7 17. Nf3 Qxf3 18. Be3 Nxe3 19. Qg3 Qxg3+ (Caro claimed a win for Black after 19... Qxe4 but I see no compensation for the lost material after 20. Rae1 N7f5 [20... Nxf1? 21. Qh3+ Qf5 22. Rxe7+] 21. Qh3) 20. hxg3 Nxf1 21. Rxf1 Ke6. The complications are over, and it looks a little better for Black, with his minor pieces.
This completes our digression; we now return to the main variation after 5...Bg4 6. d4 exd4 7. Qb3 Bxf3!
8. Bxf7+ Kf8
Position after 8...Kf8
The critical position for the entire variation. White has two alternatives:
A. 9. Bxg8 can hardly be a good idea, trading an active piece for an inactive one, yet it has been chosen frequently. 9... Rxg8 10. gxf3 and now:
Position after 10. gxf3
A1. 10... Ne5? 11. f4 Ng4 12. cxd4 Bxd4 13. Qg3 and Black's counterplay was over in Poppert-Staunton, London 1842.
A2. 10... Qd7 11. Kh1 Re8 12. Rg1 dxc3 13. Rg4 Ne5 14. Rf4+ Ke7 15. Nxc3 c6 with substantial advantage to Black in Marshall-Janowski, 1908.
A3. 10...g5! and further:
A3a. 11. Kh1 Qf6 12. f4 gxf4 13. Qxb7 Qg5 14. Qxa8+ Nd8 15. e5 c6 and Black wins. Handbuch des Schachspiels, 1916, Karl Schlecter ed.
A3b. 11. Qe6 Rg6 12. Qh3 (12. Qf5+ Kg7 also favors Black) 12... Kg7 and Black was better in Naedlitz-Goerlitz, 1870.
A3c. 11. Qd1 Qd7 12. b4 Bb6 13. Bb2 d3 14. Qxd3 Ne5 15. Qe2 Qh3 16. Nd2 g4 with a winning game for Black in Kolisch-Anderssen, Paris 1860
A3d. 11. Qxb7 Ne5 12. cxd4 was played in the 1986 postal game Kasturin-Saracino, and Black should have continued 12... Nxf3+ 13. Kg2 Nxd4 with much better piece activity and excellent attacking chances.
B. 9. gxf3
B1. 9... Ne5 10. cxd4 Nxf7 (10... Bxd4 11. Bxg8 Rxg8 12. f4 also wins for White) 11. dxc5 with a winning advantage for White. Maroczy.
B2. 9... g5 10. Bh5 Ne5 11. f4 gxf4 12. Bxf4 Nf6 13. Bh6+ Ke7 14. Kh1 Nxh5 15. Bg5+ Nf6 16. f4 Nf7 17. e5 dxe5 18. Bxf6+ Kxf6 19. fxe5+ Kxe5 20. cxd4+ Qxd4 21. Nc3 and Black resigned in Charousek-Maroczy, Budapest 1895.
B3. 9...Nf6! The Handbuch attributes this to Cordel in his work, Theorie und Praxis.
Position after 9...Nf6
B3a. 10. Bg5 Na5 11. Qe6 Qe7 with a slight advantage for Black based on his better pawns
B3b. The same is true after 10. Bd2 Na5 11. Qe6 Qe7.
B3c. 10. Be6 dxc3 11. bxc3 was played in an e-mail game Sundin-Cherner, 1996. Correct now was 11... Na5 12. Qc2 (12. Qa4 Qe8 13. Qxa5 b6 14. Qa6 Qxe6) 12... Qe8 and Black again consolidates and retains the better pawn structure
B3d. 10. Bf4 dxc3 11. bxc3 (11. Nxc3? Nd4) 11... Na5 12. Qe6 Qe7. And here also, White's two bishops fail to attone for his inferior pawn structure.
I thus complete my treatment of this lost variation.
I believe that if works of chess theory come to be recorded in magnetic form instead of printed form, just as game collections are now, the entire extent of past and present theory will come to be readily available to everyone. Until then, players contesting either side of such openings as the Evans and King's Gambits, and the Two Knights, will be wise
to attempt to recover old theory. They could do much worse than obtain a copy of the Olms reprint of Schlecter's monumental 1916 Handbuch.
Game: Clark-Morss, USCF-92CM76
Download game and main text analysis
for a zipped file (with commentary) in new ChessBase (CBH) and
Next Month: More about the Modern Two Knights (4. d4 exd4 5.