Today, everybody just "knows" nobody needs chess books any more, and all that matters are databases -- especially so as to keep up-to-date with the latest opening novelties. But let us consider what Amos Burn had to offer us -- way back in 1901.
Burn was, in his peak (ca. 1890) almost as strong, based on his international tournament record, as Steinitz was at the time, while remaining an amateur. He was also British Amateur Champion, won many strong national tournaments, etc. He was also a longtime editor of chess columns. In the Liverpool Weekly Mercury, 17 Aug. 1901, he offered a correspondence game between the Liverpool Chess Club (where he was one of the consultants) and the Edinburgh Chess Club.
The game reached the following position:
Burn notes that, after the last move, 53. Bf4!, the Liverpool CC announced mate in 45. The Edinburgh CC, in reply, accepted their analysis and resigned the game!
Why? Burn explains in his long column (I am paraphrasing):
- First, in general, White, due to his advanced passed pawn, should wins once he forces the exchange of bishops.
- But things are not that simple. If Black's king were on f7, Black could accept the exchange of bishops since then he could keep the king on d8 or e8, ready to move to d7 the moment the White's king moved to e5 (e5 and d7 are corresponding squares), putting White in a drawing zugzwang. As it is, after the forced 53. ... Bxf4 54. Kxf4, if Black plays 54. ... Kd7? 55. Ke5! and it is Black who is in a losing zugzwang.
- So 54. ... Kd6 is forced, and White then wins by moving his king to the queen side and queening a pawn there, as Black is too late to stop him after he takes the f-pawn.
- But why is the win so long? Isn't an outside (and far-advanced) passed pawn usually quickly decisive? Because, Burn's analysis makes clear, nevertheless Black can take the f-pawn (which is relatively close to the d-file) and just manage to queen as White queens, leaving the two sides with an interesting an complicated queen's ending!
55. Ke4 Ke6 56. Kd4 c5+ ('If 56. ... Kd6 57 c5+ wins' -- Burn.) 57. Kc3 Kd7 58. Kb3 Kd6 59. Ka4 Ke6 60. Kd7 61. f7 (at the right time) Ke7 62. Kc6 Kxf7 63. Kb7 b5 64. cxb5 c4 (nothing else reaches a queen's ending at all) 65. b6 c3 66. bxa7 c2 67. a8=Q c1=Q.
Now, adds Burn, this queen's ending is (due to the extra advanced pawn) won for White. But it isn't simple! He notes: 'It would have been an almost endless task to... arrive at the solution by mere plodding analysis'. Instead he offers the following strategic analysis (I shortened it) of various winning plans.
- Speaking generally, the a-pawn cannot be queened unless Black's king is on the second rank, otherwise the white king will be trapped in the corner.
- If White's king first captures Black' pawn, he can then (both his pawns defended by the queen) take refuge in a8, advance the pawn to a7, and with his queen at b7 advance the g-pawn, threatening mate or exchange of queens. Then Black's queen must capture the g-pawn, and white checks at c8 or moves his king to b8, queening the a-pawn.
- Alternatively, if the black king moves to h7 (the second rank) the a pawn can queen.
- Or else, in many cases, the a-pawn can be abandoned, so long as the king is close enough to the g-pawn to queen it.
- In all these cases White must avoid perpetual check, and to do this the black pawn must be made liable to capture by the White king.
- A game where a bishop's ending turns from a drawn to a lost pawn's ending due to one tempo;
- a pawn's ending that turns into a complex queen's ending;
- a deep analysis of the ending's various winning plans and pitfalls;
- all that leading to an announcement of mate in 45 from the original bishops' ending.
But amateurs think that they will get more out of chess by checking their constantly-updated database for opening novelties they will never play and won't know how to exploit if they did play.
They must be joking.
(Source of Burn's column: Richard Forster's superb Amos Burn: A Chess Biography, pp. 498-499).