Thursday, February 25, 2021

The Death of Correspondence Chess (and an Alleged Soviet Mole)

     After returning to correspondence chess 20 years ago after a long hiatus, I quickly discovered that even on sites where engines weren't allowed many players were using them anyway. At first it wasn't much of a problem because with some effort an engine could be beaten. Of course, things eventually reached the point where they couldn't be beaten and I quit playing correspondence chess again.
     Then around 10 years ago I discovered Lechenicher SchachServer, a site where engines are allowed and thought it might be interesting to experiment with different openings and engines. Over 80 percent of my games were drawn. Of course a few games will produce decisive results, but how can you increase your chances of winning? 
     A while back I was following a forum conversation between players who are more successful at correspondence chess than I and found their methods interesting. As one of them observed, there is not a great deal of life left in CC because Stockfish is just too strong and even on a modest laptop it rarely loses even in positions that were once felt to be promising. 
     With decent hardware, diligent time management and careful play a draw seems almost unavoidable. In fact, most losses are due to 1) entering the wrong move (it happens!) or 2) not checking the engine's suggestions properly, and 3) a scant few players aren't using the strongest engines. 
     A fourth factor that results in lost games is playing a bad opening. One fellow wrote that when playing on ICCF (engines allowed) he used Aquarium and created an opening tree with almost 5 million positions!! At the start of a game the opening tree is his main decision maker. When he gets to a position where the tree has fewer than "a few thousand" positions he begins a hunt for the best move using Infinite Analysis; other players prefer to use Aquarium's IDeA. 
     It wasn't always that way. In the old days players could consult books, but the moves were their own. And, in those days top correspondence players were REAL masters and Grandmasters, not an average player with an engine and a penchant for conducting deep research. 
     After C.J.S. Purdy won the first world correspondence championship he vowed never try and repeat the performance because it was simply too time consuming and too much work. 
     On his way to winning that first world championship Purdy lost one game and it was to a British player Graham Mitchell who was a most interesting character. He was a Correspondence IM who served in MI5 (British Intelligence), eventually rising to Deputy Director-General of the service, reporting to a fellow named Sir Roger Hollis who was the father of Adrian Hollis, a top correspondence player. Mitchell and Hollis were involved in an intriguing web of spy stuff that has been of 
interest even to this day.
 
Further Reading: 
Graham Mitchell's story on Spartacus Educational...HERE 
The Spy Who Checkmated Me: Why Postal Chess Was Banned During Wartime...HERE 
 
 

Graham R. Mitchell - C.J.S. Purdy

Result: 1-0

Site: 1st World Correspondence Championship

Date: 1950

Ruy Lopez

[...] 1.e4 e5 2.♘f3 ♘c6 3.♗b5 Even in the days before engines postal chess required careful opening preparation. And, here we are going to see one of the long variations of the Ruy Lopez. 3...a6 4.♗a4 ♘f6 5.O-O ♗e7 6.♖e1 b5 7.♗b3 O-O 8.c3 d6 9.d4 ♗g4 10.♗e3 exd4 11.cxd4 ♘a5 12.♗c2 ♘c4 13.♗c1 c5 14.b3 ♘a5 15.♘bd2 ♘c6 16.h3 ♗h5 17.g4 Theory on this position is starting to get pretty thin and white will play this sooner or later in any case, so Mitchell does it at once. However, the move seems to give black a slight advantage. The move 17.d5 might be worth investigating.
17.♗b2 ♘xd4 18.♗xd4 cxd4 19.g4 ♗g6 20.♘xd4 ♕b6 21.♘f5 Espinosa Aranda,A (2419)-Ibarra Jerez,J (2545)/Madrid 2015. The position is about equal.
17...♗g6
17...♘xg4 This move is also playable. 18.hxg4 ♗xg4 19.d5 After this black gets the advantage. Best was keeping the tension in the center with 19.Bb2. 19...♘d4 20.♗b2 ♗f6 21.♗xd4 ♗xd4 22.♖b1 ♕f6 Vidic,T (2235)-Batchimeg,T (2391)/Pardubice 2016. Black is better and went on to win.
18.♗b2 ♘d7 19.♖c1 ♖e8
19...♖c8 20.d5 ♘cb8 21.e5 ♗xc2 22.♕xc2 ♘xe5 23.♘xe5 dxe5 24.♗xe5 ♗d6 White is better. Janosevic,D (2455)-Balshan,A (2415)/Amsterdam 1978
20.♘f1 White might well have done better to have closed the center with 20.d5 20...♗f6 21.♗b1 ♕b6 After this Purdy finds himself facing problems.
21...♘xd4 22.♘xd4 cxd4 23.♗xd4 h5 would have given him some play on the K-side.
22.dxc5
22.h4 was better as after 22...d5 23.e5
23.dxc5 ♕d8 (23...♘xc5 24.♗xf6 gxf6 25.♕xd5 is winning for white.) 24.♗xf6 ♘xf6 25.exd5 ♖xe1 26.♘xe1 ♘xd5 27.♗xg6 hxg6 white is slightly better.
23...♗xb1 24.exf6 ♗e4 25.dxc5 ♘xc5 White's position is quite promising.
22...dxc5 23.♗xf6 ♘xf6 24.e5 ♖ed8 25.♕e2 ♗xb1 26.exf6 ♗d3 27.♕e3 ♘b4 28.♕xc5 ♕xc5 29.♖xc5 ♘xa2 30.♖e7 ♘b4 31.♘e3 ♖ac8 32.♖xc8 ♖xc8 33.♘d4 gxf6 34.♘df5 ♘c6 35.♖d7 ♗xf5 36.♘xf5 a5 37.♘h6+ ♔f8 38.♖xf7+ ♔e8 39.♖xh7 ♘d4 40.♖a7 Up until this point the game has been even but Purdy's next move is a serious mistake. Surprisingly the correct course is to obtain a passed a-Pawn and not the b-Pawn. 40...♘xb3
40...♘c6 41.♖b7 ♖b8 42.♖xb8+ ♘xb8 43.♘f5 a4 44.bxa4 bxa4 45.♘d6+ ♔e7 46.♘c4 and this ending is drawn. Just one example... 46...♘c6 47.♔g2 ♘e5 48.♘a3 f5 49.f4 ♘c6 50.♔g3 ♔f7 and white can make no progress because his N is tied down blockading the a-Pawn.
41.♘f5 ♖b8 After this black is dead lost.
41...♖c1+ 42.♔g2 ♖d1 43.h4 ♖d7 44.♖a6 ♘d4 45.♘g3 b4 46.h5 and black at least has a fighting chance. Shootouts using Stockfish resulted in white scoring +3 -0 =2.
42.h4 This secures the win as black is unable to successfully deal with the h-Pawn. 42...♘c5 43.h5 ♘d7 44.h6 ♘f8 Losing immediately.
44...♔f7 was also hopeless though. 45.♖xd7+ ♔e6 46.♖e7+ ♔d5 47.♖g7 ♖h8 48.h7 ♔c4 49.♘h4 b4 50.♘g6 ♖xh7 51.♖xh7 b3 52.♘f4 b2 53.♖b7 ♔c3 54.♘d5+ ♔c2 55.♘e3+ ♔c1 56.♖c7+ ♔b1 57.♘c4 ♔a2 58.♘xb2 ♔xb2 59.♖a7
45.♘g7+ It's mate in 12 so Purdy resigned.
45.♘g7+ ♔d8 46.♘e6+ ♘xe6 47.h7 ♔c8 48.h8=♕+ ♘d8 49.♕f8 ♖b7 50.♖a8+ ♖b8 51.♕c5+ ♔d7 52.♖xb8 a4 53.♕d5+ ♔c7 54.♖xd8 a3 55.♕d6+ ♔b7 56.♖b8+ ♔a7 57.♕b6#
Powered by Aquarium

No comments:

Post a Comment