Umansky’s King Takes a Walk

 

    Mikhail Umansky (1952 - 2010) was a Soviet Master and a Correspondence Grandmaster. He was the 13th ICCF World Corrrespondence Champion between 1989 and 1998. He was also USSR Correspondence Champion in 1978. He was living in Germany at the time of his death due to heart disease. 

    In over the board play he placed 2nd in the Russian Junior Championships of 1965 and 1966 and in 1968 was awarded the title of National Master. In 1997 he was awarded the IM title. 

    He placed first in the ICCF 50 Years World Champion Jubilee, a special invitational correspondence tournament involving all living former ICCF World Champions. It was this result that made him one of the most respected correspondence World champions. 

    When this game was played in 1977, a major advancement in chess computers was developed by IBM; its computer defeating the world champion, Garry Kasparov. It’s victory was viewed as a sign of the rise of chess programs. 

    In 1978, SARGON won the first tournament for microcomputers, held in San Jose, California and IM David Levy of Scotland collected his 10 year bet by defeating CHESS 4.7 in Toronto with the score of 3 wins and one draw. The drawn game was the first time a computer drew an IM. While commercially available chess computers were beginning to fles their muscles, they could not defeat strong players, so it’s safe to assume that computers played little part in this game. 

    What drew my attention to this game was Umansky’s King walk. When white resigned it looks pretty obvious that black has an easy win, but when I started analyzing, things weren’t so clear! Black has a huge advantage when it comes to Centipawns, but reaching a forced win is a long way off! Because humans don’t plat at Stickfish’s level it’s clear that white’s resignation was premature!

 

[Event "USSR Correspondence Championship"] [Site ""] [Date "1977.??.??"] [Round "?"] [White "Lev Aronson"] [Black "Mikhail Umansky"] [Result "0-1"] [ECO "C51"] [Annotator "Stockfish 17.1"] [PlyCount "70"] [EventDate "1977.??.??"] {C51: Evans Gambit: Declined} 1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 {Yes, this game was played in 1977, not 1877! White offers a P to divert the black B. If Black accepts, white can follow up with c3 and d4 establishing a strong center. } Bxb4 {The usual response although statistically black does ever so slightly better by declining the P with 4... Bb6} 5. c3 Be7 {The main alternative is 6.. .Ba4, but in my database it doesn;t make much difference in the outcome which move blavk chooses.} 6. Qb3 Nh6 {This counter-intuitive move is the main line and is definitely best.} 7. d4 Na5 8. Qa4 Nxc4 9. Qxc4 {Black's main options are 9...exd4, 9...Ng4 and 9...d6} d5 {There appears to be no known examples of this move, but it is not at all bad. White has a bewildering array of possible capture and chooses the only one that maintains equality.} (9... Ng4 {is the most interesting move.} 10. h3 Nf6 {White has two ways to capture on e5 when in either case black stands well.}) 10. exd5 e4 11. Ne5 {Oddly, playing Bxh6 does not yield any advantage.} (11. Bxh6 exf3 12. Bxg7 Rg8 13. Be5 fxg2 14. Rg1 Bf5 {Black stands very well.}) 11... f6 12. Bxh6 gxh6 {The N has no retreat.} 13. d6 (13. Nc6 bxc6 14. Qxc6+ Bd7 15. Qa6 Rg8 {and white is facing serious difficulties.} 16. g3 Kf7 17. O-O Bh3 18. Re1 Qxd5 19. Re3 f5 {Black has a winning attack.}) 13... fxe5 14. dxe7 Qxe7 15. Qb5+ c6 16. Qxe5 Qxe5 {The series of exchanges have resulted in an ending in which the chances are equal. Five Shootouts from this position were drawn.} 17. dxe5 Rg8 18. Nd2 Bf5 19. Rg1 Rd8 20. Nc4 {The N will be strongly posted at d6} Ke7 {The beginning of an amazing journey.} 21. Nd6 Ke6 22. Nxb7 (22. Nxf5 Kxf5 {and after the white's e-Pawn goes black will have the advantage.}) 22... Rd5 23. Rb1 {This slip results in the loss of the game.} (23. c4 Rd4 {Black wants to keep control of the d-file so not 23...Rxe5} 24. Rc1 Rb8 25. Nc5+ Kxe5 26. Nb3 Rd3 {Black has only a small advantage, but in Shootoutd he scored two wins and three draws.}) 23... Kxe5 24. c4 {Black's centralized pieces are more active than white's who has to get his N back into play and, also, get the R on h1 into play.} Rd3 25. Na5 c5 26. Rb5 Bg4 27. Rxc5+ {White picks up a P, but at the same time black;s K gets to see some action.} Kd4 {White is still facing a mate threat with ... Rd1} 28. Rd5+ Kc3 29. f3 {In his rush to eliminate black's e-Pawnb white makes an elementary tactical mistake...surprising from a strong player in correspondence play.} (29. Nb3 Kxc4 30. Rxd3 Kxd3 {Capturing with the P actually slightly favors white.} 31. Nc5+ Kd4 32. Nb3+ Kc3 33. f3 {Now is the time for this.} Bxf3 34. Kf2 Kb2 {Black is slightly better after 35.h3, but NOT } 35. gxf3 e3+ 36. Kf1 e2+ {and black wins.}) 29... Bxf3 {Black bow has a decisive advantage.} 30. Kf2 Bg4 31. Rc1+ Kb2 32. Re1 {Black is winning, but only if he finds the right move!} e3+ {[%mdl 512] ...which he does. This move drives the K back where it's can;t be of any assistance.} 33. Kg1 (33. Rxe3 Rf8+ {wins material.} 34. Kg3 (34. Rf3 Bxf3 35. Rxd3 Be4+ 36. Ke3 Bxd3 37. Kxd3 ) 34... Rxe3+ 35. Kxg4 Re2 {etc.}) 33... Kc2 34. Nb3 {Black must prevent Rc1+.} Rxd5 35. cxd5 Kd3 {White resigned. He probably shouldn't have. Engines piut black's advantage at nearly four Ps, but the win still looks tedious and very long. Stockfish's solution after about 15 minutes of calculation...} (35... Kd3 36. d6 e2 37. Kf2 Rf8+ 38. Kg3 Bh5 39. Kh4 Bg6 40. Nc1+ Kd2 41. Rxe2+ Kxc1 42. Re7 a5 43. Ra7 Rf4+ 44. g4 a4 45. d7 Rd4 46. Rxa4 Rxd7 47. h3 Kc2 48. Rc4+ Kd3 49. Ra4 Kd2 50. Ra6 Rd3 51. a4 Ra3 52. a5 Kd3 53. Rd6+ Kc3 54. Rc6+ Kb4 55. Rb6+ Kc5 56. Ra6 Kb5 57. Ra8 Rxa5 {...and the evaluation remains about the same. How much longer it woulf take Stockfish to force the win I did not bother with. Between humans the result does not seem at all obvious.}) 0-1