Optical illusions

Endgames often hide their beauty. Their treasure is sometimes only detected in the analysis as during the game there is no time or stress dominates. A deceptive easy looking position not seldom demands a lot of extra research compared with what we initially believed.

In positions with a very limited amount of material it is evident that the king plays a major role. The king doesn't wait anymore in a corner but is actively participating in the battle. In some endgames you get the feeling that the king does the job all alone. Recently I studied a few of those endgames and once again I was surprised how complex and beautiful chess can be.

After my very fortunate victory in round 3 of Open Leuven on Iuliia Morozova, I had to play the same evening still against the Belgian grandmaster Alexander Dgebuadze. I didn't play well the opening which forced me to spend a lot of time avoiding an immediate defeat. Later in the game I managed to fight back but in the end I run out of time. Short of time I blundered and got a lost endgame on the board after which the flag rescued me of further suffering. At least that was how we both evaluated the game till I discovered with an engine that the endgame wasn't lost at all.

[Event "Open Leuven 4de ronde"] [Date "2014"] [White "Brabo"] [Black "Dgebuadze, A."] [Result "0-1"] [ECO "C18"] [WhiteElo "2337"] [BlackElo "2510"] [SetUp "1"] [FEN "8/1k6/1p6/3p4/3PbBpP/2p3P1/5P2/5K2 w - - 0 37"] [PlyCount "6"] 37. Ke1 $4 {(A study-like draw can still be achieved with Ke2.)} ( 37. Ke2 $1 Ka6 38. Bd6 Kb5 39. Bc7 Ka4 40. Bxb6 Kb3 41. Ba5 c2 42. Bd2 Kb2 43. h5 c1=Q 44. Bxc1 Kxc1 45. Ke3 Kc2 46. Kf4 Kc3 47. Kxg4 Kxd4 48. f3 Bc2 (48... Kc3 $2 {(Now black even loses.)} 49. fxe4 d4 50. h6 d3 51. h7 d2 52. h8=Q $18) 49. Kg5 Kc5 (49... Bh7 50. Kf4 Kc3 51. g4 d4 52. g5 Bc2 53. g6 d3 54. h6 d2 55. h7 $11 ) (49... Ke5 50. f4 Kd6 51. f5 d4 52. h6 d3 53. h7 d2 54. h8=Q $11 ) 50. f4 d4 51. f5 Bb3 (51... d3 52. h6 d2 53. h7 d1=Q 54. h8=Q $11) 52. Kf4 Bg8 53. h6 Kd5 54. g4 d3 55. Ke3 $11 ) 37... Ka6 38. Kd1 {(Now the mechanism with Bd6 does not work anymore as white miss a tempo.)} (38. Bd6 Kb5 39. Bc7 Ka4 40. Bxb6 Kb3 41. Ba5 c2 42. Bd2 Kb2 43. h5 c1=Q 44. Bxc1 Kxc1 45. Ke2 Kc2 46. Ke3 Kc3 47. Kf4 Kxd4 48. Kxg4 Kc3 49. f3 d4 50. fxe4 d3 51. h6 d2 52. h7 d1=Q $19 {(Black has other wins in the last moves too but this line shows very nicely how important the tempo was.)}) 38... b5 39. Bd6 Ka5 {(My flag dropped but honestly I was relieved not having to search any futile defense anymore.)} 0-1

The illusion is that black easily wins with his fast passed pawn and this fast bishop prevents any counter-promotion. However the truth is that the slow king is sufficiently fast to support the pawns and simultaneously slow down the passed pawn. A remarkable performance which reminds me of the famous Reti-study.

A second example of an illusion I met in the book From London to Elista which I read during my holidays in Russia. I discovered the existence of the book a few month ago by accident and I immediately decided to buy it. The book is easy readable with more shortened analysis than Kasparovs books of his matches with Karpov. The book wasn't written by Kramnik but is a project of the for me unknown Russian amateur Ilya Levitov which relied for the analytic part on the strong Russian grandmaster Evgeny Bareev. This brings a special dynamic in the book with sometimes good but also less good chapters. Especially when Ilya tries to make statements about some technical aspects then I am a bit annoyed by the ignorance. In the book after the missed winning chances by Kasparov of the 8th match-game a reference is made to the strange resignation in the game Kramnik Svidler, played in Wijk aan Zee 2004. Svidler resigned in apparently a drawn position which Ilyia unjustly classifies as a beginners-mistake.

[Event "Corus"] [Site "Wijk aan Zee"] [Date "2004.01.14"] [Round "4"] [White "Kramnik, Vladimir"] [Black "Svidler, Peter"] [Result "1-0"] [ECO "B97"] [WhiteElo "2777"] [BlackElo "2747"] [SetUp "1"] [FEN "4r3/5p2/5k1p/p2B2p1/P1K2bP1/2P2R1P/8/8 b - - 0 42"] [PlyCount "14"] 42... Re3 {(Further waiting is not pleasant and the opposite bishop-endgame looks a safe harbor for the draw.)} 43. Rxe3 Bxe3 44. Kb5 Ke5 45. Bxf7 Kd6 46. c4 Bf2 47. Be8 Ke7 48. Bc6 Kd6 49. Bb7 {(Here Svidler resigned because it was inevitable that white would create 2 distanced passed pawns. It appeared an optical illusion as afterwards they soon discovered that blacks king can do much more than obstruct the c-pawn.)} (49. Bb7 Kc7 50. Bd5 Be1 (50... Kd6 51. Bf7 (51. Kxa5 Kc5 52. Ka6 Kb4 53. a5 Bd4 54. Kb7 Kxa5 55. Kc7 Kb4 56. Kd6 Bf2 $11 {(White can capture the h-pawn but black puts the king on c5 and the bishop on h4 which constructs an impenetrable defense.)}) 51... Be3 52. Be8 Bf2 53. Kxa5 Kc5 54. Bb5 Be1 55. Ka6 Bd2 56. Kb7 Ba5 57. Kc8 Kd6 {(Blacks king avoids that whites king can escape and simultaneously stops the important c-pawn.)} 58. Be8 {(Only the sacrifice of the c-pawn allows white to free the king but then black can put the king on a5 and the bishop on h4. Again this is a position which white can not win.)} Bb6 $11) 51. c5 Bd2 52. Kc4 Bf4 53. Kd3 Be5 54. Ke4 Bf4 55. Kf5 Be3 56. c6 Kd6 57. Bh1 Bf4 58. Kg6 Bg3 59. Kxh6 Bh4 60. Kg6 Kc7 61. Kf5 Kd6 62. Bg2 Kc7 63. Ke5 Bg3 64. Kd4 Be1 65. Kc5 Bb4 66. Kb5 Be1 67. Bh1 Bd2 68. h4 $18 {(I assume this position was probably what Svider thought as unavoidable when he resigned and is indeed a simple win. This reminds me of my endgame against Grochal which I earlier discussed in the article "opposite bishops".)}) 1-0

The resignation has nothing to do with being ignorant about the basic rule of 3 columns between 2 passed pawns in an oppositie coloured endgame as Ilya claims. On the contrary, it is because Svidler knows this rule that he resigned as he had the optical illusion that such lost position was unavoidable.

A third and last example of an optical illusion which I want to show here, is an endgame which I met a few years ago in an analysis. Again I was surprised that my engine immediately showed a win while I thought on first glance that it should be an easy draw. Later I tried to compose an endgame-study out of it with an introduction but I must admit that I didn't get further than a rough diamond.

[Event "Composition-idea"] [Date "2014"] [White "?"] [Black "?"] [Result "*"] [SetUp "1"] [FEN "8/4k1p1/3p3p/2p1P3/1p1P4/p1P3P1/1P5P/1K6 w - - 0 1"] [PlyCount "39"] 1. exd6 $1 Kxd6 2. dxc5 $1 Kxc5 3. cxb4 $1 Kxb4 4. bxa3 $1 Kxa3 5. Kc2 $1 Kb4 6. Kd3 $1 Kc5 7. Ke4 $1 Kd6 8. Kf5 $1 Ke7 9. Kg6 $1 Kf8 10. Kh7 $1 {(The exclamation-marks are only put to tell that there is only 1 winning move. Till this position it is just the introduction. Optically it looks a simple draw but we will quickly discover this is an illusion.)} Kf7 (10... h5 11. Kg6 $1 h4 12. g4 $1) 11. g4 $1 Kf8 12. h4 $1 Kf7 13. g5 $1 hxg5 (13... h5 14. Kh8 $1 Kg6 ( 14... g6 15. Kh7 $1) 15. Kg8 $1) 14. hxg5 $1 g6 (14... Kf8 15. g6 $1 Ke7 16. Kxg7 $1) 15. Kh6 $1 Kf8 16. Kxg6 $1 Kg8 17. Kh6 $1 Kh8 18. g6 $1 Kg8 19. g7 $1 Kf7 20. Kh7 $1 {(20 consecutive only moves which are needed for the win!)} *

Except for the funny king-walks, I find it extraordinary how white still managed to get black into zugzwang. Anybody with time and energy is free to rework the study into something looking better.

Despite the small steps, the king is often not weaker than a piece but this only becomes clear in the endgame when the king can join the action. Besides the mobility of the king is more difficult to grasp as it can reach a square via several routes in the same number of moves but optically the distance looks each time different. Maybe this is simple for somebody like Ivanchuk. When Yasser Seirawan asked him why Jobava resigned after the first round of the still ongoing Wijk aan Zee , then he replied "It is not so difficult for a good grandmaster". However for us ordinary mortals such optical illusions are much tougher to see through.

Brabo