This was the position after move 70.Qc6+. Gelfand played 70…Qf3 and the game was drawn on move 100. However, I believe from a quick glance with the help of a primitive computer software, Black is much better off playing 70…Kg1. Here is a possible line:
70… Kg1 71. Qxb6 Qd1+ 72. Ka2 Qd5+ 73. Kb1 h3 74. Qh6 Qd3+ 75. Ka2 Qc4+76. Ka1 Qf1+ 77. Ka2 Qf7+ 78. c4 Qxc4+ 79. Ka1 h2 80. Qe3+ Kg2 81. Qd2+ Kg3 82.Qd6+ Qf4 83. Qg6+ Kf2 84. Qc2+ Ke1 85. Ka2 c4 86. Qh7 c3 -+
Do you agree? Did Gelfand miss other opportunities? Perhaps IM Ken Regan can go over this line with his incredible analysis 🙂
Chess Daily News from Susan Polgar
yes, I also thought that kg1 is the natural answer during the game and did not understand qf3.
The point is that usually the side with the pawn wins such endgames when the king of the other side is exposed, and then there is almost always a way to cover Q checks with opposite check that trade queens and win, however in this endgame Leko had his king safe and prevented such exchange and got the perpetual.
The only option was when the white king is still on the back rank to block qb1+ with forced queen exchange, and this is the reason I believe it was the last chance to do it with Kg1.
After qf7+ white is forced to play c4 since ka1 seem to run into the back rank problem, and after c4 is gone we have the option to play c4-c3 and expose the white king.
However there seem to be many other variations until black play qf7+, so I am not sure it is winning, but it is certainly a better chance…
70. …Kg1
71. Qxb6 Qd1+
72. Ka2 Qd5+
73. Ka1 h3
74. Qg6+ Kf2
75. Qc2+ Kg3
76. Qg6+ Kh4
77. Qg1 Qe5
78. Qf2+ Kg5
79. a4 h2
80. Qg2+ Kh4
81. Ka2 Qg3
82. Qh1 Qg1
83. Qe4+ Qg4
84. Qe1+ Kh3
85. Qf1+ Kg3
86. Qe1+ Kg2
87. Qd2+ Kh3
88. Qe3+ Qg3
89. Qe6+ Kg2
90. Qd5+ Qf3
91. Qg5+ Kf1
92. Qc1+ Kf2
93. Qd2+ Qe2
94. Qd5 c4
95. Qh1 Kg3
96. a5 Qf3
97. Qe1+ Kh3
98. Qe6+ Kg2
99. Qg6+ Kf1
100. Qb1+ Kf2
101. Qc2+ Kg3
102. Qg6+ Qg4
103. Qd6+ Qf4
104. Qc6 Qf1
Now here is where programs may lose the draw by moving 105. Qg6+? allowing the black king to move over to a1 and then queen the pawn
instead keep the queen between the black king and the white king side of the board leading to repetition as follows:
105. Qd6+ Kg2
106. Qd5+ Qf3
107. Qd2+ Qf2
108. Qd5+ Kg1
109. Qd1+ Kg2
110. Qd5+ Kg3
111. Qd6+ Kg2
112. Qd5+ 1/2-1/2
Hey Kg1 anon,you’re the analysis man of the year!!!
Chess is no longer chess when theories are now getting to 30-40 deep.
Interesting idea. I saw the first few moves of it while watching ICC. It looked the most promising to me, since it seemed to me that the key to the position for black was to force white’s king onto the back rank. Since this comes closer to achieving it, it is worth a try, although probably a draw.
Dear friendly people, I am a little confused on when a player should not play on to win any more, risking being rude. Some discussions here went to a moment that Kramnik should not have played on a draw endgame. Are there some rules on when to offer a draw? Thank you!
70… Kg1 71. Qxb6 Qd1+ 72. Ka2 Qd5+ 73. Kb1 h3 74. Qh6 Qd3+ 75. Ka2 Qc4+76. Ka1 Qf1+ 77. Ka2 Qf7+ 78. c4 Qxc4+ 79. Ka1 h2
Susan, I did not see the point of 78. c4, because Black can always back to f7 after 79…Qf1+ 80. Ka2 Qf7+, now White has no second c4 to give…
Susan, thanks for the vote of confidence, “your wish is my command”: I did a few hours of analysis and posted a PGN file over in chessdiscussion.com, click “Chess Events” then “Boris Gelfand 7 hours 100 move game!?”
In your line, White can play 73.Ka1, which is better “on principle” as the King is more shielded from checks. Then on 73…h3, 74.Qh6 is an only-move (the tempting 74.Qg6+ loses to …Qg2!), but it suffices.
However, 73.Kb1 is OK so long as after 73…h3 74.Qh6 Qd3+ 75.Ka2 Qc4+, you ignore the “on principle” and play 76.Kb1! The difference from 76.Ka1 (which indeed loses) appears to show up in one key line where Black’s King goes to d1 and Queen to e1, setting up a battery that will support h1=Q. But White can play Qg2, threatening mate by Qc2 supported by the King on b1! With WK on a1 there is no support for this, and Black wins. That’s how subtle Queen endings can be! The whole line is in my PGN file.
Enjoy, and as I wrote on your discussion site, I’ll try to add more if time allows. (My comments in the PGN file are a bit sketchy as-is.)
Hi KWRegan,
I posted a reply to the original topic of
Boris gelfand 7 hours 100 move game !?
There is a link to a better presentation of your analyses and some suggestions.
I’m posting it here
QP analyses as a shortcut.