Interesting endgame from the Ontario Open

[Hugh Siddeley]

Black has just played 40..Kg7-f6 to reach the first time control with 11 seconds to spare. Was this a mistake? Assess the position. Should White now play Qxe4?

[Tom O'Donnell]

Re: Interesting endgame from the Ontario Open

You don't mention how much time you spent on ...Kf6 but I will assume less than 30 seconds. I think you made a mistake in the metagame sense. A "pass" move like ..Qe1 or ...Qe3 is a lot less risky than entering into a K and P ending where some finesse might be the difference between a draw and a loss. Even if ...Kf6 draws (and I suspect it does) you still made a mistake, imo.

[Hugh Siddeley]

Re: Interesting endgame from the Ontario Open

You don't mention how much time you spent on ...Kf6 but I will assume less than 30 seconds. I think you made a mistake in the metagame sense. A "pass" move like ..Qe1 or ...Qe3 is a lot less risky than entering into a K and P ending where some finesse might be the difference between a draw and a loss. Even if ...Kf6 draws (and I suspect it does) you still made a mistake, imo.

I had about 1-2 minutes to consider 40..Kf6. I certainly considered a ''pass'' move but opted to take a chance with the text. The real question is, again, what happens after 41. Qxe4, which is what my opponent subsequently played. Assess the resultant pawn ending.

[Egidijus Zeromskis]

Re: Interesting endgame from the Ontario Open

White might win after Qe4 as when Black gets to c2 and b3 , White will start a push with g and h in a way that they will quen with the check on g8.
If Black does not get to c2, White would create a pass pawn and then collect on a queen side other pawns.
That would be estimations. Conceate variations might show a different outcome.

[Hugh Siddeley]

Re: Interesting endgame from the Ontario Open

White might win after Qe4 as when Black gets to c2 and b3 , White will start a push with g and h in a way that they will quen with the check on g8.
If Black does not get to c2, White would create a pass pawn and then collect on a queen side other pawns.
That would be estimations. Conceate variations might show a different outcome.

I need some cement, some concrete analysis there Mr. Z. I think you can do better than that ;)

I have the answer btw

[Egidijus Zeromskis]

Re: Interesting endgame from the Ontario Open

I need some cement, some concrete analysis there Mr. Z. I think you can do better than that ;)

Have enough own games :)

But the rule of a farther located 'passed' pawn (h3-h4-h5) should dominate over a blocked e4.

[Adam Cormier]

Re: Interesting endgame from the Ontario Open

White seems to have winning chances due to the possibility of an outside pawn, and I can't really see a good line for black.

Qxe4 fxe4 Kg2 Ke5 Kf2...Here I looked at a few moves none of them seem to be any good

a5? ke3 kf5 g4 and if he goes for the h-pawn black loses by kg5 kxe4 kh4 ke5 kxh3 g5. If he doesn't go for the pawn than ke5 h4 is game over. If he moves the king to d5 initially white goes for the same outside passed pawn idea (starting with h4)

b5!? axb5 axb5 Ke2! (Ke3 runs into b4 and then white must either retreat the king allowing d4 or give up key squares on the kingside, example: g4 g5. or h4 Kf5) and now if b4 Ke3 and black is in zugswang, but white only wins a pawn and both sides get a queen so it's probably a draw-the line goes, kf5 g4 kg5 Kxe4 kh4 g5 kxh3 kd5 kg4 kxc5 kxg5 kxb4 and both sides will make a queen at the same time, probably drawn but practically white has winning chances.

Ke6?! Ke3 Ke5 c3 and I don't see a way for black to survive as every move hurts black's position (if a5 c4, or kf5 g4 )

The b5 line probably draws with best play.

[Ed Seedhouse]

Re: Interesting endgame from the Ontario Open

For what it's worth Houdini 3 thinks this is equal after the exchange of queens.

[Hugh Siddeley]

Re: Interesting endgame from the Ontario Open

White seems to have winning chances due to the possibility of an outside pawn, and I can't really see a good line for black.

Qxe4 fxe4 Kg2 Ke5 Kf2...Here I looked at a few moves none of them seem to be any good

a5? ke3 kf5 g4 and if he goes for the h-pawn black loses by kg5 kxe4 kh4 ke5 kxh3 g5. If he doesn't go for the pawn than ke5 h4 is game over. If he moves the king to d5 initially white goes for the same outside passed pawn idea (starting with h4)

b5!? axb5 axb5 Ke2! (Ke3 runs into b4 and then white must either retreat the king allowing d4 or give up key squares on the kingside, example: g4 g5. or h4 Kf5) and now if b4 Ke3 and black is in zugswang, but white only wins a pawn and both sides get a queen so it's probably a draw-the line goes, kf5 g4 kg5 Kxe4 kh4 g5 kxh3 kd5 kg4 kxc5 kxg5 kxb4 and both sides will make a queen at the same time, probably drawn but practically white has winning chances.

Ke6?! Ke3 Ke5 c3 and I don't see a way for black to survive as every move hurts black's position (if a5 c4, or kf5 g4 )

The b5 line probably draws with best play.

Thanks for your reply Adam. And your analysis is spot on.

White certainly has the more dangerous threat of creating an outside passed pawn. And Black must play very accurately to stave off a loss in the 43..b5 line you mentioned. If White manages to swap off the newly promoted queens it's game over for Black.

White is not playing completely without risk from the previous diagram either however, the game continued 42. Kg2 Ke5 43. c3 (preventing Black from infiltrating with 43..Kd4), reaching the following diagram. What happens now?

[Andrew Peredun]

Re: Interesting endgame from the Ontario Open

In the initial position (after 40... Kf6):

41. Qxe4 fxe4 42. Kg2 b5! is the only move that draws (although I had to tablebase all of the resulting queen endgames to make sure):

a) 43. axb5 axb5 44. Kf2 b4 leads to a position where black must make sure to only play Ke5 after white has played Ke3. White can even lose spectacularly in a line such as 45. Ke2 Kf5 46. Kd2 Ke6 47. c3 c4! when black breaks through. 44. c3 leads to a similar position where black will not lose as long as he does not get into a Ke3 vs Ke5 zugzwang.

b) 43. Kf2 bxa4 44. bxa4 c4! creates the same Ke3 vs Ke5 situation where both sides have a tempo to spend on the queenside before the pawns become blocked (draw after 45. a5 c3 or 45. c3 a5).


42... Ke5 (instead of b5) loses:

43. Kf2 b5 44. axb5 axb5 45. Ke2 Kf5 46. Ke3 Ke5 47. c3 Kf5 48. g4 Kg5 49. Kxe4 Kh4 50. g5 Kxh3 51. Kd5 Kg4 52. Kxc5 Kxg5 53. Kxb5 is mate in 59 according to tablebases (not a very practical assessment during a tournament game, to be sure) although it's interesting that in order to win white must queen the b pawn not the c pawn.

In the position after 42. Kg2 Ke5 it is not necessary to play 43. c3 as 43. Kf2 Kd4 loses to 44. h4 Kc3 45. Ke3 Kxc2 46. g4 Kxb3 47. h5 gxh5 48. g6! when white queens with check.

After 43. c3, black wins by 43... a5 when the only way for white to prevent ... b5 and ...c4 followed by queening the a pawn is to play c4 himself, after which black wins by 44... Kd4 45. Kf2 Kd3 46. Ke1 Ke3! winning the kingside pawns, or to play 44. Kf2 b5 45. c4, which loses to ... bxa4 46. bxa4 Kd4 when black capturing the c pawn and queening is faster than white pushing the kingside pawns.

Adam's line (42... Ke5 43. Kf2 b5 44. axb5 axb5 45. Ke2 b4 46. Ke3 Kf5 47. g4 Kg5 48. Kxe4 Kh4 49. g5 Kxh3 50. Kd5 Kg4 51. Kxc5 Kxg5 52. Kxb4) is also a tablebase win for white, but much cleaner is 49. Kf4! Kxh3 50. Kg5 Kg3 51. Kxg6 Kxg4 52. Kf6 Kf4 53. Ke6 Ke4 54. Kd6 Kd4 55. Kc6 c4 56. Kb5.

In many positions it's possible for white to play h4 too early, but this seems to always lose to Kf5 -> Kg4 followed by sacrificing the e pawn to win both of white's kingside pawns, so I didn't include those.

[Hugh Siddeley]

Re: Interesting endgame from the Ontario Open

In the initial position (after 40... Kf6):

41. Qxe4 fxe4 42. Kg2 b5! is the only move that draws (although I had to tablebase all of the resulting queen endgames to make sure):

a) 43. axb5 axb5 44. Kf2 b4 leads to a position where black must make sure to only play Ke5 after white has played Ke3. White can even lose spectacularly in a line such as 45. Ke2 Kf5 46. Kd2 Ke6 47. c3 c4! when black breaks through. 44. c3 leads to a similar position where black will not lose as long as he does not get into a Ke3 vs Ke5 zugzwang.

b) 43. Kf2 bxa4 44. bxa4 c4! creates the same Ke3 vs Ke5 situation where both sides have a tempo to spend on the queenside before the pawns become blocked (draw after 45. a5 c3 or 45. c3 a5).


42... Ke5 (instead of b5) loses:

43. Kf2 b5 44. axb5 axb5 45. Ke2 Kf5 46. Ke3 Ke5 47. c3 Kf5 48. g4 Kg5 49. Kxe4 Kh4 50. g5 Kxh3 51. Kd5 Kg4 52. Kxc5 Kxg5 53. Kxb5 is mate in 59 according to tablebases (not a very practical assessment during a tournament game, to be sure) although it's interesting that in order to win white must queen the b pawn not the c pawn.

In the position after 42. Kg2 Ke5 it is not necessary to play 43. c3 as 43. Kf2 Kd4 loses to 44. h4 Kc3 45. Ke3 Kxc2 46. g4 Kxb3 47. h5 gxh5 48. g6! when white queens with check.

After 43. c3, black wins by 43... a5 when the only way for white to prevent ... b5 and ...c4 followed by queening the a pawn is to play c4 himself, after which black wins by 44... Kd4 45. Kf2 Kd3 46. Ke1 Ke3! winning the kingside pawns, or to play 44. Kf2 b5 45. c4, which loses to ... bxa4 46. bxa4 Kd4 when black capturing the c pawn and queening is faster than white pushing the kingside pawns.

Adam's line (42... Ke5 43. Kf2 b5 44. axb5 axb5 45. Ke2 b4 46. Ke3 Kf5 47. g4 Kg5 48. Kxe4 Kh4 49. g5 Kxh3 50. Kd5 Kg4 51. Kxc5 Kxg5 52. Kxb4) is also a tablebase win for white, but much cleaner is 49. Kf4! Kxh3 50. Kg5 Kg3 51. Kxg6 Kxg4 52. Kf6 Kf4 53. Ke6 Ke4 54. Kd6 Kd4 55. Kc6 c4 56. Kb5.

In many positions it's possible for white to play h4 too early, but this seems to always lose to Kf5 -> Kg4 followed by sacrificing the e pawn to win both of white's kingside pawns, so I didn't include those.

Excellent stuff Andrew. My initial reason for posting this endgame was to show how tricky this position is and how complicated chess can be even with just kings and pawns. Do we really need knights, bishops, rooks, and queens?

As Andrew points out, the position after 41. Qxe4 fxe4 is a draw with perfect play, but it's amazing how both sides can go astray in one single move. After 42. Kg2 Ke5?? (Black goes from drawing to losing) 43. c3?? (White goes from winning to losing) 43..b5?? (Black goes from winning to losing) as was played in the game. Just sickening :o