Reno Redux — Round 3

by admin on October 28, 2009

I’m going to show you some of the highlights from my Reno tournament, starting with Round 3. I came into this round with a score of ½-1½ and was paired against Walter Shipman. I admire his play a lot — very solid and tough to beat. He plays what Jesse Kraai would call “old man chess,” but he has a right to, because he is 80 years old! I sure hope I’m playing as well as Shipman when I get to age 80.

Shipman is also the solution to my personal “six degrees of Bobby Fischer” puzzle. What is the shortest link of players connecting me to Bobby Fischer, so that I beat player 1, who beat player 2, …, who beat player n, who beat Bobby Fischer?

In my case the answer is n = 2. I beat Walter Shipman (2003 and 2009), who beat Sammy Reshevsky (1956), who beat Fischer and just about every other world champion of the mid-twentieth century.

Now, getting back to this weekend. I’ll skip over the opening and middlegame, although they were interesting, and take you straight to the endgame, which was the most interesting part. Here is the position after White’s 49th move, 49. Kb4.

 

I played the cautious 49. … Bc6, but after 50. Kc3 I didn’t see a way to make progress. If 50. … e4 51. fe fe White can play 52. Nf1, setting up a blockade on e3. I concluded that Black needs to have his bishop on the a6-f1 diagonal, and that means I have to sac the a-pawn. So I continued 50. … Bb5 51. Kb4 Be2!

To be honest, I was surprised that Shipman took the pawn. But that’s because I didn’t realize how truly awful White’s position if he declines. After 52. Kc3 e4 53. fe+ fe 54. Nb1 e3 his king is cut off and his knight has no moves. White is, in fact, in zugzwang or will be as soon as he runs out of pawn moves. So White basically loses without a fight in this line.

Therefore, Shipman was right to play 52. Kxa4! I followed up with 52. … Kd4, and it looked to me as if the king was just going to shepherd my passed e-pawn through to victory. But it’s not quite that easy! He played 53. Kb4 Bd1 54. a4 e4 55. fe fe and now a key moment was reached, although I did not realize it at all during the game.

Diagram 2: White to play and make it as hard as possible for Black to win.

In fact, I don’t think that Shipman thought very long about this move, either. The move seemed obvious — he played 56. a5? While you’re thinking about what White should have played instead, I’ll show you how the game ended. There’s a very nice little twist at the end, as you’ll see. I played 55. … e3 56. a6 and now Black has to avoid a trap.

Black to play and win. Don’t screw it up now!

White was undoubtedly hoping for 56. … e2??, which loses to 57. Nf3+! Instead, the only winning move for Black is 56. … ed! Shipman may have thought that White was still going to draw here, because both sides promote their pawns. But it’s always important, when you have a pawn race, to look at what happens after the pawns queen. After 57. a7 Bf3! (getting the bishop out of the way with a gain of tempo) 58. ef d1Q Shipman resigned, because after 59. a8Q the newly created queen falls victim to an x-ray check: 59. … Qb1+ 60. Ka5 Qa1+.

I was pretty proud of myself for finding this winning combination, and I thought this was about the narrowest margin of victory you can have in an endgame. The pawn race was basically a tie. If only White’s king had been on b5 he would have had nothing to worry about. But as you’ll see if you go back, he did not have a chance to play Kb5 earlier instead of Kb4. Chess games are sometimes decided by such subtle things!

However, I did not even suspect how subtle the endgame was until I went over the game with the computer. When we got to the position of Diagram 2 the computer recommended the following move:

Diagram 2 again, for your convenience.

56. Nxe4! …

Of course, during the game I was aware of this possibility, but to be honest I scarcely even looked at it. It just seemed to me that Black must be obviously winning. However, this piece sac actually leads to a very difficult endgame that I almost certainly would have misplayed — and so this represents a terrific opportunity for learning.

First, let me show you how I would probably have failed to win. This is the first line that I tried against the computer:

56. … Kxe4 57. a5 Be2 58. Kc5 g5 Of course, Black wants to go after White’s kingside pawns, but he also doesn’t want to lose touch with his g7 pawn.

59. Kd6! … This is the point. White wants to bring his king over to the kingside, so that his threats on the g-pawn will prevent Black from carrying out his plan.

59. … Bf1 60. Ke6! … There is no point for White in trying to hold the g-pawn. In fact, he’s glad to give it up because it lets him advance the a-pawn closer to the queening square.

60. … Bxg2 61. a6 g4 62. a7 Kf4 63. Kf6 Be4 64. Ke6 …

Position after 64. Ke6 (analysis). Draw!

This is a typical drawn position. Black can never win the h-pawn because White will bring his king to g5 and h5. In order to win the h-pawn, Black has to give up his defense of the g4 pawn. Even if Black puts his bishop on f3, it does not really defend the g-pawn, because as soon as Black’s king takes on h7 White will play a8Q Bxa8 Kxg4, draw.

The moral to draw from this is that you should not be too impressed by Black’s extra piece. When the player with an extra piece is down to his last pawn, the paramount issue is whether he can keep that pawn defended. Similarly, for the player who is a piece down, the question is whether he can place the pawn under attack. Another important issue here is the passed a-pawn. Once it gets to a7, it restricts Black’s bishop to one diagonal. In effect, it turns the bishop into a much weaker piece — instead of a three-point bishop, it’s more like a one-point bishop.

Now that we have seen how Black should not play the endgame, how should he play it?

The answer is that he needs to do everything he can to prevent White’s king from reaching the kingside and harassing the g-pawn. And this means that he has to move his king back, away from the White’s pawns on g2 and h2. To me, this is extremely counterintuitive. I mean, it’s obvious that eventually Black’s king will have to capture on h2. How could moving backward help?

As it turns out, the reason Black doesn’t head toward h2 with his king is that he must win the a-pawn first! Only after he achives that will his bishop turn back into a normal, three-point bishop instead of a crippled one-point bishop.

So the basic strategy for Black is first, keep White’s king as far away from the g-pawn as possible, and second, if White really commits to the kingside with his king, then Black’s king can run to the queenside and win the a-pawn. Here’s the crucial line that I came up (together with Fritz) with that demonstrates the idea. Note: Because I did this analysis on the computer, I can’t guarantee that it is 100 percent correct. If you see a mistake, please let me know.

Starting again from Diagram 2:

56. Nxe4! Kxe4 57. a5 Be2 58. Kc6 g5 59. Kd6 Kf5!

This is the key idea, which I never would have found over the board. Black makes it as hard as possible for White to penetrate on the kingside. Let’s see what happens if White won’t take no for an answer: 60. Ke7 Bc4 61. Kf8 Kf6! Again, retreating to victory! Black has now cut off the White king invasion.

62. Ke8 g4 63. Kf8 Bf1 Now it’s finally time to go after the g-pawn.

Diagram 4. Position after 63. … Bf1 (analysis).

Now White is at a crossroads. Does he play 64. Kg8, insisting on the kingside? Does he play 64. Ke8, trying to get back over to the queenside? Or does he play 64. g3, passing?

First, I should tell you that 64. g3 is bad. The move g3 is always bad for White because it allows Black to blockade the kingside pawns with no effort at all, keeping his bishop on f8. The bishop will finally move to h3 when it has to, but by then Black’s king will have gotten within safe range of the a-pawn.

Also, 64. Kg8 allows Black to win fairly simply, because his king can again cross over to the queenside and win the a-pawn. Although I say it’s “simple,” Black succeeds by just one tempo:

64. Kg8 Bxg2 65. a6 Ke6 66. a7 Kd6 67. Kg7 Kc7 68. Kg6 Kb7 69. Kf5 Bh3 (see diagram)

Diagram 5. After 69. … Bh3 (analysis).

Success! This is the basic won position that Black was aiming for. His bishop held off the a-pawn’s advance just long enough for his king to get there, and then switched to defense on h3, where it blockades White’s pawn. Notice also that it was very important that White’s king was chased all the way to the back rank. Otherwise Black would not have had time to execute this plan.

Finally, going back to Diagram 4, if 64. Ke8 Black simply keeps the opposition with 64. … Ke6. Every step toward the a-pawn helps him, and if White ever goes the other way then Black will take on g2 and transpose into the line we just saw.

This is, of course, not an exhaustive analysis of the endgame. White has other plans — he could have kept his king on the queenside and forced Black’s bishop eventually to sac itself for the a-pawn. Also, there is the technical matter of how Black wins from Diagram 5 after he captures the a-pawn — but I will let you work that out. The main point of the endgame, though, is that it’s definitely not a brainless case of winning the g2 and h2 pawns and pushing the g-pawn. And it’s particularly important to realize that if Black’s bishop gets tied down to defending a8, it turns into only a one-point bishop and Black loses his apparent material advantage.

In this case, I think that the lesson was almost as important as the outcome of the game! However, this was certainly a key game for me in the tournament, putting me into the right frame of mind for the final three rounds.

Print Friendly

{ 5 comments… read them below or add one }

Eric Fingal October 28, 2009 at 12:39 pm

Dana, Have you ever played (or beaten) Victor Pupols ? He is a friend of my old friend Jim Blackwood in Arizona and a solid master who has played many times in the Open section in Reno when I have been there. He beat Bobby in the US Junior championship in 1955.That would get you even closer than Walter Shipman in degrees of seperation.
-Eric

Reply

Ashish October 28, 2009 at 1:46 pm

Wow. Just reading that made my head hurt. Nice win, and great analysis!

Reply

Ashish October 28, 2009 at 1:50 pm

One more thing: Assuming Fischer number works like ErdÅ‘s number, “n=1” would apply to those who themselves defeated Fischer. So your n is actually 3. As is (I’m pretty sure) mine, if we allow beating future GMs when they were children.

Reply

admin October 28, 2009 at 1:54 pm

Eric, great idea! Yes, I have played Pupols but unfortunately I lost to him. So that link doesn’t quite work. 🙂

Ashish, sorry about the headache! All that analysis is just fantasy chess anyway, until I can actually use it in a game.

Reply

admin October 28, 2009 at 2:04 pm

By the way, for what it’s worth, my Fischer number equals my Erdos number. I wonder how many people can say that?

(My Erdos number is also three: I coauthored a paper with Frank Morgan, who wrote a paper with Zoltan Furedi, who wrote ten papers with Erdos.)

Reply

Leave a Comment

Previous post:

Next post: