Boy, if it isn’t one catastrophe, it’s another. First there was the coronavirus. Then the protests in many U.S. cities (though, thankfully, not very much unrest in Santa Cruz, where I live). Now there are fires in the mountains north of Santa Cruz, set off by a lightning storm this weekend.
In the eastern and midwestern U.S., thunderstorms are a regular occurrence in summer — sometimes almost daily. But in California, we usually go at least four or five months without significant rain, from May through September. So the vegetation gets really, really dry, and if something does happen to start a fire, you’d better watch out.
So on Sunday we had a storm with 2500 strokes of lightning in the Santa Cruz mountains (who counts these things?), and by yesterday (Tuesday) several major fires were burning, all of them completely uncontrolled. I want you to know that I am not in danger: the fires would have to expand to biblical proportions to get close to my house. But several of my friends who live in the hills have been forced to evacuate. I have seen pictures in the news of houses burned down to the ground, in areas that I have driven through many times.
Here’s a picture of the overnight ash fall in my back patio, taken yesterday morning. It looks as if someone took a wood chipper to the forest, launched all the debris into the atmosphere and it rained down on Santa Cruz.
While people are evacuating out of the hills, their animals are being evacuated into the animal shelter where Kay and I volunteer. This meant that all the animals who were previously being housed at the shelter needed to be re-housed in volunteer homes. Kay and I are now taking care of a very friendly black kitten named Prince. We don’t know for how long. We were already fostering two kittens named Bramble and Poppy, who were scheduled to go back to the shelter this week, but until this situation clears up we are continuing to take care of them. So now, for the first time, we have an upstairs kitten and two downstairs kittens, plus our permanent cat Max, who is allowed to commute between the upstairs and downstairs.
This being a chess blog, I should write a little bit at least about chess. Here’s an interesting position I got to yesterday against the computer.
Which way would you get out of check? Would you play 42. Kd3, 42. Kd2, 42. Kf4, or 42. Rxc5? More importantly, what is your plan for winning? (The Mike Splane Question.)
FEN: 2R5/3nk2p/6p1/1PbP4/4B3/4K3/6P1/8 w – – 0 43
With a lead of an exchange and a pawn, and with two passed pawns, it seems as if White should be winning easily. And in fact, Fritz evaluates the position at +5 pawns for White. But computer evaluations don’t win the game for you. You need a plan.
When I looked at this position, I saw that Black has a pretty good blockade of White’s pawns. I also saw that Black has a weakness that is hard to defend — the two kingside pawns. But the direct approach of attacking them with my rook seems dubious; if I ever play Rh8 my rook might get shut in after … Nf8. So I thought, why not go after the pawns with my king instead? So I chose 43. Kf4!?
In retrospect I don’t know quite what to say about this move. It works. Everything works, in this position. But I came perilously close to squandering my advantage, and it’s only because of a somewhat serendipitous idea at move 49 that it works at all. Let’s see what happened:
43. Kf4!? Bd6+ 44. Kg5 Bc5 45. Kh6!? …
Almost leading to disaster. I think that what bothers me is that I didn’t actually see the danger until it was too late.
45. … Be3+
And only now did I realize that 46. Kxh7 would cost me a piece after 46. … Nf6+ 47. Kxg6 Nxe4. If you have computer-like tactical vision, you can see that White nevertheless has a forced tactical win with 48. b6!! Nd6 (or 48. … Bxb6 49. Rc6, when White fortuitously wins back the piece after 50. Re6+) 49. b7!! Nxb7 50. Rc7+ Kd6 51. Rxb7 Kxd6 with an easy win for White, thanks to the g-pawn that has been a bystander for the whole game.
I did not have computer-like tactical vision. I only saw that 46. Kxh7 would lose a piece, so I played
46. Kg7 …
And now the computer surprised me by not playing 46. … Bd4+, which I thought was the obvious move. It played 46. … Bc5, after which I was able to move my bishop to a safe square on f3 and eventually round up the h- and g-pawns, just like I planned. Here are the moves: 46. … Bc5?! 47. Rc7 Bd6 48. Ra7 Bc5 49. Ra6 Bd6 50. Bf3 Bc5 51. Bg4 Nb6 52. Be6 Be3 53. Kxh7 and White wins. If 53. … g5 54. Kg6 Nc4 55. Bg4 Black is in zugzwang: he cannot protect the g-pawn and blockade the d-pawn and blockade the b-pawn.
But the real question is how does White win after 46. … Bd4+?
FEN: 2R5/3nk1Kp/6p1/1P1P4/3bB3/8/6P1/8 w – – 0 47
Of course I would be bitterly disappointed to take a draw with 47. Kh6 Be3+ 48. Kg7 Bd4+. The only other option is to sacrifice my bishop. But as it turns out, the bishop sacrifice is completely winning! After 47. Kxh7 Nf6+ 48. Kxg6 Nxe4 comes the “scorpion’s sting at the tail end of the combination,” to quote Bobby Fischer’s wonderful phrase: 49. Rc4! skewering the bishop and knight. Black has no checks or anything else that could get him out of the skewer.
To me, this was both good news and bad news. It’s good news that in the end, my plan was correct. But the bad news is that I did not see the “scorpion’s sting” that makes it all feasible. If I didn’t see that, I should not even have played this plan on move 43, or at least I should have turned back on move 45 rather than playing 45. Kh6, which committed me to the piece sacrifice.
So was it foresight or fortune? A little bit of both.
Now, going back to the original position, what about the other possibilities? Both 43. Kd3 and 43. Kd2 are possible, of course. The computer prefers 43. Kd2 because 43. Kd3 would allow 43. … Ne5+, which gives Black’s pieces a slightly better blockading position. Unfortunately, it’s harder to get the computer to explain what its plan is. For some reason, it seems to want to bring its rook to c6 and then e6. More understandably, it wants to improve the bishop’s status by playing Bf3 and Bg4. So I guess I would call this plan “gradual improvement of the pieces while cutting Black’s king off from the queenside pawns by putting the rook on the sixth rank.” But it still seems pretty amorphous to me. In fact, the game variation (shown above) was pretty similar to this, with the additional ingredient that I was able to make my king an active participant in the battle. So I would say that my plan was a more incisive version of the computer’s plan.
Finally, some of you may have been tempted to play 43. Rxc5. Of course, this move works too, but I think that such a move shows a certain amount of impatience. It’s true that the best thing to do with an extra exchange is sometimes to give it back at the right moment. Especially if you can get a pawn for it, or otherwise inflict fatal damage on your opponent’s position.
But is this really the right moment? Does White get a pawn for the exchange? No. Has White done all that he can to improve his position before giving back the exchange? I think not. After 43. Rxc5 Nxc5 44. b6 Kd6, White’s plan for making progress is not exactly clear, and his bishop is still a pretty sad piece. It’s better to save the exchange sacrifice for a moment when it really breaks the defender’s resistance.
Lessons:
- It’s better to have a faulty plan than no plan at all. Sometimes the chess gods will help you out.
- True master chess involves the combination of sound planning with opportunistic tactical vision.
- When you’re up an exchange in the ending, often the right plan is to give the exchange back at just the right moment. But resist the urge to give it back too early. Improve your position to the maximum extent. Torture your opponent a little bit, and then play the sac if it’s the most effective way to break through.
- Don’t forget to look for the “scorpion’s sting.” This is a quiet move that occurs at the end of a forcing tactical sequence, often at a point where most of us would think that the tactics have ended.
{ 3 comments… read them below or add one }
I also liked the computer’s suggestion of rook to c6. Understanding what constitutes a weakness is the key to understanding this position.
“A weakness is a square that has to be defended by a piece. ” Black can not allow the white b pawn to reach b7 so putting the rook on c6 ties down both black pieces as they are both needed as guardians of the b6 square. This more active rook location also restricts counterplay by the Black king, keeping the d4 pawn safe.
Then White simply needs to force the knight to move. After Kd2, Rc6, Bf3 and Bg4. the defense collapses and stopping the b pawn will cost Black a piece. No sharp tactical analysis is needed.
As a side note, White would be thrilled to give up the d4 pawn if it involved the trade of bishop for knight. The R Vs B ending is easily won even without the d4 pawn.
Thank you for an excellent comment! I was wondering what would be a logical train of thought that would lead me to the plan of Rc6, and you give a good explanation. Here’s another way to say it: A threat, even if your opponent can defend it in the short run, may still be useful in the long run when it ties down his forces. I didn’t see the point of Rc6 because it makes two threats (b5-b6 and d5-d6) that were already defended, so I thought, “What’s the point?”
It’s ironic that in the game, I ended up moving the rook to the sixth rank and the bishop to g4 anyway. So my whole plan of the king invasion was a little bit of a red herring; Black’s position really started crumbling after 51. Bg4.
A second misconception I had was thinking that the rook on c8 was well placed. It’s not! The only thing it can do on the back rank is to harass the h-pawn, but as I mentioned, Rh8 is a really bad idea because my rook could get trapped. The rook became more useful when I moved it to the seventh rank, and then it became really useful when I moved it to the sixth.
So, just from the point of view of improving my pieces, it should have been a no-brainer to see that my rook would be better on c6.