Y Vetemaa vs Shabalov, 1986 (20...?)

Black to play and win.

Material: B for N+P. The White Kc1 has 1 legal move, d2, which Qd7 x-rays. The White Bc4 has the absolute burden of preventing Nc5-b6#. The Black Rc8 x-rays Kc1 through the Black Nc5 and the White Bc4 and Nc3, making Qd7-b5 a possible candidate. The Black Kg8 is secured from check.

Candidates (20...): Qb5

20…Qb5 (threatening 21.Qxb2#)

Counterattack is infeasible, because only check has priority over a mate-in-1. Now, we just run through the TRIP mnemonic (take-run-interpose-protect). CRIPT (<agb2002>’s suggestion) scoops up all the responses to a threat.

(1) 21.Bxc5 Nb3#

(2) 21.Nxb5 [or Nxa4] Nb3#

(Rc8 pins Bc4 to Kc1, as the preliminary analysis notes)

(3) 21.Kd2 [or R along 1-st rank] Qxb2+ then 22…Qxc3+

(4) 21.Rd2 [Rd3 drops at least an exchange]

21…Nxc3 22.Qxc3 [bxc3 Qb1#]

22…Nb6+ 23.Kd1 [Qxb3 Qxb3] [Bxb3 Rxc3]

Black would deliver mate with 20... Nb3 if 21.Bxb3 were not possible. This suggests 20... Qb5, threatening 21... Qxb2#:

A) 21.Bxb5 (or 21.Nxb5, 21.Nxa4, 21.Na2(b1,d2)) Nb3#.

B) 21.Ne4 Nb3+ 22.Kc2 Rxc4+ - +.

C) 21.Rd2 Nxc3

C.1) 22.Bxb5 Nb3#.

C.2) 22.bxc3 Qb1#.

C.3) 22.Qxc3 Nb3+ 23.Kd1 (23.Qxb3 Qxb3 - +; 23.Bxb3 Rxc3+ - +) Rxc4 24.Qxc4 (24.Qe3 Rc1#) Qxc4 - +.

D) 21.b3 Nxb3+ 22.Bxb3 Qxb3 23.Kd2 (23.B(R)d2 Qb1#) Rxc3 - +.

<20...♘e4>

The idea is not too complicated. If black clears the c file, the threat ...♖xc4 ...♖c8 appears. But this move is not forcing enough.

<21.♗d2>

But not 21.♘xe4?? ♖xc4+ 22.♕xc4 ♖c8, and black wins.

---

Only after looking at the position, perplexed, for some five minutes, I noticed that BOTH ♗c4 and the ♘c3 are pinned to prevent 20...♘b3#. (What I previously thought was that, if either moved, the other one would be automatically pinned.) Thus, removing one of the pinned pieces from its place was completely unnecessary. What I had to do is exploit the pin, not try to create one.

The obvious way to proceed is 20...b5. But there is a stronger move.

<20...♕b5!>

Creating the threat 21...♕xb2#.

<21.♖d2 ♘xc3>

And, as the following sequence shows, material loss is unavoidable.

<22.♕xc3 ♕xc4 23.♕xc4 ♘b3+ 24.♔d1 ♖xc4>

20...b5 doesn't eliminate the bishop since 21. Ba2 allows the bishop to continue protecting the b3 sqaure. The knight on a4 can also attack the c4, but the bishop can still escape to 21. Ba2. The only other piece available to attack the bishop c4 is the queen by 20...Qb5.

Upon examining this move the casual observer sees that 21...Qxb2 is also mate. Then if 21. Bxb5 Nb3++ or 21. Nxb5 Nb3++ as the bishop is pinned by the rook on c8. But black can play 21. Rd2 defending b2. Oh boy now what.

To make a long story short, 21...Nxc3, if 22. bxc3 Qb1+++ or 22. Qxc3 Nb3+ 23. (Q or B)xb3 Qxb3 and then queen cannot be retaken due to the pin by the rook on c8. So black goes up materially with a powerful attack to continue.

The question is when did white resign.

Old Wolf: "I don't get why this dzechiel guy has fans. Does that sort of post really interest anyone?"

The hidden double-pin on c is reminiscent of M Brodsky vs Kramnik, 1991, after the line <18..Bh6!!> 19.Qxh6?? Rxc2+:

click for larger view

Yes. This is quite rare. GM Auerbach once played an opponent who did just that. Auerbach politely notified his opponent about the "illegal move", and was in return informed, that this was fully legal.

The story was written in some old chess book under the headline "Grand Master learns the rules of Chess".

1) White was up a pawn
2) White had the two active bishops
3) White had a crushing central majority advantage 2-0. The central e5 and d5 passers were well supported by white army

4) White controlled the d-file. The major pieces on d file were well placed

Black was winning on tactical ground:

1) The c8 rook was actively placed for c-file invasion

2) The White king was fatally exposed. Its`escape route was cut off by the f5 bishop

3) b3 square was insufficently protected. A knight there would be checkmate or winning the white queen

4) The a4 knight was ready to break down resistance by attacking the c3 knight and the b2 pawn

Black still needed to include the queen and f8 rook for the final assult. It was not easy to activate the f8 rook as the only reasonable square was c8. There was, however, only one reasonable way to include the queen into the attack

candidate: 20 Qb5!

a) 21 b4 Nb3+ 22 Bxb3 Rxc3+ 23 Kd2 Rd3+ black had queen vs rook

b1) 21 Rd2 Nxc3! 22 Rc2 (22 Qxc3 Nb3+ 23 Kd1 Rxc4 ) Nb3+ 23 Bxb3 Ne2+ 24 Kd1 Nxd4

I am always a student of chess and I want to improve even further. However, after my improvement I just realized the puzzles are not challenging enough anymore. My suggestion to CG is that we should have more <tough weeks> in the future.

As puzzles become tougher, hopefully there will be more titled players kibitzing in the puzzle section. Their input will definitely help pazter like me improve.

Black to play (20...?) "Insane"

With White's impressive center, at first I thought this was White to play and win! But with Black to play, the first thing I realized was this: If you remove either the Nc3 or Bc4, Black mates with Nb3#. That helped me to find Black's move.

20...Qb5!!

Now, of course, 21. Bxb5 or 21. Nxb5 results in 21...Nb3#, as does 21. Nxa4 Nb3#. Black is now threatening both 21...Qxb2# as well as 21...Qb3, from where it could go to a2 and a1, terrorizing White's king! Now:

A) 21. Rd2 Nb3+ 22. Bxb3 Qxb3 and Black wins with the threat of 23...Nxc3 24. bxc3 Rxc3+. White is helpless now that the remaining bishop is deprived of the d2 square for defense!

B) 21. b3 Qxb3! (or Nxb3+) threatens two different mates, and White is helpless since 22. Bxb3 Nxb3 is still mate.

C) 21. b4 Nb3+! 22. Bxb3 Qxd4 23. Rxd4 Nxc3+ recovers the knight and wins the bishop to boot.