As is my custom, I only say which side is to play: and not giving an idea if the move wins or otherwise, unless on occasion I think signposting would be helpful. Instead, the problems are posed with the instruction to decide what you would play, as in a game.

**Black to play: evaluate 1….Be8**

Reg Clucas v Allan Beardsworth, Bramhall 11/2/2020

**Solution**

I played 1…Re2+ and won, but I considered and decided against 1…Be8 because of 2 f7: White either wins the Bishop back, or queens his pawn.

But no: 1…Be8! is Komodo’s first choice and if 2 f7 Re2+!!

Either the rook is not captured, in which case 2…Kg7 and the Re2 guards e8 whilst the King guards f8; or the rook is captured, when the zwischenschach 3..Bb5+ wins since 4…Kg7 guards the now sole queening square.

Lovely. And it would have been impossible for me to have seen.

**FEN**

8/6Rp/5P1k/1b6/1P6/4K3/7r/8 b – – 0 46

]]>As is my custom, I only say which side is to play: and not giving an idea if the move wins or otherwise, unless on occasion I think signposting would be helpful. Instead, the problems are posed with the instruction to decide what you would play, as in a game.

**White to play: how to meet 1…Bf4, as played in the game?**

Paravyan D v Golubov S, St Petersburg Korchnoi Memorial op 20/8/2018

**Solution**

1…Bf4 2 Qc7!!

If 2…Bc7 then 3 g5 smothered mate; if 2…Be6 then 3 Qf4+; so 2…Bh6 was played, met with by 3 Qe5+ Kg5[]

4 Rg3! or 4 h4+ and mates.

White, David Paravyan, is the winner of the recent Gibraltar 2020 tournament.

**FEN**

r4r2/pp3p2/5kpB/3p1b2/q2P1bP1/1R6/P1Q2P1P/4R1K1 w – – 0 24

]]>As is my custom, I only say which side is to play: and not giving an idea if the move wins or otherwise, unless on occasion I think signposting would be helpful. Instead, the problems are posed with the instruction to decide what you would play, as in a game.

**White to play**

Paravyan D v Golubov S, St Petersburg Korchnoi Memorial op 20/8/2018

**Solution**

1 Bh6+! is the natural first move, its first point being 1…Kh6 2 Rh3+ and picks up the LPDO Qa4.

In the game, Black played 1…Kf6, met with by 2 g41

2….Bc2 3 Rf3+ Bf5[] 4 g5 is an unusual smothered mate.

**FEN**

r4r2/pp3pk1/3b1Np1/3p1b2/q2P4/1R6/P1QB1PPP/4R1K1 w – – 0 22

]]>

**White to play**

Leenhouts v Van Wely, Amsterdam 5/7/2018

**Solution**

White missed the best move, but won after a melee anyway. But 1 Ra8+! is a standard tactic; either diverting the K (1…Ka8 2 Qa4+ 3 Qa7+ 4 Qb6 and mate) or putting the B on an unfortunate square (1…Ba8 2 Qa4 and either 2..Bb7 met as before, or Qa7+ Q*a8)

**FEN**

1k1r1br1/1b1p3p/1p2p1qP/1N4p1/2PP3n/1Q3p2/1P4PN/R1B2RK1 w – – 0 21

]]>

**White to play: what happens after 1…h6?**

Mamedyarov v Karjakin, Saint Louis blitz 15/8/2018

**Solution**

1..h6 is met by 2 h6! and **pins and wins. **

Next move, B*f5, then the Bg3 falls.

**FEN**

3r1rk1/pbq2ppp/1p2p2N/2p2nP1/2P4P/PP1PP1bB/1B2KQ2/R5R1 b – – 0 22

]]>

**White to play**

Mamedyarov v Karjakin, Saint Louis blitz 15/8/2018

**Solution**

1 Nh6+! has to be tried.

If 1…Kh8, then what?

Only one move wins.

**Solution**

2 Bg7+! either enabling a knight fork after 2…Kg7 3 Nf5+, after which the Bg3 is captured, or if 2…Ng7, the Bg3 can again be captured.

**FEN**

3r1rk1/pbq2ppp/1p2p3/2p2nP1/2P3NP/PP1PP1bB/1B2KQ2/R5R1 w – – 0 22

]]>

**Black to play**

Samborski v Renkowski, Pokrzywna 7/9/2018

**Solution**

I would play 1..Kg3 or 1..Kg2 “at random”; in fact, I would likely play 1…Kg3 just to be “nearer” to the Pf4; and not 1..Kg2, the move played in the game.

Well, 1…Kg2 is a draw, the game result; whilst 1…Kg3 wins. The difference turns out to be that after 1…Kg3, 2 Nd3 is met by 2..f3, and the b-pawn is poisoned, since 2..f2 3 Ke2 Kg2 and the f-pawn promotes.

Whereas after 1…Kg2 2 Nd3 f3 3 Nb4! is possible.

3…f2 4 Nc2! or 4 Nd5! and a Ne3+ forks if the Pawn is promoted (other than to a N).

Chess is a very deep game!

**FEN**

8/8/8/2N5/1p3p1p/5k1P/3K4/8 b – – 0 63

]]>

**White to play**

Iturrizaga Bonelli v Warmerdam, Vlissingen, 9/8/2018

**Solution**

This took me a couple of sittings to solve, but, when I got the light-bulb moment, it was a moment of pleasure.

1 Nc6! disrupts Black’s counterplay sufficiently that Black is overpowered, typically by Qg5+.

**FEN**

2r1b1k1/1p2pp2/1q1p1b2/3P1PpR/3N2P1/1P3P2/3Q2K1/B7 b – – 0 36

]]>

**White to play**

You can tell from the Chessbase evaluation that “something happened”. Aronian missed something in the position below.

Aronian v Nakamura, Saint Louis Rapid 12/8/2018

**Solution**

I doubt I would ever, not even in a month of Sunday’s, have played the move Levon missed: 1 c3; Komodo gives it as the only move which gives a winning advantage; all other moves, including Levon’s 1 d7, losing.

1 d7 lost to 1…Rb4; from this can be seen that 1 c3 is “simple” prophylaxis. If Black doesn’t mate White, e.g if he plays 1…Bc3 2 Nc3 Rc3, then 3 d7 and the pawn queens, since 3…Nf6 loses to 4 Qg7 mate.

**FEN**

4nrk1/5pp1/1q1P4/1p6/2r1N2P/p6B/PbP3QP/1K1R2R1 w – – 0 30

]]>

**Black to play and draw**

K Arkell v K Toma, Hull 30/7/2018

**Solution**

In the game, Black played 1…Kf7 and lost; a typical Arkell endgame grind.

I struggled (failed) this one: trying to make 1…Be4 or 1…g5 a draw, without success.

But 1…Ke7 is a draw: I don’t really know how/why, but having played around with a few lines, can just about see that it is. The engines are their unhelpful best selves, showing 0.00 as tablebase outcomes.

(Having spent a good while on this puzzle, I am not much the wiser. It seems to be just a matter of calculation; after 1…Ke7, 2 e5+ can be met by 2…Bc6+ pushing the k away; whereas if 1…Kf7 2 e5+ then 2…Ba6+ is met by 3 Kd6).

**FEN**

8/8/5kp1/3K4/3NPP2/8/8/1b6 b – – 0 54

]]>