The Lemon Fool

I've been trying to improve my chess game for the last few months and came across this difficult but beautiful puzzle. It is very difficult and I will post a link to the solution in one week's time. I hope you enjoy it.

White to play and win.



RC

I could have been a bit more helpful and included a board link:

https://lichess.org/editor/8/3P3k/n2K3p/2p3n1/1b4N1/2p1p1P1/8/3B4_w_-_-_0_1

RC

As expressed the puzzle seems a bit ambiguous.

Some puzzles say "White to play and mate in n moves". The fact that yours says only "win" could be interpreted as getting to an obviously winning position rather than actually mating the black king.

It might be possible to come up with a number of variations where white can build a clearly winning position. But your use of the word "beautiful" indicates that there is an elegant mate in there somewhere. And in that case it is usual to state "mate" rather than "win". And to cite the number of moves to that mate, with best play by both sides.

ReformedCharacter wrote:I've been trying to improve my chess game for the last few months and came across this difficult but beautiful puzzle. It is very difficult and I will post a link to the solution in one week's time. I hope you enjoy it.

White to play and win.



RC

2hrs later ... I give in

Must be a great move

Looking forward to the answer

AiY

Lootman wrote:As expressed the puzzle seems a bit ambiguous.

Some puzzles say "White to play and mate in n moves". The fact that yours says only "win" could be interpreted as getting to an obviously winning position rather than actually mating the black king.

It might be possible to come up with a number of variations where white can build a clearly winning position. But your use of the word "beautiful" indicates that there is an elegant mate in there somewhere. And in that case it is usual to state "mate" rather than "win". And to cite the number of moves to that mate, with best play by both sides.

Mate in 14.

RC

AsleepInYorkshire wrote:

ReformedCharacter wrote:I've been trying to improve my chess game for the last few months and came across this difficult but beautiful puzzle. It is very difficult and I will post a link to the solution in one week's time. I hope you enjoy it.

White to play and win.

2hrs later ... I give in

Must be a great move

The thing with chess puzzles, as opposed to actual over-the-board play, is that the first move is invariably not "great" in any normal sense of the word. Rather it is typically a quiet move that unlocks the "key" to the puzzle. As such it is usually not anything obvious like a capture or a check.

Apparently it is mate in 14 whereas most puzzles are mate in 2 to 5 moves, so it is probably not a rewarding puzzle to invest that much time in. A chess player, as opposed to a puzzle solver, would probably look at this position and think:

1) White is winning as he has a pawn on the 7th rank. However immediately promoting it to a queen fails to 1 . . Nf7+

2) Black has 2 dangerous passed pawns on the 3rd rank, but they are covered by the white knight and bishop. 1 Bc2+ blockades the c pawn, whilst the e pawn can be immediately captured by 1 Ne3.

The awkward black move is 1 . . c4 which is a discovered check that forces the white king off its preferred diagonal and where it really wants to be (e7). Other squares lead to annoying checks from the two black knights.

Lootman wrote:As expressed the puzzle seems a bit ambiguous.

Some puzzles say "White to play and mate in n moves". The fact that yours says only "win" could be interpreted as getting to an obviously winning position rather than actually mating the black king.

It might be possible to come up with a number of variations where white can build a clearly winning position. But your use of the word "beautiful" indicates that there is an elegant mate in there somewhere. And in that case it is usual to state "mate" rather than "win". And to cite the number of moves to that mate, with best play by both sides.

"Play and win" puzzles are certainly much less common than "play and mate in n moves" puzzles, and are sometimes known as "studies" or "endgame studies" rather than "puzzles" - but they are a type of chess puzzle. Not just my opinion, by the way - see https://en.wikipedia.org/wiki/Endgame_study, which starts:

In the game of chess, an endgame study, or just study, is a composed position—that is, one that has been made up rather than played in an actual game—presented as a sort of puzzle, in which the aim of the solver is to find the essentially unique way for one side (usually White) to win or draw, as stipulated, against any moves the other side plays. There is no limit to the number of moves which are allowed to achieve the win; this distinguishes studies from the genre of direct mate problems (e.g. "mate in 2"). Such problems also differ qualitatively from the very common genre of tactical puzzles based around the middlegame, often based on an actual game, where a decisive tactic must be found.

Also note that it is linked to from https://en.wikipedia.org/wiki/Chess_problem, which is turn is linked to from https://en.wikipedia.org/wiki/Chess_puzzle in a 'taxonomy' of types of chess puzzle.

Within the first link, the puzzle in https://en.wikipedia.org/wiki/Endgame_s ... ngineering seems quite similar in general characteristics to this one - white is down materially, by enough that one would normally expect black to be able to win, and yet white can avoid losing (though with the difference that white can only achieve a draw on that one, as opposed to a win on this one). Neither actually specifies a number of moves to achieve the desired result - and if they were to do so, that number would be big enough to rule out brute-force search as a plausible way to solve them.

Gengulphus

Lootman wrote:Apparently it is mate in 14 whereas most puzzles are mate in 2 to 5 moves, so it is probably not a rewarding puzzle to invest that much time in. A chess player, as opposed to a puzzle solver, would probably look at this position and think:

1) White is winning as he has a pawn on the 7th rank. However immediately promoting it to a queen fails to 1 . . Nf7+

2) Black has 2 dangerous passed pawns on the 3rd rank, but they are covered by the white knight and bishop. 1 Bc2+ blockades the c pawn, whilst the e pawn can be immediately captured by 1 Ne3.

The awkward black move is 1 . . c4 which is a discovered check that forces the white king off its preferred diagonal and where it really wants to be (e7). Other squares lead to annoying checks from the two black knights.

Another awkward black move is Nb8 - if black is allowed to play it before the passed white pawn on d7 is able to safely move to d8 and promote, he can then capture that pawn without worrying about whether it results in the capture of his knight, and while a subsequent white win with just knight and bishop isn't totally impossible, it does look pretty unlikely (though of course, involving such a win might be exactly what makes the puzzle "difficult but beautiful"!).

And yet another awkward black move is e2, threatening promotion either on d1 or e1. Of course, the white bishop can respond by capturing that pawn, but then c2 looks on a pretty certain path to promotion.

Gengulphus

Gengulphus wrote:

Lootman wrote:As expressed the puzzle seems a bit ambiguous.

Some puzzles say "White to play and mate in n moves". The fact that yours says only "win" could be interpreted as getting to an obviously winning position rather than actually mating the black king.

It might be possible to come up with a number of variations where white can build a clearly winning position. But your use of the word "beautiful" indicates that there is an elegant mate in there somewhere. And in that case it is usual to state "mate" rather than "win". And to cite the number of moves to that mate, with best play by both sides.

"Play and win" puzzles are certainly much less common than "play and mate in n moves" puzzles, and are sometimes known as "studies" or "endgame studies" rather than "puzzles" - but they are a type of chess puzzle.

Another way to phrase a distinction is that there are puzzles that are specifically designed as a puzzle, often using a position that would be difficult or impossible to achieve in a real game. These typically lead to an elegant mate.

And then there are positions that are taken from real games where the reader is invited to determine the best play from there, which may lead to mate but may also just lead to an overwhelming material, tactical or positional advantage.

This position could happen in real play, but it is rather unusual in the configuration of the four passed pawns. In a game with a time limit I would probably play 1Ne3 and 2Bc2+ (or the other way around) to nullify the advanced black pawns, but I am fairly sure that isn't the solution because they are too obvious.

Damn, just lost a reply when I accidentally closed a browser tab!

Lootman wrote:This position could happen in real play, but it is rather unusual in the configuration of the four passed pawns. In a game with a time limit I would probably play 1Ne3 and 2Bc2+ (or the other way around) to nullify the advanced black pawns, but I am fairly sure that isn't the solution because they are too obvious.

That seems to me futile: it hands black a tempo, and with it the white passed pawn on a plate, whereupon black still has a material advantage and could seek to use it to promote another pawn. White's best hope then is a draw, assisted perhaps by opposite-colour bishops.

My instinct would be check, but that too is far too obvious. For example

N-f6 K-g7
N-h5 K-g6
N-f4 K-g7
N-e6, whereupon the obstacle to queening is exchanged away

These black moves are not forced but alternatives hand white a free tempo: queen with check, or check with the bishop. The latter either at h5 followed immediately by queening, or at c2 where it exerts power and prepares to bring the white king out of trouble and into play. However, at move 3 the black king can allow the c2 check and escape his corner, via 3 ... K-f5, so that line appears to peter out. So maybe instead

N-f6 K-g7
N-h5 K-g6
N-f4 K-f5
K-d5

But though that threatens immediate queening, it also hands black that tempo to block it with ... B-a5 or to do something more aggressive. So my instinctive line is no better than yours.

I wonder if the startling beauty could be that the pawn gets promoted to a knight rather than queen?

UncleEbenezer wrote:I wonder if the startling beauty could be that the pawn gets promoted to a knight rather than queen?

I had wondered that as well. If the black king can be driven to f7 then the white d pawn can advance and be promoted to a knight with check.

Then there might be a possible forced mate using a bishop and 2 knights, which would explain the cited 14 moves.

UncleEbenezer wrote:
N-f6 K-g7
N-h5 K-g6

Correct.

RC

If anyone is still interested...

3. Bc2 Kxh5

RC

ReformedCharacter wrote:If anyone is still interested...

OK, I'll bite.

3. Bc2 Kxh5

RC

Still eludes me. B-f5 is an obvious followup to that, but then ... N-f7 with check gives black time to follow with ... B-a5, and the queening is stone-dead while black's prospects for a queen remain very much alive. That leaves only(?) the unlikely immediate knighting of that pawn - which I know not how to notate - as move 4, whereupon black's choices include advancing the pawn to e2 in what looks like cruising to victory.

You're going to explain next how white sacrifices the pawn and the bishop

UncleEbenezer wrote:

3. Bc2 Kxh5

Still eludes me. B-f5 is an obvious followup to that, but then ... N-f7 with check gives black time to follow with ... B-a5, and the queening is stone-dead while black's prospects for a queen remain very much alive. That leaves only(?) the unlikely immediate knighting of that pawn - which I know not how to notate - as move 4, whereupon black's choices include advancing the pawn to e2 in what looks like cruising to victory.

You're going to explain next how white sacrifices the pawn and the bishop

Continuing the kamikaze theme, how about:

4. g4+ Kg4
5. Bg6

Now the black king is cut off and the white bishop's control of the f7 square assures that the white d pawn can advance to queendom.

But black still has 5 . . c4+ as well as an option to advance its c or e pawn. So the win still eludes me as well.

I've struggled

But I wonder if whites pawn becomes a bishop?

AiY

Lootman wrote:

UncleEbenezer wrote:

3. Bc2 Kxh5

Still eludes me. B-f5 is an obvious followup to that, but then ... N-f7 with check gives black time to follow with ... B-a5, and the queening is stone-dead while black's prospects for a queen remain very much alive. That leaves only(?) the unlikely immediate knighting of that pawn - which I know not how to notate - as move 4, whereupon black's choices include advancing the pawn to e2 in what looks like cruising to victory.

You're going to explain next how white sacrifices the pawn and the bishop

Continuing the kamikaze theme, how about:

4. g4+ Kg4
5. Bg6

Now the black king is cut off and the white bishop's control of the f7 square assures that the white d pawn can advance to queendom.

Nope. The black king is now under no pressure, and 5 ... B-a5 kills white's queening stone dead.

The bishops a rubbish thought

AiY

UncleEbenezer wrote:

Lootman wrote:

UncleEbenezer wrote:Still eludes me. B-f5 is an obvious followup to that, but then ... N-f7 with check gives black time to follow with ... B-a5, and the queening is stone-dead while black's prospects for a queen remain very much alive. That leaves only(?) the unlikely immediate knighting of that pawn - which I know not how to notate - as move 4, whereupon black's choices include advancing the pawn to e2 in what looks like cruising to victory.

You're going to explain next how white sacrifices the pawn and the bishop

Continuing the kamikaze theme, how about:

4. g4+ Kg4
5. Bg6

Now the black king is cut off and the white bishop's control of the f7 square assures that the white d pawn can advance to queendom.

Nope. The black king is now under no pressure, and 5 ... B-a5 kills white's queening stone dead.

Bugger, I missed 5 . . Ba5.

AsleepInYorkshire wrote:I've struggled

But I wonder if whites pawn becomes a bishop?

AiY

Surely a bishop has no advantage over a queen, except where stalemate is in play? We're a long way from that!

The point of the knight is that it has different capabilities. Notably being not-vulnerable to being forked by the black knight.