2024/04/06 DPA: "Slicing The Gordian Knot"
White to move.
White wants to play Rc8 but Black's Queen and Bishop are guarding c8.
Black has 4 attackers on the Knight.
Black has dark square weaknesses, namely f8 and g7. Black has a dark-squared Bishop but it's currently not defending f8.
The Bishop on h3 could attack if the Knight moved.
1. Nxe7+ Qxe7 removes one defender of c8. It also exposes White's Queen to Black's Rook.
1. Rc8+ Qxc8 2. Nxe7#. So Black cannot take the Rook but must block check instead: 1. ... Ne8. So maybe White eliminates the Knight first?
1. Qxf6 Bxf6 frees up h8 for the King and also guards e7.
1. Rc8+ Ne8 2. Nxe7+ Qxe7 3. Qxe7
- A) If 3. ... Bxc8 4. Qxe8#
- B) If 3. ... Re1+ 4. Bf1, avoiding 4. Kg2 Bxh3+, winning back the Queen and protecting the Knight.
- C) If 3. ... Bxh3, discovering an attack on the Queen by the Rook, 4. Rxe8#; or 4. Qf8#
This analysis was correct: the puzzle chose 1. ... Qxc8. I thought line B was the most interesting and it gave White a chance to blunder with 4. Kg2??. But the total solution would have been 8 moves, which would have taken the spotlight away from the ideas of the puzzle.
The keys were recognizing:
- Black's back rank weakness
- Black's King's lack of mobility
- The Overloaded Defender Black Queen [guarding both c8 and e7]
- Visualizing that, after 1. ... Ne8, the White Queen now attacks e7 also
- The move order [1. Nxe7 followed by 2. Rc8+ would have failed]
.
This will unleash the usual firestorm of protest that the puzzle is wrong or the puzzle is bad or the puzzle creator doesn't know what he's doing, etc. These claims are false: puzzles may contain sub-optimal moves by the opponent [but not the solver]. It's at the discretion of the creator, based on how best to showcase the idea.
https://www.chess.com/blog/EnPassantFork/no-the-puzzle-is-not-wrong