2025/08/15 DPA: "Orthogonal Action, Diagonal Finish"

Black to move:

.

Black has a Queen + Bishop battery along the a8-h1 diagonal but the White Queen guards g2.

One thing I see that does NOT work:  1. ... Qxg2+. 2. Qxg2 Bxg2+. 3. Kxg2 gxf6 [pawn is longer pinned and so can capture the Rook]. 4. Bh7+ Kxh7. 5. Rxd8.

Wait:  that's wrong.  Black's Rooks are still doubled and thus protected.

But even so, Black only wins the exchange and a pawn but began down a piece so this is not winning.

Black's Rook battery can't do much as long as the White Bishop is on d3.

We'd like to remove the Queen from guarding g2.  Can Black create a threat that drives the Queen away?

1. ... Rg4 doesn't work since 2. Qxg4 maintains protection of g2.

No BQMs [Backward Queen Moves].

Moving the Queen forwards, either vertically or diagonally, doesn't do anything.

1. ... Qc3 threatens the Knight and if 2. Bh7+ Kxh7  3. Qxc3 Rxd1+. 4. Qe1 Rxe1#.

But 2. Rf1 defends the back rank.  2. ... Qxa3. 3. Bh7+ Kxh7...oh, but Black still has doubled Rooks!  So Black wins a piece.

4. Rxd4 Qxg3. 5. hxg3 Rxd4.

I don't see anything else that produces material gain.

What if 2. Nb1?  2. ... Qxc2 and if 3. Bxc2 Rxd1+, leading to checkmate.

But White doesn't have to capture:  he could play 3. Rf1 but now his Bishop is hanging.  And if 3. Bh7+ Kxh7. 4. Rxd4 Qd1+.

No, I was wrong.

.

The key was the simultaneous checkmate threats against g2 if the Queen leaves and the back rank if the Rook leaves.

1. ... Rxd3 captures a Bishop and encourages recapture.

2. ... Rxd3 threatens the Queen and Rook.  There are ways to avoid checkmate but at the loss of the Queen.

3. ... Qxg2#.

.

My error was two-fold:

• I saw the possibility of driving the Queen away from guarding g2 but didn't consider the Rook sac.
• My 1. ... Qc3 idea was riddled with flaws, one of them being Rxf7, which leads to checkmate if Black accepts, rather than Bh7+.  But there were others too.  A sloppy analytical job.