Richter Position 14

Peter Grayson · Post by **Peter Grayson** » 02/08/2021, 18:47

Finally got a solution to the Richter test position 14.

An interesting history with the original key move given as 1.Qf3 that may initially make the viewer sit up and wonder what is going on. Earlier old school engines found this move early on but as the engines became faster and stronger, there was a migration away to a different move that was 1.Qd8
However, analysis engines often chose 1.Qf3 whereas game engines stuck with 1.Qd8+"
I had never seen a definitive solution but subsequently included 1.Qd8+ as an alternative solution because it clearly obtained a winning advantage.

Now however allowing Stockfish 14 to run through in analysis until a mating solution was found revealed all of the engines choosing 1.Qd8 were wrong and the fastest route to mate was by 1.Qf3 ... unless anyone can provide alternative analyis?

One move in it. Should 1.Qd8 stay or be classed as a failure?

1.Qf3 leads to #26

1.Qd8+ leads to #27

peterm · Post by **peterm** » 03/08/2021, 5:39

Peter Grayson wrote: ↑02/08/2021, 18:47 Should 1.Qd8 stay or be classed as a failure?

Backwarding your lines with SlowChess (has a very fine mate search of its own, hasn't it?) brings deviation 3...Re1 instead of ...Rc1 in your 1.Qf3- line, thus giving DTM 27 to both moves at starting position then in MultiPV=2 mode:

1k6/p7/1p3P2/3QP3/8/4q3/6PP/2r2R1K w - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1. +- (#27): 1.Qd8+ Kb7 2.Qd7+ Kb8 3.Qf5 Qe1 4.Kg1 Qe3+ 5.Qf2 Rxf1+ 6.Kxf1 Qc3 7.f7 Qc1+ 8.Ke2 Qc4+ 9.Ke3 Qc1+ 10.Kf3 Qc6+ 11.Kf4 Qh6+ 12.Ke4 Qf8 13.Qf6 Kb7 14.Kf5 Ka6 15.Kg6 Qa8 16.Qf1+ b5 17.f8Q Qe4+ 18.Kf6 Qh4+ 19.Ke6 Qg4+ 20.Kd5 Qd7+ 21.Qd6+ Qxd6+ 22.exd6 Kb6 23.d7 Ka5 24.d8Q+ Ka4 25.Qa1+ Kb3 26.Kc5 Kc2 27.Qdd1#

2. +- (#27): 1.Qf3 Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Rxe5 6.Qg7+ Kd6 7.h4 Re7 8.Qf6+ Re6 9.Qf8+ Kd7 10.h5 Re8 11.Qg7+ Ke6 12.h6 a5 13.h7 a4 14.Qg6+ Ke7 15.Qxe8+ Kxe8 16.h8Q+ Kd7 17.f4 a3 18.Qh3+ Kd6 19.Qxa3+ Kd7 20.f5 b5 21.Qa7+ Kd6 22.f6 Kd5 23.f7 Kc4 24.f8Q b4 25.Qe3 Kb5 26.Qec5+ Ka4 27.Qxb4#

BTW e.g 1.Qd6+ isn't very much longer to mate, after Forward- Backward with this one:

1k6/p7/1p1Q1P2/4P3/8/4q3/6PP/2r2R1K b - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1...Kc8 2.Qe6+ Kd8 3.Qd5+ Kc7 4.Qf3 Re1 5.f7 Qxf3 6.gxf3 Rxf1+ 7.Kg2 Re1 8.f8Q Rxe5 9.Qg7+ Kd6 10.h4 Re2+ 11.Kg3 Re7 12.Qf8 Kd7 13.h5 Re8 14.Qg7+ Kc6 15.h6 b5 16.h7 b4 17.Qg6+ Kd7 18.Qxe8+ Kxe8 19.h8Q+ Kd7 20.Qd4+ Kc6 21.Qxb4 a6 22.Kf4 Kd5 23.Kf5 a5 24.Qa4 Kc5 25.Ke5 Kb6 26.Kd6 Kb7 27.Qxa5 Kb8 28.Kc6 Kc8 29.Qa8#
+- (#28) Depth: 27/128 00:01:09 982MN, tb=26926168

peterm · Post by **peterm** » 03/08/2021, 5:52

peterm wrote: ↑03/08/2021, 5:39
Peter Grayson wrote: ↑02/08/2021, 18:47 Should 1.Qd8 stay or be classed as a failure?
Backwarding your lines with SlowChess (has a very fine mate search of its own, hasn't it?) brings deviation 3...Re1 instead of ...Rc1 in your 1.Qf3- line, thus giving DTM 27 to both moves at starting position then in MultiPV=2 mode:

1k6/p7/1p3P2/3QP3/8/4q3/6PP/2r2R1K w - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1. +- (#27): 1.Qd8+ Kb7 2.Qd7+ Kb8 3.Qf5 Qe1 4.Kg1 Qe3+ 5.Qf2 Rxf1+ 6.Kxf1 Qc3 7.f7 Qc1+ 8.Ke2 Qc4+ 9.Ke3 Qc1+ 10.Kf3 Qc6+ 11.Kf4 Qh6+ 12.Ke4 Qf8 13.Qf6 Kb7 14.Kf5 Ka6 15.Kg6 Qa8 16.Qf1+ b5 17.f8Q Qe4+ 18.Kf6 Qh4+ 19.Ke6 Qg4+ 20.Kd5 Qd7+ 21.Qd6+ Qxd6+ 22.exd6 Kb6 23.d7 Ka5 24.d8Q+ Ka4 25.Qa1+ Kb3 26.Kc5 Kc2 27.Qdd1#

2. +- (#27): 1.Qf3 Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Rxe5 6.Qg7+ Kd6 7.h4 Re7 8.Qf6+ Re6 9.Qf8+ Kd7 10.h5 Re8 11.Qg7+ Ke6 12.h6 a5 13.h7 a4 14.Qg6+ Ke7 15.Qxe8+ Kxe8 16.h8Q+ Kd7 17.f4 a3 18.Qh3+ Kd6 19.Qxa3+ Kd7 20.f5 b5 21.Qa7+ Kd6 22.f6 Kd5 23.f7 Kc4 24.f8Q b4 25.Qe3 Kb5 26.Qec5+ Ka4 27.Qxb4#

To late to edit: yet #26 after 1.Qf3 3...Re1 too:

1k6/p7/1p3P2/4P3/8/4qQ2/6PP/2r2R1K b - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1...Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Re2+ 6.Kg3 Rxe5 7.Qg7+ Kd6 8.h4 Re7 9.Qf8 Kd7 10.h5 Re8 11.Qg7+ Re7 12.h6 a5 13.Qxe7+ Kxe7 14.h7 a4 15.h8Q a3 16.Qa8 a2 17.Qxa2 Kd6 18.f4 Kc5 19.f5 Kd6 20.f6 b5 21.f7 Ke7 22.Qa7+ Kf8 23.Qd7 b4 24.Qe8+ Kg7 25.f8Q+ Kh7 26.Qef7#
+- (#25) Depth: 29/128 00:00:57 792MN, tb=20150964

I really love SlowChess, here's another one, that bites the dust (1.Qg8+)

1k6/p7/1p3P2/3QP3/8/4q3/6PP/2r2R1K w - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1. +- (#26): 1.Qf3 Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Rxe5 6.Qg7+ Kd6 7.h4 Re2+ 8.Kg3 Re7 9.Qf8 Kd7 10.h5 Re8 11.Qg7+ Kc6 12.h6 b5 13.h7 b4 14.h8Q Rxh8 15.Qxh8 b3 16.f4 Kc5 17.f5 Kd5 18.Qb8 a6 19.Qxb3+ Ke5 20.Qe6+ Kd4 21.f6 Kc3 22.f7 Kb2 23.f8Q Kc2 24.Qc4+ Kd2 25.Qf2+ Kd1 26.Qcc2#

2. +- (#27): 1.Qd8+ Kb7 2.Qd7+ Kb8 3.Qf5 Qe1 4.Kg1 Qe3+ 5.Qf2 Rxf1+ 6.Kxf1 Qc3 7.f7 Qc1+ 8.Ke2 Qc4+ 9.Ke3 Qb3+ 10.Kd4 Qb4+ 11.Kd5 Qb3+ 12.Kd6 Qb4+ 13.Kd7 Qb5+ 14.Ke6 Qc4+ 15.Kf6 Qb4 16.Kg6 Qf8 17.e6 Kb7 18.Qf4 Qe7 19.Qf5 Qf8 20.Qe4+ Ka6 21.e7 Qb8 22.f8Q Qd6+ 23.Qf6 Qxf6+ 24.Kxf6 Ka5 25.e8Q b5 26.Qd8+ Ka6 27.Qc6#

3. +- (#28): 1.Qd6+ Ka8 2.Qd8+ Kb7 3.Qd7+ Kb8 4.Qf5 Qe1 5.Kg1 Qe3+ 6.Qf2 Rxf1+ 7.Kxf1 Qc3 8.f7 Qc1+ 9.Ke2 Qc4+ 10.Ke3 Qb3+ 11.Kd4 Qb4+ 12.Kd5 Qb3+ 13.Kd6 Qb4+ 14.Kd7 Qb5+ 15.Ke6 Qc4+ 16.Kf6 Qb4 17.Kg6 Qf8 18.e6 Kb7 19.Qf4 Qe7 20.Qf5 Qf8 21.Qe4+ Ka6 22.e7 Qb8 23.f8Q Qd6+ 24.Qf6 Qxf6+ 25.Kxf6 Ka5 26.e8Q b5 27.Qd8+ Ka6 28.Qc6#

4. +- (#29): 1.Qg8+ Kc7 2.Qh7+ Kd8 3.Qe7+ Kc8 4.Qe6+ Kd8 5.Qd5+ Kc7 6.Qf3 Kd8 7.Qa8+ Kd7 8.Qb7+ Kd8 9.Qb8+ Kd7 10.e6+ Kxe6 11.Qe8+ Kd6 12.Qd8+ Ke6 13.Qe7+ Kd5 14.Qd7+ Ke5 15.Qf5+ Kd4 16.f7 Rxf1+ 17.Qxf1 Qh6 18.f8Q Qxf8 19.Qxf8 b5 20.h4 b4 21.Qxb4+ Ke5 22.h5 Kf6 23.h6 Kg6 24.Qd6+ Kf7 25.Qd7+ Kg6 26.h7 a5 27.h8Q Kg5 28.Qe5+ Kg6 29.Qdg7#

5. = (0.02): 1.Qb5 Qf4 2.Qe8+ Kc7 3.Qf7+ Kc8 4.Qf8+ Kb7 5.Qe7+ Kc8 6.Qe6+ Kc7 7.Qd6+ Kc8 8.Qf8+

Peter Grayson · Post by **Peter Grayson** » 03/08/2021, 19:42

peterm wrote: ↑03/08/2021, 5:52
peterm wrote: ↑03/08/2021, 5:39
Peter Grayson wrote: ↑02/08/2021, 18:47 Should 1.Qd8 stay or be classed as a failure?
Backwarding your lines with SlowChess (has a very fine mate search of its own, hasn't it?) brings deviation 3...Re1 instead of ...Rc1 in your 1.Qf3- line, thus giving DTM 27 to both moves at starting position then in MultiPV=2 mode:

1k6/p7/1p3P2/3QP3/8/4q3/6PP/2r2R1K w - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1. +- (#27): 1.Qd8+ Kb7 2.Qd7+ Kb8 3.Qf5 Qe1 4.Kg1 Qe3+ 5.Qf2 Rxf1+ 6.Kxf1 Qc3 7.f7 Qc1+ 8.Ke2 Qc4+ 9.Ke3 Qc1+ 10.Kf3 Qc6+ 11.Kf4 Qh6+ 12.Ke4 Qf8 13.Qf6 Kb7 14.Kf5 Ka6 15.Kg6 Qa8 16.Qf1+ b5 17.f8Q Qe4+ 18.Kf6 Qh4+ 19.Ke6 Qg4+ 20.Kd5 Qd7+ 21.Qd6+ Qxd6+ 22.exd6 Kb6 23.d7 Ka5 24.d8Q+ Ka4 25.Qa1+ Kb3 26.Kc5 Kc2 27.Qdd1#

2. +- (#27): 1.Qf3 Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Rxe5 6.Qg7+ Kd6 7.h4 Re7 8.Qf6+ Re6 9.Qf8+ Kd7 10.h5 Re8 11.Qg7+ Ke6 12.h6 a5 13.h7 a4 14.Qg6+ Ke7 15.Qxe8+ Kxe8 16.h8Q+ Kd7 17.f4 a3 18.Qh3+ Kd6 19.Qxa3+ Kd7 20.f5 b5 21.Qa7+ Kd6 22.f6 Kd5 23.f7 Kc4 24.f8Q b4 25.Qe3 Kb5 26.Qec5+ Ka4 27.Qxb4#

To late to edit: yet #26 after 1.Qf3 3...Re1 too:

1k6/p7/1p3P2/4P3/8/4qQ2/6PP/2r2R1K b - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1...Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Re2+ 6.Kg3 Rxe5 7.Qg7+ Kd6 8.h4 Re7 9.Qf8 Kd7 10.h5 Re8 11.Qg7+ Re7 12.h6 a5 13.Qxe7+ Kxe7 14.h7 a4 15.h8Q a3 16.Qa8 a2 17.Qxa2 Kd6 18.f4 Kc5 19.f5 Kd6 20.f6 b5 21.f7 Ke7 22.Qa7+ Kf8 23.Qd7 b4 24.Qe8+ Kg7 25.f8Q+ Kh7 26.Qef7#
+- (#25) Depth: 29/128 00:00:57 792MN, tb=20150964

I really love SlowChess, here's another one, that bites the dust (1.Qg8+)

1k6/p7/1p3P2/3QP3/8/4q3/6PP/2r2R1K w - - 0 1

Analysis by SlowChess Blitz 2.6 sse:

1. +- (#26): 1.Qf3 Qxf3 2.gxf3 Rxf1+ 3.Kg2 Re1 4.f7 Kc7 5.f8Q Rxe5 6.Qg7+ Kd6 7.h4 Re2+ 8.Kg3 Re7 9.Qf8 Kd7 10.h5 Re8 11.Qg7+ Kc6 12.h6 b5 13.h7 b4 14.h8Q Rxh8 15.Qxh8 b3 16.f4 Kc5 17.f5 Kd5 18.Qb8 a6 19.Qxb3+ Ke5 20.Qe6+ Kd4 21.f6 Kc3 22.f7 Kb2 23.f8Q Kc2 24.Qc4+ Kd2 25.Qf2+ Kd1 26.Qcc2#

2. +- (#27): 1.Qd8+ Kb7 2.Qd7+ Kb8 3.Qf5 Qe1 4.Kg1 Qe3+ 5.Qf2 Rxf1+ 6.Kxf1 Qc3 7.f7 Qc1+ 8.Ke2 Qc4+ 9.Ke3 Qb3+ 10.Kd4 Qb4+ 11.Kd5 Qb3+ 12.Kd6 Qb4+ 13.Kd7 Qb5+ 14.Ke6 Qc4+ 15.Kf6 Qb4 16.Kg6 Qf8 17.e6 Kb7 18.Qf4 Qe7 19.Qf5 Qf8 20.Qe4+ Ka6 21.e7 Qb8 22.f8Q Qd6+ 23.Qf6 Qxf6+ 24.Kxf6 Ka5 25.e8Q b5 26.Qd8+ Ka6 27.Qc6#

3. +- (#28): 1.Qd6+ Ka8 2.Qd8+ Kb7 3.Qd7+ Kb8 4.Qf5 Qe1 5.Kg1 Qe3+ 6.Qf2 Rxf1+ 7.Kxf1 Qc3 8.f7 Qc1+ 9.Ke2 Qc4+ 10.Ke3 Qb3+ 11.Kd4 Qb4+ 12.Kd5 Qb3+ 13.Kd6 Qb4+ 14.Kd7 Qb5+ 15.Ke6 Qc4+ 16.Kf6 Qb4 17.Kg6 Qf8 18.e6 Kb7 19.Qf4 Qe7 20.Qf5 Qf8 21.Qe4+ Ka6 22.e7 Qb8 23.f8Q Qd6+ 24.Qf6 Qxf6+ 25.Kxf6 Ka5 26.e8Q b5 27.Qd8+ Ka6 28.Qc6#

4. +- (#29): 1.Qg8+ Kc7 2.Qh7+ Kd8 3.Qe7+ Kc8 4.Qe6+ Kd8 5.Qd5+ Kc7 6.Qf3 Kd8 7.Qa8+ Kd7 8.Qb7+ Kd8 9.Qb8+ Kd7 10.e6+ Kxe6 11.Qe8+ Kd6 12.Qd8+ Ke6 13.Qe7+ Kd5 14.Qd7+ Ke5 15.Qf5+ Kd4 16.f7 Rxf1+ 17.Qxf1 Qh6 18.f8Q Qxf8 19.Qxf8 b5 20.h4 b4 21.Qxb4+ Ke5 22.h5 Kf6 23.h6 Kg6 24.Qd6+ Kf7 25.Qd7+ Kg6 26.h7 a5 27.h8Q Kg5 28.Qe5+ Kg6 29.Qdg7#

5. = (0.02): 1.Qb5 Qf4 2.Qe8+ Kc7 3.Qf7+ Kc8 4.Qf8+ Kb7 5.Qe7+ Kc8 6.Qe6+ Kc7 7.Qd6+ Kc8 8.Qf8+

Thanks for the additional mate information.

I am always wary using backward analysis with Stockfish based engines because some strange outputs can often be seen once the distance is 20 moves (40 ply) or more away from the mate. It may be an issue with the hash holding.

Historically there were only two moves considered by engines therefore I ran Stockfish from the start with 2 x Multi-PV. 1.Qf3 showed #40 at 52 ply and both were recording there respective final mate depths by 55 primary ply depth therefore I decided if there was no further change in depth to mate after an additional 8 ply depth it was unlikely there would be any further improvement. Hence both moves were analysed by the engine from the start through to 63 ply depth.

Peter

Outskirts CheSS ForuM

Outskirts CheSS ForuM