The Halloween Chess Puzzle 2018

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
zullil
Posts: 5593
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: The Halloween Chess Puzzle 2018

Post by zullil » Thu Nov 01, 2018 10:12 am

peter wrote:
Wed Oct 31, 2018 10:07 pm
peter wrote:
Wed Oct 31, 2018 9:30 pm
Still would lay my bets rather on #37, any other findings from other mate- solvers?
:)
Will let it run a little longer with full hash and will come back later maybe.
And here we are again already:

Analysis by asmFishWCP_2018-10-28_popcnt:

1.e5+ Ke6 2.Sf4+ Kxe5 3.f7 Td4+ 4.Kb3 Txf4 5.Sxd7+ Kd5 6.f8D Txf8 7.Sxf8 c5 8.Sd7 c4+ 9.Kb2 c3+ 10.Kc2 Kc6 11.Sf6 Kc5 12.Kd3 a4 13.Se4+ Kc6 14.h4 a3 15.Sxc3 bxc3 16.Kxc3 a2 17.Kb2 a1D+ 18.Kxa1 Kd5 19.h5 Ke6 20.Kb2 Kf5 21.f4 Kf6 22.Kc3 Ke6 23.Kd4 Kf5 24.Kd5 Kf6 25.h6 Kg6 26.f5+ Kxh6 27.f6 Kg6 28.Ke6 Kg5 29.f7 Kf4 30.f8D+ Ke4 31.Df2 Kd3 32.Kd5 Kc3 33.De3+ Kc2 34.Kc4 Kd1 35.Df2 Kc1 36.Kc3 Kb1 37.Db2#
+- (#37) Tiefe: 61/76 00:25:10 79473MN
Hi Peter,

Here's a straight search using Stockfish-dev. Not yet seeing the #37:

info depth 79 seldepth 76 multipv 1 score mate 38 nodes 3052750541876 nps 95913130 hashfull 999 tbhits 0 time 31828286 pv e4e5 d6e6 g2f4 e6e5 f6f7 d2d4 c4b3 d4f4 b8d7 e5d5 f7f8q f4f8 d7f8 c7c5 b3b2 c5c4 f8d7 d5c6 d7f6 c4c3 b2c2 c6c5 c2d3 a5a4 f6e4 c5c6 e4c3 a4a3 d3c2 b4c3 c2c3 a3a2 c3b2 a2a1q b2a1 c6d6 f3f4 d6d7 h3h4 d7e6 h4h5 e6f6 a1b2 f6f5 b2c3 f5f6 c3d4 f6f5 d4e3 f5f6 e3e4 f6f7 e4e5 f7f8 e5e6 f8g7 f4f5 g7h6 f5f6 h6h5 f6f7 h5g4 f7f8q g4g3 f8f1 g3g4 f1e2 g4g5 e2e3 g5h5 e6f5 h5h4 e3a3 h4h5 a3h3

Louis

peter
Posts: 1775
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: The Halloween Chess Puzzle 2018

Post by peter » Thu Nov 01, 2018 10:19 am

zullil wrote:
Thu Nov 01, 2018 10:12 am
Here's a straight search using Stockfish-dev. Not yet seeing the #37:

info depth 79 seldepth 76 multipv 1 score mate 38
Let's not have too much argueing about the single one move, Louis :)

I had long Backward before standalone- computing and don't think SF (or asmFish) is wrong as for too short lines, rather for too long ones most of the times, but I don't plan to try it all over again with MateFinder neither.

Even less because it doesn't matter at all as for a study other than mate in x- problem, which this one isn't.
But thanks for trying too.
Peter.

ernest
Posts: 1853
Joined: Wed Mar 08, 2006 7:30 pm

Re: The Halloween Chess Puzzle 2018

Post by ernest » Thu Nov 01, 2018 1:11 pm

peter wrote:
Thu Nov 01, 2018 6:34 am

https://github.com/Counterply/asmFish/t ... CounterPly
As for my personal pov, asmFish is much further up to date to SF dev. now, than it was in June with Antonov's latest versions.
Yes Peter I know, we can only be thankful to Counterplay for sustaining life to asmFish,
but I feel uneasy when I read in the github :

Updated: 10/22/18. Corresponds approximately to Stockfish 220718. https://www.counterply.com/asmFishCP

peter
Posts: 1775
Joined: Sat Feb 16, 2008 6:38 am
Full name: Peter Martan

Re: The Halloween Chess Puzzle 2018

Post by peter » Thu Nov 01, 2018 4:13 pm

ernest wrote:
Thu Nov 01, 2018 1:11 pm
but I feel uneasy when I read in the github :

Updated: 10/22/18. Corresponds approximately to Stockfish 220718. https://www.counterply.com/asmFishCP
Of course it would be somewhat finer, to know all patches of SF dev being in asmFish already too, but I'm used to see each and every branch as something of its own anyhow.
And yes, I'm happy too, to see the development continued at all.
Peter.

zullil
Posts: 5593
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: The Halloween Chess Puzzle 2018

Post by zullil » Thu Nov 01, 2018 8:57 pm

peter wrote:
Thu Nov 01, 2018 10:19 am
zullil wrote:
Thu Nov 01, 2018 10:12 am
Here's a straight search using Stockfish-dev. Not yet seeing the #37:

info depth 79 seldepth 76 multipv 1 score mate 38
Let's not have too much argueing about the single one move, Louis :)

I had long Backward before standalone- computing and don't think SF (or asmFish) is wrong as for too short lines, rather for too long ones most of the times, but I don't plan to try it all over again with MateFinder neither.

Even less because it doesn't matter at all as for a study other than mate in x- problem, which this one isn't.
But thanks for trying too.
No argument from me!

info depth 84 seldepth 74 multipv 1 score mate 37 nodes 7973601091763 nps 101958442 hashfull 999 tbhits 0 time 78204422 pv e4e5 d6e6 g2f4 e6e5 f6f7 d2d4 c4b3 d4f4 b8d7 e5d5 f7f8q f4f8 d7f8 c7c5 f8d7 c5c4 b3b2 d5c6 d7f6 c4c3 b2c2 c6c5 c2d3 a5a4 f6e4 c5c6 e4c3 a4a3 f3f4 b4c3 d3c3 a3a2 c3b2 a2a1q b2a1 c6d6 h3h4 d6e6 a1a2 e6f6 h4h5 f6f5 a2b2 f5f6 b2c2 f6f5 c2d3 f5e6 h5h6 e6f6 f4f5 f6f7 d3d4 f7f6 d4e4 f6f7 e4e5 f7f8 f5f6 f8g8 e5e6 g8h7 f6f7 h7h6 f7f8q h6g5 f8f3 g5h4 f3g2 h4h5 e6f5 h5h6 g2g6

Post Reply