1m+2s Man versus machine at 173.110.000 speed. Pablo Ignacio Restrepo versus super engine running at 173.110KN/s draw.

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
Father
Posts: 1271
Joined: Sun Mar 19, 2006 3:39 am
Location: Colombia
Full name: Pablo Ignacio Restrepo

1m+2s Man versus machine at 173.110.000 speed. Pablo Ignacio Restrepo versus super engine running at 173.110KN/s draw.

Post by Father » Sat Oct 12, 2019 3:43 am


Again I wish to thank the super chess computer operator for allowing me to play chess in my capacity as a human being against such a portentous supercomputer engineering.


I would never have the opportunity to test and measure the pulse and strength as chess player playing against such a quality chess machine.

Since childhood my parents taught me to be grateful. And that is what I wish to express at this moment: "Thank you with all my heart for allowing me to fight against such a wonderful machine with the Chinese flag."


I am aware that I lack every chess title and that my only color and heart is to fight to learn to fight against the machines in the rest of my life.

This example of a man versus machine chess game has two special features: 1o It was played at minute and; 2o with increase time per play of two seconds.

All those who demerit and falsify my achievements against computers without scientific foundation, it is time for them to keep silent.

Facts and events are the best and main spiritual and material tests that life and exitence can provide.

Best regards,

Pablo Ignacio Restrepo
Father and grand-father playchess user

Ping Grand-Father = 1.0s Rated (Classic) 1m+2s No. of players receiving your offer: 4 Rated (Classic) 1m+2s No. of players receiving your offer: 4 Rated (Classic) 1m+2s Grand-Father - Noobpwnftw ½-1½ New Elo: 2000 (+16), Classic. BookFish (8 cores): 41.8 plies; 173.110kN/s Intel(R) Xeon(R) Silver 4116 CPU @ 2.10GHz 2095MHz.

I am thinking chess is in a coin.Human beings for ever playing in one face.Now I am playing in the other face:"Antichess". Computers are as a fortres where owner forgot to close a little door behind. You must enter across this door.Forget the front.

lkaufman
Posts: 3683
Joined: Sun Jan 10, 2010 5:15 am
Location: Maryland USA
Contact:

Re: 1m+2s Man versus machine at 173.110.000 speed. Pablo Ignacio Restrepo versus super engine running at 173.110KN/s dr

Post by lkaufman » Sat Oct 12, 2019 6:08 am

I was a bit curious to see whether this way of playing for a draw would work against Komodo or Lc0, or whether it just works against versions of Stockfish. So I looked at the key mistakes where Stockfish allowed the game to become more and more hopelessly blocked up. The most obviously ridiculous move was 20...a4??, which probably no human player rated over 1800 or so would play, as it obviously makes the win far more difficult. I confirmed that the latest Stockfish does indeed like this ridiculous move. Then I checked it on Komodo, and regardless of whether MCTS is on or off or whether high contempt is set or not, it never shows this move. Same with various versions of Lc0, they all immediately see it as a clear mistake. Then the next questionable move is 23...g4?, which at least makes some sense (unlike 20...a4??) but blocks the game too much. Here Komodo generally avoids it but may choose it depending on time, Contempt, and MCTS setting; Stockfish just loves it, Lc0 avoids it. Then very late in the game ...Nxg3? pretty much guarantees the draw; both Komodo and Lc0 (but not Stockfish) recognize that Black should try sacrificing a pawn with b4 to bring a knight to b5 and aim for a pawn break. Whether that wins or not, it is the best try. So the short answer is that I believe you will never get a draw against either Komodo or Lc0 with your strategy (assuming no games terminated due to lag), but you have demonstrated that although Stockfish may outrate Komodo and will defeat it in a match, it does not understand that if a position is too blocked up, it will end in a draw. Both Komodo and Lc0 don't seem to have this problem.
Komodo rules!

Post Reply