Paloma wrote: ↑Tue Jun 04, 2019 11:05 pm
Hi Marek,
what is exactly the difference between
forward 1 play,
forward 14 play,
play forward 1,
play forward 6
in your examples from above.
What means the number 1, 14 , 6
The number, after forward, means the count of skipped moves from PV-line returned earlier by searching.
It is your choise how many moves you want to skip.
I recommend to stop searching and to skip move(s), if it (they) satisfies you and to start searching from the new place. Next or instead of the new
searching you may start auto-playing, if you want to get the full solution.
The command forward goes to a new place definitely.
BTW. "forward 1 play" or "play forward 1" means the same.
lech wrote: ↑Tue Jun 04, 2019 12:38 pm
I see that the auto-playing is an interesting new tool and you like it.
One more example.
When your longanimity (searching) ends, you can go forward and start searching from the new place or (or next) can check the new position by auto-playing. You can always stop auto-playing and display diagram (d) or score (eval) and start auto-playing with new time per move or start searching at the new place.
bold = input lines
[d]8/5k2/6p1/4BP1p/6p1/p1p3PP/6P1/1n2K2R b K - 0 1 sting-sf-15
Sting 15 based on Stockfish 2.1.1 by Marek Kwiatkowski setoption name hash value 1024
setoption name clear hash value true
setoption name threads value 1
position fen 8/5k2/6p1/4BP1p/6p1/p1p3PP/6P1/1n2K2R b K - 0 1
go movetime 3600000
info depth 35 multipv 1 score cp 557 nodes 365201479 nps 736082 time 496142 pv a3a2 f5g6 f7g6 e5c3 b1c3 e1g1 c3b1 h3h4 a2a1q g1h2 a1a5 f1b1 g6f5 b1e1 a5c3 e1e7 c3c6 e7e1 c6c7 e1c1 c7a7 c1a1 a7h7 a1e1 f5f6 e1e8 h7f7 e8e1 f7g6 e1d1 f6e5 d1a1 g6f7 a1a5 e5d4 a5a4 d4e3 a4a5 e3f2 a5e5 f2f1 e5a5 f1e2
info depth 36 multipv 1 score cp 557 nodes 442224980 nps 729709 time 606029 pv a3a2 f5g6 f7g8 <click Ctrl keyboard button to stop solution>
stop
info depth 37 multipv 1 score cp 557 nodes 445023885 nps 723819 time 614827 pv a3a2 f5g6 f7g8 go movetime 60000 play forward 3
g6g7 a2a1q e1g1 a1a7 g1h2 a7c5 f1f8 c5f8 g7f8q g8f8 h2g1 f8f7 e5f4 b1d2 g1f2 d2b3 f2e2 c3c2 e2d3 c2c1q f4c1 b3c1 d3d2 g4h3 g2h3 c1b3 d2e3 f7g6 e3f4 b3d4 h3h4 d4e6 f4e5 e6c5 e5f4 c5d3 f4e4 d3f2 e4f3 f2g4 f3f4 g6f6 f4e4 g4e5 e4f4 e5g6 f4e4 f6e6 g3g4 h5g4 h4h5 g6e5 h5h6 e6f7 h6h7 f7g7 h7h8n g7h8 e4f4 h8g8 f4e3 g8g7 e3e2 g7g6 e2d1 g4g3 d1e2 g6f6 e2e1 f6f5 e1f1 e5d3 f1g1 f5f4 g1g2 f4g4 g2f1 g4f3 f1g1 g3g2 g1h2 f3f2 h2h3 g2g1q h3h4 g1g6 h4h3 g6g3 White is mated
What happened here?
The moves 1... a2 2. fxg6+ Kxg6 lead to fortress.
Black can avoid the fortress by 2... Kg8 but it leads to a very long difficult endgame.
Hi lech,how did you find 2....Kg8?
thank you
sorry,my PC is old and something was bad,after 5 minutes it found 2....Kg8
Maybe I'm missing something, but I think Kg8 in this position loses. With this this position I see a huge plus 10 score for white.
[d]6k1/8/6p1/4BP1p/6p1/p1p3PP/6P1/1n2K2R w K - 1 2
with some back and forth ...
lech wrote: ↑Tue Jun 04, 2019 12:38 pm
I see that the auto-playing is an interesting new tool and you like it.
One more example.
When your longanimity (searching) ends, you can go forward and start searching from the new place or (or next) can check the new position by auto-playing. You can always stop auto-playing and display diagram (d) or score (eval) and start auto-playing with new time per move or start searching at the new place.
bold = input lines
[d]8/5k2/6p1/4BP1p/6p1/p1p3PP/6P1/1n2K2R b K - 0 1 sting-sf-15
Sting 15 based on Stockfish 2.1.1 by Marek Kwiatkowski setoption name hash value 1024
setoption name clear hash value true
setoption name threads value 1
position fen 8/5k2/6p1/4BP1p/6p1/p1p3PP/6P1/1n2K2R b K - 0 1
go movetime 3600000
info depth 35 multipv 1 score cp 557 nodes 365201479 nps 736082 time 496142 pv a3a2 f5g6 f7g6 e5c3 b1c3 e1g1 c3b1 h3h4 a2a1q g1h2 a1a5 f1b1 g6f5 b1e1 a5c3 e1e7 c3c6 e7e1 c6c7 e1c1 c7a7 c1a1 a7h7 a1e1 f5f6 e1e8 h7f7 e8e1 f7g6 e1d1 f6e5 d1a1 g6f7 a1a5 e5d4 a5a4 d4e3 a4a5 e3f2 a5e5 f2f1 e5a5 f1e2
info depth 36 multipv 1 score cp 557 nodes 442224980 nps 729709 time 606029 pv a3a2 f5g6 f7g8 <click Ctrl keyboard button to stop solution>
stop
info depth 37 multipv 1 score cp 557 nodes 445023885 nps 723819 time 614827 pv a3a2 f5g6 f7g8 go movetime 60000 play forward 3
g6g7 a2a1q e1g1 a1a7 g1h2 a7c5 f1f8 c5f8 g7f8q g8f8 h2g1 f8f7 e5f4 b1d2 g1f2 d2b3 f2e2 c3c2 e2d3 c2c1q f4c1 b3c1 d3d2 g4h3 g2h3 c1b3 d2e3 f7g6 e3f4 b3d4 h3h4 d4e6 f4e5 e6c5 e5f4 c5d3 f4e4 d3f2 e4f3 f2g4 f3f4 g6f6 f4e4 g4e5 e4f4 e5g6 f4e4 f6e6 g3g4 h5g4 h4h5 g6e5 h5h6 e6f7 h6h7 f7g7 h7h8n g7h8 e4f4 h8g8 f4e3 g8g7 e3e2 g7g6 e2d1 g4g3 d1e2 g6f6 e2e1 f6f5 e1f1 e5d3 f1g1 f5f4 g1g2 f4g4 g2f1 g4f3 f1g1 g3g2 g1h2 f3f2 h2h3 g2g1q h3h4 g1g6 h4h3 g6g3 White is mated
What happened here?
The moves 1... a2 2. fxg6+ Kxg6 lead to fortress.
Black can avoid the fortress by 2... Kg8 but it leads to a very long difficult endgame.
Hi lech,how did you find 2....Kg8?
thank you
sorry,my PC is old and something was bad,after 5 minutes it found 2....Kg8
Maybe I'm missing something, but I think Kg8 in this position loses. With this this position I see a huge plus 10 score for white.
I think you are really missing something!
Try all the problem again!
MikeB wrote: ↑Fri Jun 07, 2019 8:17 am
Maybe I'm missing something, but I think Kg8 in this position loses. With this this position I see a huge plus 10 score for white.
[d]6k1/8/6P1/4BP1p/6p1/p1p3PP/6P1/1n2K2R w K - 1 2
Hi Mike,
the right position is after 1... a2 2. fxg6+ Kxg6 lead to fortress.
Black can avoid the fortress by 2... Kg8 but it leads to a very long difficult endgame. (lech)
lech wrote: ↑Wed Jun 05, 2019 9:58 am
The number, after forward, means the count of skipped moves from PV-line returned earlier by searching.
It is your choise how many moves you want to skip.
I recommend to stop searching and to skip move(s), if it (they) satisfies you and to start searching from the new place. Next or instead of the new searching you may start auto-playing, if you want to get the full solution.
The command forward goes to a new place definitely.
So is it like selective depth search (some) engines have, except with user control?
lech wrote: ↑Wed Jun 05, 2019 9:58 am
The number, after forward, means the count of skipped moves from PV-line returned earlier by searching.
It is your choise how many moves you want to skip.
I recommend to stop searching and to skip move(s), if it (they) satisfies you and to start searching from the new place. Next or instead of the new searching you may start auto-playing, if you want to get the full solution.
The command forward goes to a new place definitely.
So is it like selective depth search (some) engines have, except with user control?
A selective search, by external commands, for other engines helps to solve a position and it is only a trick in fact.
"Forward" and "auto-playing" help only to confirm and extend solution. The fact that "auto-playing" is to able to help to solve a position (a very long solution) it is only a question of the future. I am working at this and it looks very interesting.