I got the impression the players did not want to hear engine scores, so I'll refrain from providing one. But, let me just say that I would be quite surprised if this is not a win.Dann Corbit wrote: ↑Sun May 17, 2020 4:56 am That's why I couch my language with things like "almost certainly drawn". I have seen SF show a score of zero from ply 30 to ply 113 and then an ply 114 fail high for an hour and end up with mate in 35.
So, though it is pretty rare, I never say never. Oops. I just said it, twice.
Zenmastur vs Ovyron No it's not another GROB II!
Moderators: hgm, Rebel, chrisw
-
- Posts: 29
- Joined: Sun Feb 16, 2020 11:13 pm
- Full name: Drew D. Rue
Re: Zenmastur vs Ovyron No it's not another GROB II!
-
- Posts: 12541
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Zenmastur vs Ovyron No it's not another GROB II!
you put a bit of fear into my heart, because I only gave it two hours on a 3970x and I know you have the hardware dreams are made of. More than a billion tablebase hits, so the engine spent quite a bit of time talking to God, but SF does have an unnecessarily powerful draw magnet.drewdrew wrote: ↑Sun May 17, 2020 7:46 amI got the impression the players did not want to hear engine scores, so I'll refrain from providing one. But, let me just say that I would be quite surprised if this is not a win.Dann Corbit wrote: ↑Sun May 17, 2020 4:56 am That's why I couch my language with things like "almost certainly drawn". I have seen SF show a score of zero from ply 30 to ply 113 and then an ply 114 fail high for an hour and end up with mate in 35.
So, though it is pretty rare, I never say never. Oops. I just said it, twice.
At any rate, this is a lot more interesting than the grob match, which was already interesting
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 6442
- Joined: Tue Jan 09, 2007 12:31 am
- Location: PA USA
- Full name: Louis Zulli
Re: Zenmastur vs Ovyron No it's not another GROB II!
I don't have drewdrew's hardware either, but I also think Black can win from the current position. That said, Black can also draw from the current position, by playing imprecisely. So I think this remains a nice test of Ovyron's hardware and methodology. As far as I know, he has passed all the previous tests, so I'm actually rooting for him.Dann Corbit wrote: ↑Sun May 17, 2020 8:06 amyou put a bit of fear into my heart, because I only gave it two hours on a 3970x and I know you have the hardware dreams are made of. More than a billion tablebase hits, so the engine spent quite a bit of time talking to God, but SF does have an unnecessarily powerful draw magnet.drewdrew wrote: ↑Sun May 17, 2020 7:46 amI got the impression the players did not want to hear engine scores, so I'll refrain from providing one. But, let me just say that I would be quite surprised if this is not a win.Dann Corbit wrote: ↑Sun May 17, 2020 4:56 am That's why I couch my language with things like "almost certainly drawn". I have seen SF show a score of zero from ply 30 to ply 113 and then an ply 114 fail high for an hour and end up with mate in 35.
So, though it is pretty rare, I never say never. Oops. I just said it, twice.
At any rate, this is a lot more interesting than the grob match, which was already interesting
Depth 94 is not very deep; suppose Black wins here (hypothetically) by forcing an 11-man position that is a "tablebase win" that requires 77 moves. Would Stockfish see that from the root at depth 94 ply, even with existing tablebases?
-
- Posts: 12541
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Zenmastur vs Ovyron No it's not another GROB II!
possibly accidentally because SF does have some extensions. Also, there is qsearch, but I do not remember if SF probes in qsearch. On the other hand, it could be in a branch of the tree that is heavily pruned and it may take even beyond SF's maximum depth to find it from the current root.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 10303
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Zenmastur vs Ovyron No it's not another GROB II!
The final result of the game is not going to prove some claims of ovyron even if ovyron is going to win the game
1)"by your own admission you wouldn't know to win with f4"
Zenmastur did not admit it.
zenmastur only admitted not finding a win at some time.
facts are the following:
a)ovyron played f4 in 9.5 2:13 pm
b)zenmastur posted in 13.5 7:10 am
"I am not 100% sure just yet but I think you gave me a draw by playing f4"
time between the events is only 88 hours and 57 minutes.
Not finding a win in less than 89 hours is not equivalent to not finding a win practically in a correspondence game
2)"If anything it's show the uselessness of big depth and how unassisted engines can't win positions like these by themselves"
It is related to the possibility that Ovyron wins the game.
Even if we assume Ovyron is going to win the game it does not prove that unassisted engines can't win positions like these by themselves when they get significantly more than 88 hours and 57 minutes and it is possible to give them significantly more time in a correspondence game.
Maybe they can win it by themselves but zenmastur did not give them enough time at the time of posting and maybe the correct engine that zen did not use can win it by itself.
1)"by your own admission you wouldn't know to win with f4"
Zenmastur did not admit it.
zenmastur only admitted not finding a win at some time.
facts are the following:
a)ovyron played f4 in 9.5 2:13 pm
b)zenmastur posted in 13.5 7:10 am
"I am not 100% sure just yet but I think you gave me a draw by playing f4"
time between the events is only 88 hours and 57 minutes.
Not finding a win in less than 89 hours is not equivalent to not finding a win practically in a correspondence game
2)"If anything it's show the uselessness of big depth and how unassisted engines can't win positions like these by themselves"
It is related to the possibility that Ovyron wins the game.
Even if we assume Ovyron is going to win the game it does not prove that unassisted engines can't win positions like these by themselves when they get significantly more than 88 hours and 57 minutes and it is possible to give them significantly more time in a correspondence game.
Maybe they can win it by themselves but zenmastur did not give them enough time at the time of posting and maybe the correct engine that zen did not use can win it by itself.
-
- Posts: 4556
- Joined: Tue Jul 03, 2007 4:30 am
Re: Zenmastur vs Ovyron No it's not another GROB II!
It would prove that the time it took me to play 3...f4 was enough to build a plan that would defeat any defense, while an unassisted engine is unable to find progress no matter how high depth it reaches.
I'm doing this with a 2300 k/ns average engine without 7men or 6men tablebases, with some methodology as dumb and simple as "fake evaluation hit when material is exchanged" - it's just about blowing up black's eval while keeping material on the board, no huge insight here.
If I can't win then everything I saw was a mirage and I would be exposed as a delusional clown.
If I win, it would show how you drive your car is more important than how much horsepower it has. And that drewdrew's "hardware dreams are made of" can perform the same as my tricycle in a game.
-
- Posts: 919
- Joined: Sat May 31, 2014 8:28 am
Re: Zenmastur vs Ovyron No it's not another GROB II!
Sorry for taking so long to move. I scrambled the DB with my analysis in it. I'm having to recreate it from scratch. Tomorrow morning early my wife has surgery. If I don't make a move by tonight it likely won't happen until after Wednesday morning.Ovyron wrote: ↑Mon May 18, 2020 7:21 amIt would prove that the time it took me to play 3...f4 was enough to build a plan that would defeat any defense, while an unassisted engine is unable to find progress no matter how high depth it reaches.
I'm doing this with a 2300 k/ns average engine without 7men or 6men tablebases, with some methodology as dumb and simple as "fake evaluation hit when material is exchanged" - it's just about blowing up black's eval while keeping material on the board, no huge insight here.
If I can't win then everything I saw was a mirage and I would be exposed as a delusional clown.
If I win, it would show how you drive your car is more important than how much horsepower it has. And that drewdrew's "hardware dreams are made of" can perform the same as my tricycle in a game.
Regards,
Zenmastur
Only 2 defining forces have ever offered to die for you.....Jesus Christ and the American Soldier. One died for your soul, the other for your freedom.
-
- Posts: 4556
- Joined: Tue Jul 03, 2007 4:30 am
-
- Posts: 1470
- Joined: Mon Apr 23, 2018 7:54 am
Re: Zenmastur vs Ovyron No it's not another GROB II!
Which previous tests? (e.g. do you count mate-in-N tests, which I previously said I'm not convinced are suitable tests?)
Do you mean DTZ 77?
-
- Posts: 4556
- Joined: Tue Jul 03, 2007 4:30 am