Correspondence ICCF Champion Vs Top Engine ?

[Cumnor]

Reminds me of when ICCF GM Arno Nickle Played Hydra in a correspondence match in 2004 https://amici.iccf.com/issues/issue_05/ ... hydra.html

[towforce]

Reminds me of when ICCF GM Arno Nickle Played Hydra in a correspondence match in 2004 https://amici.iccf.com/issues/issue_05/ ... hydra.html

Interesting: Hydra lost those games, but both the hardware and the software have improved since them. Time for a rematch!

[jr66]

2004 is indeed the Stone Age regarding engines !
Although one year later i think, Adams lost against this Hydra ?

[Milos]

I don't understand Milos sorry because it seem you speak only about engine vs engine match which is not the subject of the topic ?

There is nothing but engine vs. engine. Some ppl obviously wrongly believe that ICCF matches are like human-engine, while they are nothing but engine-engine matches. Just because you select first couple of moves from a database that is essentially again created by engines (selected by humans but that's irrelevant) and the rest it 99% pure engine decision and remaining 1% about taking a second or third engine move from the list instead of first or meditating when two top engines you use for analysis disagree.
What I wrote is the minimum requirement in terms of nodes searched per move so that the best ICCF player, using whatever he normally uses for analysis, is able to maybe win 1 game in 100.

[jr66]

I don't understand Milos sorry because it seem you speak only about engine vs engine match which is not the subject of the topic ?

There is nothing but engine vs. engine. Some ppl obviously wrongly believe that ICCF matches are like human-engine, while they are nothing but engine-engine matches. Just because you select first couple of moves from a database that is essentially again created by engines (selected by humans but that's irrelevant) and the rest it 99% pure engine decision and remaining 1% about taking a second or third engine move from the list instead of first or meditating when two top engines you use for analysis disagree.
What I wrote is the minimum requirement in terms of nodes searched per move so that the best ICCF player, using whatever he normally uses for analysis, is able to maybe win 1 game in 100.

Who are you for say that please? An ICFF GM ?
The criticism is easy at this time of very strong NNUE SF but i can tell you engines was not so strong some years ago and it was a pleasure to improve their play !
So, if i admit CC is in difficulty now with too much draws, it does not mean humans was not necessary ???
Finally, databases fondation were not created by engines but by hundreds of years humans games !!!
Not so long ago, engines Always played the French for example wihout Book and so, it was not the top ???

[Milos]

I don't understand Milos sorry because it seem you speak only about engine vs engine match which is not the subject of the topic ?

There is nothing but engine vs. engine. Some ppl obviously wrongly believe that ICCF matches are like human-engine, while they are nothing but engine-engine matches. Just because you select first couple of moves from a database that is essentially again created by engines (selected by humans but that's irrelevant) and the rest it 99% pure engine decision and remaining 1% about taking a second or third engine move from the list instead of first or meditating when two top engines you use for analysis disagree.
What I wrote is the minimum requirement in terms of nodes searched per move so that the best ICCF player, using whatever he normally uses for analysis, is able to maybe win 1 game in 100.

Who are you for say that please? An ICFF GM ?
The criticism is easy at this time of very strong NNUE SF but i can tell you engines was not so strong some years ago and it was a pleasure to improve their play !
So, if i admit CC is in difficulty now with too much draws, it does not mean humans was not necessary ???
Finally, databases fondation were not created by engines but by hundreds of years humans games !!!
Not so long ago, engines Always played the French for example wihout Book and so, it was not the top ???

ICCF GM's are already almost irrelevant, NNUE is kind of final nail in the coffin. Ppl that don't understand this are simply dinosaurs. I don't have any comparable ICCF rating, but I understand engines better than 99.99% of ICCF GMs.
Databases created by "hundreds of years humans games" were simply trashed by AlphaZero that is miles behind the current strong 100% engine generated books like CerebellumMerged. And as I mentioned they are also irrelevant because SFdev or Lc0 yield the strongest book moves if one just lets them analyze for an hour on a strong hardware.
Your "not so long ago" perception of engines seems to be before AlphaZero times (you seem stuck in SF8/9 times). And that might look recent to you, but in terms of engine strength and progress made is light years away.

[jr66]

AO Times is simply SF8 Times , no ?
And the SF8 play was not perfect as you say yourself ...
I don't doubt you know very well engines structures and performances ( i don't know you ) but what you say about ICCF players show you don't know how they use engines and databases ???
Enough for me !

[Milos]

AO Times is simply SF8 Times , no ?
And the SF8 play was not perfect as you say yourself ...
I don't doubt you know very well engines structures and performances ( i don't know you ) but what you say about ICCF players show you don't know how they use engines and databases ???
Enough for me !

You can't know what I know or don't know. Stating this is just presumptuous. What is obvious OTOH is that you don't give any arguments. That means beside an empty title (which today doesn't mean anything) you have nothing useful to offer in the discussion.

[towforce]

Your "not so long ago" perception of engines seems to be before AlphaZero times (you seem stuck in SF8/9 times). And that might look recent to you, but in terms of engine strength and progress made is light years away.

You mean Mac Hack running on a PDP-6 isn't still the strongest chess computer???

[jr66]

AO Times is simply SF8 Times , no ?
And the SF8 play was not perfect as you say yourself ...
I don't doubt you know very well engines structures and performances ( i don't know you ) but what you say about ICCF players show you don't know how they use engines and databases ???
Enough for me !

You can't know what I know or don't know. Stating this is just presumptuous. What is obvious OTOH is that you don't give any arguments. That means beside an empty title (which today doesn't mean anything) you have nothing useful to offer in the discussion.

What discussion ?
You don't consider me as human but as engine !
Indeed, i have nothing to offer to you ...
This topic has a title and if somebody could speak about it instead to derivate into "anti correspondance chess players" ???