ICGA WCCC and WCSC in Santiago de Compostela

[Engin]

“We must admit that chess has fallen out of the sp It’s time to put an end to this matter. Today’s programs are approximately of the same level. They play 30 games to a draw, then one of them wins, but this no longer arouses any interest,” Vladimir Lvovich noted.

Not sure what you really mean by " how in line are this legends...".
Chess is not solved. Thus, we don't know how to play perfectly.
I recall years ago many SF fanatics claimed computer chess was dead because SF was at the top
for so long and thought that it played nearly perfect. But, then Alpha Zero and Leela come along and destroyed that thought.
Then some thought that Leela was impossible to beat, but new ideas came that could beat Leela and SF. The SF team adopted them quickly.
Now, some claim it again.
I doubt the search algorithm in SF plays perfectly: it prunes heavily, maybe too much.
So, new ideas will come .... and as history suggests they will not be from SF.
To my WCCC colleagues and friends: some of you liked my ideas to make neural nets more efficient and want to try them.
That is fine. I plan to publish on the topic after gathering more data. Please don't repeat the ideas for the SF team to get before I can publish.

Even though LC0 wasn't there, and the SF derivative was woefully underpowered on only 6 cores , it was great to see Stoofvlees (tens of thousands of nps) trading equal blows with Jonny (billions of nps). It's lovely to still have diversity of approach.

Looking forward to the publication Charles

Stockfish is too drawish. They should insert an Aggressive mode, they have even removed the contempt factor option, instead...

Most of the games are drawish by perfectly played, just humans still make blunders.

[Engin]

When asked by some of the passing researchers if their program was more like AlphaZero or more like Stockfish, one of the Sashchess guys immediately said "it IS Stockfish". It's not like they're making any misrepresentations here in that regard.

Same for Tornado https://www.chessprogramming.org/Tornado

Tornado is not Stockfish or has to do anything with it, i did it wrote from stratch and got only just ideas from other programmers and sources but i never did copy and paste anything from them.

OK i am using the probing code from Daniel Shawul for the NNUE's, not more using the Stockfish networks anymore, just i generated from my own games and data new NNUE's to use it.

why should I reinvent the well knowing wheel?

To write an engine nowadays to take well knowledge of the past has nothing to do with cloning or derivatives as long you understanding the algorithm or working on such like to make it stronger then others.

People who shout out loud that this is a clone have no idea about programming.

I don't have to justify to anyone how I wrote my program, nor do I have to reveal the source code to anyone, it's not open source but more of a private hobby program.

[Engin]

Most of the games are drawish by perfectly played, just humans still make blunders.

it should be drawish logically by perfecty playing of a game, because white has for the first move the tempo advantage but black can refute that tempo too. so every played move from white can be refuted by black, the outcome must be drawish.

just only by making blunders can the other part win the game

[jdart]

WCCC-games.pgn has some errors:

No king move possible to e5.
File wccc-games.pgn: Line number: 1679
Failed to make move 47... Ke5+ in the game:
........
.p......
..p....p
.....k..
PP.Kp..p
.......P
........
........

Tech 4 - Shashchess World Computer Chess Championship 2024 Santiago de Compostela 2024.10.22
File wccc-games.pgn: Line number: 1679
No king move possible to d5.
File wccc-games.pgn: Line number: 2289
Failed to make move 31... Kd5 in the game:
........
ppp.....
..p.k...
........
....P...
....K...
PPP....R
..r.....

Jonny - Rofchade World Computer Chess Championship 2024 Santiago de Compostela 2024.10.23
File wccc-games.pgn: Line number: 2289

[mclane]

Dont the participants export the games digitally? Do they still use paper ? With wrong written notation?

[Engin]

do you tried this ones?

forum3/viewtopic.php?t=84442

[Engin]

WCCC-games.pgn has some errors:

No king move possible to e5.
File wccc-games.pgn: Line number: 1679
Failed to make move 47... Ke5+ in the game:
........
.p......
..p....p
.....k..
PP.Kp..p
.......P
........
........

Tech 4 - Shashchess World Computer Chess Championship 2024 Santiago de Compostela 2024.10.22
File wccc-games.pgn: Line number: 1679
No king move possible to d5.
File wccc-games.pgn: Line number: 2289
Failed to make move 31... Kd5 in the game:
........
ppp.....
..p.k...
........
....P...
....K...
PPP....R
..r.....

Jonny - Rofchade World Computer Chess Championship 2024 Santiago de Compostela 2024.10.23
File wccc-games.pgn: Line number: 2289

This is because black has won, there for they put the black king to e5 that the arbiter knows that black has won.

[CRoberson]

Dont the participants export the games digitally? Do they still use paper ? With wrong written notation?

All game scores recorded via use of DGT boards, pieces and clocks generously supplied by DGT.

[Gian-Carlo Pascutto]

The tiebreak games are (still) missing, but they are on lichess: https://lichess.org/broadcast/icga-worl ... 1/eMN1appo