@maniwa wrote:I know it was not the same book in scct.
But in the tournament of Fantasmitadel50, your last book is not in the lead in 3 mins.
Take this!
Welcome to Chess2U - The Ultimate Computer Chess Forum!
If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post and use all thefeatures of the Chess2U forums. To start viewing messages, select the forum that you want to visit from the selection below.
Go to page : 
1, 2, 3, 4
Message [Page 2 of 4]
Take this!
Massimiliano Goi wrote:
My book has come back to the first place with a large score edge towards the other books. Pity the newest Medulla and Hiarcs 14zb didn't participate... At MZChess they celebrated when Pippo's Caudine Forks book took over mine for a while but, as we say here, "the one who really laugh is the one who laugh at the end". I wouldn't mind if Pippo's book won the tournament, he deserves because he's a great bookmaker. What is really important here is the distance between my book and Vitamin 13.
Now you have a third party's proof that my tests are legit!
Link: https://sites.google.com/site/computerschess/scct-book-cs-xx-results.
Games: https://www.mediafire.com/file/vidlmj3njxmuweb/scct_cs_20.7z/file.
Here you can buy the book: https://chess.massimilianogoi.com/shop/goiopeningbook/
Massimiliano Goi wrote:This was my last challenge about CTG books (unless I'll get at leat 10 preorders for the next version, as I've said), and as always it wasn't easy, mostly because my opponents kept to switch book versions.
But this time, managing now very well the Goi 5 technique, it was also amusing. With the latest CTG compilation I didn't lose a single game on Playchess, even with the strongest opponents! Except a single game today for... time (?)...
I guess those at Playchess should fix this time issue related to the delay of their server, the position was even in my favour... Anyway, there you go:
Goi 5.1 CTG:
It contains the most fresh lines from the most powerful computer games > 2480 in the Zagor Almanah Playchess tournament, the almost entire 2018 repertoires of two strong Playchess players (one of them at the top between the strongest), my latest games on Playchess, with the lost games corrected, the SCCT XX tournament games, the matches with the nowadays top books, some thousands games from my rent Xeon E7 (coffeeone) server et cetera...
The result is an average sized book (172 MB), having 2.389.034 positions. The lines are very, very strong. So strong that I had to botch some interesting opening made by Grand Masters like Aronian, in order to beat the other books... On the other hand this has opened my eyes: I have checked the Mega Database 2018 for that opening and with that black move I have botched the white statistically won.
Another interesting thing is that I have inserted some trick advised by the IM Sergio Durán Vega to enhance the performance.
The book was tested several days on Playchess, in which I kept my elo score above 2500. You can have fun by set up matches between books, but if you want the real thing you must test your book in the Playchess arena, where there is always some big fighter like Leonidas or SASSY MUFFIN ready to destroy your games. I've survived to them these days.
As last release, Goi 5.1 CTG must be something special: this book will be subdivided in three updates, I will experiment alot in the further time. And a special book comes with a special price.
Depending on how the sales will go I will decide the future of this very website, I have alot of things in mind, but I have also personal issues to face off.
Benchs:
Goi 5.1 CTG vs Caudine Forks 1977-25.10.18 (on Raubfisch X39d3b2):
Download here the games archives in Chessbase and PGN format.
Same engine cloned, 'Permanent Brain' disabled, four cores for each engine, 128 MB of hash table, game 1 minute.
- The settings of Goi 5.1 CTG were: Tournament book: ticked, Minimum games: 0, Up to move: 100, Variety of play: 10, Influence of learn value: 100, Learning strength: 0.
- The settings of Caudine Forks 1977-25.10.18 were: Tournament book: ticked, Minimum games: 0, Up to move: 100, Variety of play: 150, Influence of learn value: 100, Learning strength: 0.
Goi 5.1 CTG vs Hiarcs 14zb (on Raubfisch X39d3b2):
Download here the games archives in Chessbase and PGN format.
Same engine cloned, 'Permanent Brain' disabled, four cores for each engine, 128 MB of hash table, game 1 minute.
- The settings of Goi 5.1 CTG were: Tournament book: ticked, Minimum games: 0, Up to move: 100, Variety of play: 10, Influence of learn value: 100, Learning strength: 0.
- The settings of Hiarcs 14zb were: Tournament book: ticked, Minimum games: 0, Up to move: 100, Variety of play: 147, Influence of learn value: 100, Learning strength: 0.
Goi 5.1 CTG vs Vitamin13 n6s3h RC7 21oct2018 (on Raubfisch X39d3b2):
Download here the games archives in Chessbase and PGN format.
Same engine cloned, 'Permanent Brain' disabled, four cores for each engine, 128 MB of hash table, game 1 minute.
- The settings of Goi 5.1 CTG were: Tournament book: ticked, Minimum games: 0, Up to move: 100, Variety of play: 10, Influence of learn value: 100, Learning strength: 0.
- The settings of Vitamin13 n6s3h RC7 21oct2018 were: Tournament book: ticked, Minimum games: 0, Up to move: 70, Variety of play: 127, Influence of learn value: 100, Learning strength: 0.
Oh! I didnt know 5.1 was ctg format...Sad in my own use cuz I prefer bin formats.
________________
Ch355 2 u m3mb3r WALDEMAR...g00d g4m3 f0r 4ll...y4h000000!
Goi 5.1 won 28 games of 51
Medulla 311018 won 23 games
So the winner is Goi 5.1
Played with : Same engine - Brainfish 181105
Best move -Not Ticked
5xmen syzygy
1min each
Very Very strong book.!
Medulla 311018 won 23 games
So the winner is Goi 5.1
Played with : Same engine - Brainfish 181105
Best move -Not Ticked
5xmen syzygy
1min each
Very Very strong book.!
That guy scored bad Goi 5.1 BIN even when it was being the SCCT XX champion, so he's not attendible. Many errors in the games happened because of his misconfiguration of the engines or books during the games.
You say bin.book , but then its shows ctg?
I m not sure how you run the test but in all my test nobody could win goi 5.1
But then, Im only testing it on the Arena gui.
So then also you tell me the scct tournaments are also wrong?
Look at the elo of Goi5.1 bin it is quite remarkable.
he was using bin before but no effect at all and updated to ctg here is the result updated book competion today here the result 
Goi 5 and 5.1 are weak books. Cerebellum Light Poly. bin to drawish.
No chance on Infinity online chess site.
No chance on Infinity online chess site.
@Andries wrote:https://chess.massimilianogoi.com/download/
We should focus on the free and not the paid. The purpose of this forum is not to advertise
@jorge91 wrote:That guy scored bad Goi 5.1 BIN even when it was being the SCCT XX champion, so he's not attendible. Many errors in the games happened because of his misconfiguration of the engines or books during the games.
We should focus on the free and not the paid. The purpose of this forum is not to advertise.
@Alexia_Nadine wrote:Goi 5 and 5.1 are weak books. Cerebellum Light Poly. bin to drawish.
No chance on Infinity online chess site.
Did you try it to say that? I bet not....
@jorge91 wrote:That guy scored bad Goi 5.1 BIN even when it was being the SCCT XX champion, so he's not attendible. Many errors in the games happened because of his misconfiguration of the engines or books during the games.
Very good for everyone!
1) First of all I want to say that goi5.bin or ctg are excellent books and I congratulate maximiliano goi alias jorge91 for his work.
2) The tournament I am doing in MZ chess forum, is a serious, responsible tournament, and I thank all those participants who trusted me to take it forward, giving me their books for free (except goi5) to make an evaluation.
3) Although it is a competition, this is taking place within an atmosphere of camaraderie, where everyone tries to contribute positively and nobody gets angry about any unfavorable result of their book.
4) Regarding the tournament itself, I disagree with the comment of Jorge91, when he tries to justify the bad round of his book, claiming that there was a misconfiguration of his book and engine.
This is not the case in any way.
The configuration of a book is set by the author of the book and nobody else, and the configuration of the engine is the same for all and these characteristics are written in the corresponding section of the MZ forum, for all those who want to see it.
There are also screenshots, with the engine configuration used.
What did change the hash, the tournament started with 384 MB of hash and several rounds ago it was put in 1024 hash.
5) I believe that a good book should be able to play and have a good performance with different configurations, otherwise it is not a good book, if it would not be the champion or the best within a certain configuration.
6) If you observe carefully in the championship of sedat canbaz the difference is within the margin of error of the test, apart from something very important, in this tournament ONLY 100 to 150 matches were played per book, very few to determine the science Certain if one book is better than another.
Especially within the first positions.
Message [Page 2 of 4]
Go to page : 1, 2, 3, 4
Related Topics
|
|
|