Hi Jouni,
possible!
I think Robert Houdert tuned on different positions:
- avaid draw with Contempt = 0, must be inside the sources.
- higher evals for black pieces (more aggressive play as Stockfish between 8 and 9). I remember on a test in Rybka forum. Rybka 4 versions are around 30 Elo stronger with such tunings. Not a new idea!
- avoid to switched the queen and knight / bishop with many pieces on board (if I compare with Stockfish 8-9).
- different other things I found.
If so, Robert Houdert have good ideas he added in all of his Houdini program versions, not only v6.03.
A long time Houdini is on place 1 with the tunings Robert do.
No Komodo or Stockfish are stronger, do you remember?
Can't be bad the work Robert do here.
The advantage from my tourney is ...
All files are inside in the download.
No problem to delete programs from the database and calculate all again.
Best
Frank
PS: Yes many strengths in Houdini 6.03 looks like to be a Stockfish (for me too, endgames, transposition into endgame, pawn structur other things) but fact is that all the time the same good tunings are in all the Houdini versions. Can be easy see with easy statistics. But I am quit sure, that in many other chess programs are the good known ideas inside, not Houdini only. Fruit / Senpai sources for an example seems to be populare. The reason that I will not public all of the statistics to the still running tourney because I have no interest on such discussion. I like the style from Houdini but I dislike that Houdini is commercial available with closed sources. And I dislike to read reviews about commercial products, with the information that Houdini is a sensational chess engines. Speaking from reviews to Chessbase programs. I think such things ended but I read it for 2 months again. No words about Stockfish inside the review. Allways the same ...
Have a look on:
Equinox = IPP with better endgames (many difficult endgames are stronger as from best available IPPs)
GullChess = Improved IPP (perfect work because king safty is much better and bugfree, most IPPs have a lot of bugs)
Fire = IPP sources inside (strong work, I like the eval from the middlegames ... means eval is balanced and not speculative, eval I like most ... from all available chess engines).
Rybka / Fruit ... for me is Rybka a Fruit with a lot of endgame knowledge.
The middlegame is the same.
And so on ...
FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Moderators: hgm, Rebel, chrisw
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Small changes:
1.
Statistics to short games improved.
2.
At the moment we are working on spelling corrections in Excel file.
Klaus is working often on the Excel during the tournament.
Best
Frank
Have a look on xiphos in detail (short won games with white / black)!!
Looks interesting ... because only the programmer know why xiphos is very passive with black pieces!
1.
Statistics to short games improved.
2.
At the moment we are working on spelling corrections in Excel file.
Klaus is working often on the Excel during the tournament.
Best
Frank
Have a look on xiphos in detail (short won games with white / black)!!
Looks interesting ... because only the programmer know why xiphos is very passive with black pieces!
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi there,
here the ...
WASP-MISSING-Control-Center ... after round 21 of 50
Easy to see ... Wasp goes with a bad start in the tourney!
For a higher resolution:
http://www.amateurschach.de/t2020/commo ... mple-1.jpg
"The Missing Control Center" can be found to all of the 41 programs and much more!!
The complete Excel files (Excel 2010 or higher):
http://www.amateurschach.de/klaus/fcp-t ... lotzka.zip
In *.zip file you can find a short readme.txt.
Best
Frank
here the ...
WASP-MISSING-Control-Center ... after round 21 of 50
Easy to see ... Wasp goes with a bad start in the tourney!
For a higher resolution:
http://www.amateurschach.de/t2020/commo ... mple-1.jpg
"The Missing Control Center" can be found to all of the 41 programs and much more!!
The complete Excel files (Excel 2010 or higher):
http://www.amateurschach.de/klaus/fcp-t ... lotzka.zip
In *.zip file you can find a short readme.txt.
Best
Frank
Last edited by Frank Quisinsky on Sat Aug 29, 2020 1:28 am, edited 1 time in total.
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi there,
I wrote for days that we make a break from September 09th, - end of October 2020.
Around after round 25 of 50.
After the break the next 25 rounds will be follow.
Please be informed that after the tourney we replay the complete tourney with an other time-control.
5x lesser time ...
50 rounds with 256Mb for hash and: Game in 4 minutes + 1 second Fischer.
Same conditions, same engine versions!
The reason should be clear:
We will see the different of results with 5x lesser time for each of the 41 programs.
So we can not switched to "newer" versions of engines or a comparsion isn't possible!
All the years never produced a person such a comparsion.
It's time for it!
Best
Frank
High-Tech work around computer chess tounaments, never be available before!
And all this with tournament category XXXII (with 41 of best chess programs) on fastest Intel hardware (with 1 core).
A little thanks to all the engines programmers by Klaus Wlotzka and myself!
I wrote for days that we make a break from September 09th, - end of October 2020.
Around after round 25 of 50.
After the break the next 25 rounds will be follow.
Please be informed that after the tourney we replay the complete tourney with an other time-control.
5x lesser time ...
50 rounds with 256Mb for hash and: Game in 4 minutes + 1 second Fischer.
Same conditions, same engine versions!
The reason should be clear:
We will see the different of results with 5x lesser time for each of the 41 programs.
So we can not switched to "newer" versions of engines or a comparsion isn't possible!
All the years never produced a person such a comparsion.
It's time for it!
Best
Frank
High-Tech work around computer chess tounaments, never be available before!
And all this with tournament category XXXII (with 41 of best chess programs) on fastest Intel hardware (with 1 core).
A little thanks to all the engines programmers by Klaus Wlotzka and myself!
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi there,
now the results from round 22 out of 50 are online.
Frank's Chess Page
https://www.amateurschach.de
Best
Frank
now the results from round 22 out of 50 are online.
Frank's Chess Page
https://www.amateurschach.de
Best
Frank
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi there,
now the results from round 23 out of 50 are online.
New:
In *.zip file (games, statistics) I added new statistics:
Ratings "Games endet with 20-59, 60-79, 80-99, 100-300 moves"
Ratings "End positions with 02-06, 07-12, 13-32 pieces on board"
Klaus Wlotzka will added new statistics in Excel file soon.
Frank's Chess Page
https://www.amateurschach.de
Best
Frank
now the results from round 23 out of 50 are online.
New:
In *.zip file (games, statistics) I added new statistics:
Ratings "Games endet with 20-59, 60-79, 80-99, 100-300 moves"
Ratings "End positions with 02-06, 07-12, 13-32 pieces on board"
Klaus Wlotzka will added new statistics in Excel file soon.
Frank's Chess Page
https://www.amateurschach.de
Best
Frank
-
- Posts: 550
- Joined: Tue Nov 19, 2019 8:48 pm
- Full name: Alayan Feh
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
With double-digit W/D/L results, the cross-table has a formatting issue, with losses overflowing to the right into wins of the next cell. It makes it hard to look at results.
Removing the unnecessary spaces between wins, draws and losses would help, slightly expanding the cell size might also be required.
Removing the unnecessary spaces between wins, draws and losses would help, slightly expanding the cell size might also be required.
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi,
not sure which of the tables you have the problems!
1.
My easy *.txt files?
2.
The excel files by Klaus generated for the web site (links over cockpit from the detail page)
3.
Or the excel files by Klaus!
https://www.amateurschach.de/klaus/fcp- ... lotzka.zip (around 5Mb, Excel 2010 or higher)
I checked all the tables with google chrome and firefox again.
All works fine for me.
Hint:
The format from my *.txt files aren't to changed or Klaus have a lot of work with Excel.
Best
Frank
not sure which of the tables you have the problems!
1.
My easy *.txt files?
2.
The excel files by Klaus generated for the web site (links over cockpit from the detail page)
3.
Or the excel files by Klaus!
https://www.amateurschach.de/klaus/fcp- ... lotzka.zip (around 5Mb, Excel 2010 or higher)
I checked all the tables with google chrome and firefox again.
All works fine for me.
Hint:
The format from my *.txt files aren't to changed or Klaus have a lot of work with Excel.
Best
Frank
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi there,
on two different position you can find the download link to all files around the tourney (all stats I generated, inside the games in *.cbh and *.pgn).
Postion 1:
https://www.amateurschach.de/ftptrigger ... -2020.html
Around 90Mb for the moment!
Maybe I should added a second download file with statistics only and without games?
Position 2:
On the detail page:
On detail page under "logs, NEWS" you can find the news around the tourney and the download link to all the files.
For loading my *.txt files you need the frame from my website.
Two possibilites:
1. Over the cockpit is the frame loaded.
2. Directly over the menue system from my site
Best
Frank
on two different position you can find the download link to all files around the tourney (all stats I generated, inside the games in *.cbh and *.pgn).
Postion 1:
https://www.amateurschach.de/ftptrigger ... -2020.html
Around 90Mb for the moment!
Maybe I should added a second download file with statistics only and without games?
Position 2:
On the detail page:
On detail page under "logs, NEWS" you can find the news around the tourney and the download link to all the files.
For loading my *.txt files you need the frame from my website.
Two possibilites:
1. Over the cockpit is the frame loaded.
2. Directly over the menue system from my site
Best
Frank
-
- Posts: 6808
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: FCP Tourney-2020: 41.000 games 20+5 to play on Intel i9-10900k
Hi,
very big tables I produced here ...
Maybe a problem with the resolution on your system.
Here I can nothing do.
You can download the excel files by Klaus or my own *.txt statistic files, inside the 90Mb download.
Working with: Win 10 Prof. x64, Radeon RX580, Firefox browser, speed = 250mbit
iiyama Prolite XUB2792QSU-B1 A (27'' IPS panel technology), WQHD (2560x1440)
Means I am working with 2560x1440!
If you are working with 1024x768 ...
HOUSTON, we have a problem!
Here what you get if you download the current tournament files:
Best
Frank
very big tables I produced here ...
Maybe a problem with the resolution on your system.
Here I can nothing do.
You can download the excel files by Klaus or my own *.txt statistic files, inside the 90Mb download.
Working with: Win 10 Prof. x64, Radeon RX580, Firefox browser, speed = 250mbit
iiyama Prolite XUB2792QSU-B1 A (27'' IPS panel technology), WQHD (2560x1440)
Means I am working with 2560x1440!
If you are working with 1024x768 ...
HOUSTON, we have a problem!
Here what you get if you download the current tournament files:
Best
Frank