Is EMAN the future number one?

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

Glarean
Posts: 262
Joined: Sun Oct 05, 2008 1:04 pm
Location: Switzerland
Full name: Walter Eigenmann

Re: Is EMAN the future number one?

Post by Glarean »

Ovyron wrote: Tue Jun 09, 2020 3:04 am
Glarean wrote: Tue Jun 09, 2020 12:17 am Match Eman vs Honey 27:23
Try now with Honey's Profound setting activated.
But "Profound" is already a default setting in Honey XI-r1?
(I usually let the engines play with the default settings, except threads and hash)

.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is EMAN the future number one?

Post by MikeB »

Glarean wrote: Tue Jun 09, 2020 8:10 pm
Ovyron wrote: Tue Jun 09, 2020 3:04 am
Glarean wrote: Tue Jun 09, 2020 12:17 am Match Eman vs Honey 27:23
Try now with Honey's Profound setting activated.
But "Profound" is already a default setting in Honey XI-r1?
(I usually let the engines play with the default settings, except threads and hash)

.
That is correct - I had not noticed the time control that you used?
Image
Dann Corbit
Posts: 12791
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Is EMAN the future number one?

Post by Dann Corbit »

corres wrote: Tue Jun 09, 2020 7:26 pm I think Eman is the Houdini 7.
Houdart thought they work such a many on Stockfish problems the Houdini is fully his own work.
Khalid Omar said thanking to his work on Stockfish up to now there is no any connection between Stockfish and Eman, Eman is an independent chess engine created by him.
Here is what he says on his own official chess site: https://eman.zohosites.com/
"Eman is a UCI chess engine by Khalid Omar
It is derived from Stockfish, a very strong world class and open-source chess engine."

Are you calling him a liar?
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.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is EMAN the future number one?

Post by MikeB »

Dann Corbit wrote: Tue Jun 09, 2020 9:01 pm
corres wrote: Tue Jun 09, 2020 7:26 pm I think Eman is the Houdini 7.
Houdart thought they work such a many on Stockfish problems the Houdini is fully his own work.
Khalid Omar said thanking to his work on Stockfish up to now there is no any connection between Stockfish and Eman, Eman is an independent chess engine created by him.
Here is what he says on his own official chess site: https://eman.zohosites.com/
"Eman is a UCI chess engine by Khalid Omar
It is derived from Stockfish, a very strong world class and open-source chess engine."

Are you calling him a liar?
I read his post as sarcasm based on what he said regarding Houdart/Stockfish.
Last edited by MikeB on Tue Jun 09, 2020 10:55 pm, edited 2 times in total.
Image
corres
Posts: 3657
Joined: Wed Nov 18, 2015 11:41 am
Location: hungary

Re: Is EMAN the future number one?

Post by corres »

Dann Corbit wrote: Tue Jun 09, 2020 9:01 pm
corres wrote: Tue Jun 09, 2020 7:26 pm I think Eman is the Houdini 7.
Houdart thought they work such a many on Stockfish problems the Houdini is fully his own work.
Khalid Omar said thanking to his work on Stockfish up to now there is no any connection between Stockfish and Eman, Eman is an independent chess engine created by him.
Here is what he says on his own official chess site: https://eman.zohosites.com/
"Eman is a UCI chess engine by Khalid Omar
It is derived from Stockfish, a very strong world class and open-source chess engine."
Are you calling him a liar?
I think only developer team of Stockfish can make such statement.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is EMAN the future number one?

Post by MikeB »

corres wrote: Tue Jun 09, 2020 10:41 pm
Dann Corbit wrote: Tue Jun 09, 2020 9:01 pm
corres wrote: Tue Jun 09, 2020 7:26 pm I think Eman is the Houdini 7.
Houdart thought they work such a many on Stockfish problems the Houdini is fully his own work.
Khalid Omar said thanking to his work on Stockfish up to now there is no any connection between Stockfish and Eman, Eman is an independent chess engine created by him.
Here is what he says on his own official chess site: https://eman.zohosites.com/
"Eman is a UCI chess engine by Khalid Omar
It is derived from Stockfish, a very strong world class and open-source chess engine."
Are you calling him a liar?
I think only developer team of Stockfish can make such statement.
?? That is certainly not true (imho) , but I am curious as to your logic behind that statement. Perhaps a google translation issue?
Image
corres
Posts: 3657
Joined: Wed Nov 18, 2015 11:41 am
Location: hungary

Re: Is EMAN the future number one?

Post by corres »

MikeB wrote: Tue Jun 09, 2020 10:54 pm
corres wrote: Tue Jun 09, 2020 10:41 pm
Dann Corbit wrote: Tue Jun 09, 2020 9:01 pm
corres wrote: Tue Jun 09, 2020 7:26 pm I think Eman is the Houdini 7.
Houdart thought they work such a many on Stockfish problems the Houdini is fully his own work.
Khalid Omar said thanking to his work on Stockfish up to now there is no any connection between Stockfish and Eman, Eman is an independent chess engine created by him.
Here is what he says on his own official chess site: https://eman.zohosites.com/
"Eman is a UCI chess engine by Khalid Omar
It is derived from Stockfish, a very strong world class and open-source chess engine."
Are you calling him a liar?
I think only developer team of Stockfish can make such statement.
?? That is certainly not true (imho) , but I am curious as to your logic behind that statement. Perhaps a google translation issue?
Nothing Google translation.
Maybe you think the Stockfish license owner is the Free Software Foundation?
Or it is in communal property?
Or perhaps you know a committee what can decide about the question?
Dann Corbit
Posts: 12791
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Is EMAN the future number one?

Post by Dann Corbit »

If I say "I used Joe's lawnmower" Joe does not have to verify that I used it. Maybe he did not see me do it, but I admitted to it.
The Eman site says it's a clone. Therefore, the SF group does not have to say "It's a clone."
We already know it because Khalid Omar said so.
Furthermore, every time he does a patch (for example : https://cumnor-chess-server.com/phpbb/v ... .php?t=217) he explains that he has back-patched with the current Stockfish code. That is only possible if it is a clone.

He says its a clone. It is a clone. No puzzling or wondering or throwing up our hands and furrowing our worried brow.
It walks like a duck, it quacks like a duck, it has a label on it that says, "I am a duck."
If you would like to imagine that it's not a duck, nobody will stop you.
But it's still a duck.
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.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is EMAN the future number one?

Post by MikeB »

Dann Corbit wrote: Wed Jun 10, 2020 12:51 am If I say "I used Joe's lawnmower" Joe does not have to verify that I used it. Maybe he did not see me do it, but I admitted to it.
The Eman site says it's a clone. Therefore, the SF group does not have to say "It's a clone."
We already know it because Khalid Omar said so.
Furthermore, every time he does a patch (for example : https://cumnor-chess-server.com/phpbb/v ... .php?t=217) he explains that he has back-patched with the current Stockfish code. That is only possible if it is a clone.

He says its a clone. It is a clone. No puzzling or wondering or throwing up our hands and furrowing our worried brow.
It walks like a duck, it quacks like a duck, it has a label on it that says, "I am a duck."
If you would like to imagine that it's not a duck, nobody will stop you.
But it's still a duck.
+1 well said, i have nothing else to add ...
except for what was dropped off to me earlier today... anonymously

Code: Select all

 ** Running: C:\cluster.mfb\Popcnt-LP\Eman.exe
** Arguments: bench 2048 64 26 default depth
 _________   ____    ____        __        ____  _____ 
|_   ___  | |_   \  /   _|      /  \      |_   \|_   _|
  | |_  \_|   |   \/   |       / /\ \       |   \ | |  
  |  _|  _    | |\  /| |      / ____ \      | |\ \| |  
 _| |___/ |  _| |_\/_| |_   _/ /    \ \_   _| |_\   |_ 
|_________| |_____||_____| |____|  |____| |_____|\____|

Eman 5.60 64-bit POPCNT by Khalid Omar
Compiled by g++ (GNUC) 7.3.0 on MinGW64

Processor          : AMD Ryzen Threadripper 3970X 32-Core Processor 
NUMA nodes         : 1
Processor cores    : 32
Logical processors : 64
Hyper-Threading    : Yes
L1/L2/L3 cache     : 2 MB / 16 MB / 131 MB
RAM                : 134 GB

info string Memory LargePages is available
info string NUMA optimization is not available
info string LiveBook initialized successfully
info string Allocated 2048 MB LargePage memory for Transposition table

Position: 1/55
Position: 2/55
Position: 3/55
Position: 4/55
Position: 5/55
Position: 6/55
Position: 7/55
Position: 8/55
Position: 9/55
Position: 10/55
Position: 11/55
Position: 12/55
Position: 13/55
Position: 14/55
Position: 15/55
Position: 16/55
Position: 17/55
Position: 18/55
Position: 19/55
Position: 20/55
Position: 21/55
Position: 22/55
Position: 23/55
Position: 24/55
Position: 25/55
Position: 26/55
Position: 27/55
Position: 28/55
Position: 29/55
Position: 30/55
Position: 31/55
Position: 32/55
Position: 33/55
Position: 34/55
Position: 35/55
Position: 36/55
Position: 37/55
Position: 38/55
Position: 39/55
Position: 40/55
Position: 41/55
Position: 42/55
Position: 43/55
Position: 44/55
Position: 45/55
Position: 46/55
Position: 47/55
Position: 48/55
Position: 49/55
Position: 50/55
Position: 51/55
Position: 52/55
Position: 53/55
Position: 54/55
Position: 55/55

===========================
Total time (ms) : 94843
Average depth   : 26.00
Nodes searched  : 8988478736
Nodes/second    : 94772189
uci
id name Eman 5.60 64-bit POPCNT
id author Khalid Omar

option name Debug Log File type string default
option name Contempt type spin default 24 min -100 max 100
option name Dynamic Contempt type check default true
option name Threads type spin default 1 min 1 max 512
option name Only Instance type check default false
option name Large Pages type check default true
option name Verbose type check default false
option name Clear Hash type button
option name Hash type spin default 16 min 1 max 131072
option name Full Analysis Depth type spin default 0 min 0 max 245
option name Pruning Threshold type spin default 32 min 0 max 32
option name Null Move Pruning type check default true
option name Experience File type string default Eman.exp
option name Experience Read Only type check default false
option name Experience MultiPV type check default true
option name Experience Min Depth type spin default 12 min 1 max 64
option name Experience Book type check default false
option name Experience Book Min Depth type spin default 27 min 12 max 245
option name Ponder type check default false
option name MultiPV type spin default 1 min 1 max 500
option name Move Overhead (ms) type spin default 10 min 0 max 5000
option name Time Buffer (s) type spin default 0 min 0 max 20
option name Slow Mover type spin default 100 min 10 max 1000
option name UCI_Chess960 type check default false
option name Live Book type check default false
option name Live Book URL type string default http://www.chessdb.cn/cdb.php
option name Live Book Timeout type spin default 5000 min 0 max 10000
option name Live Book Depth type spin default 100 min 1 max 100
option name Live Book Diversity type check default false
option name Live Book Contribute type check default true
option name Book 1 type string default Eman.bin
option name Book 1 Learning Enabled type check default true
option name Book 1 Learn From Draw type check default true
option name Book 1 Width type spin default 1 min 1 max 20
option name Book 1 Depth type spin default 100 min 1 max 100
option name Book 2 type string default <empty>
option name Book 2 Learning Enabled type check default true
option name Book 2 Learn From Draw type check default true
option name Book 2 Width type spin default 1 min 1 max 20
option name Book 2 Depth type spin default 100 min 1 max 100
option name OutOfBook ThinkHard Moves type spin default 9 min 0 max 16
option name SyzygyPath type string default <empty>
option name SyzygyProbeDepth type spin default 1 min 1 max 100
option name Syzygy50MoveRule type check default true
option name SyzygyProbeLimit type spin default 7 min 0 max 7
uciok
Initial glance tells you this is farther removed from SF as compared to Honey, Crystal or Corchess. One feature that I have not seen before , is that it appears it can learn from just feeding it pgn files? I will have to test , but we all know the value of Brainfish and even it just gets close to brainfish ( and who knows , it could be far beyond brainfish - this is my very first look at it ) but learning or experience could easily add some relatively big elo.

First I will run some tests with no learning enabled or being used. Then repeat with learning on and then repeat with learning on and being used. I suspect this might do very well indeed if the learning was implemented correctly.
Image
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Is EMAN the future number one?

Post by MikeB »

MikeB wrote: Tue Jun 09, 2020 7:19 pm
Dann Corbit wrote: Tue Jun 09, 2020 6:11 pm I have logging turned on by default so you will want to turn that off with UCI. I am getting really interesting results with solving problem sets that were previously very difficult.
A small match at time 5 min with 6 sec increment,

Honey-XIr2 ( this is the soon to be release)
cur-dev Stockfish
Bluefish-XIr2 ( also release version)
cur-dev- corchess (aka Corchess 6.0)

Both Honey and Bluefish are using Profound Analysis turned on with Pro Value value of 26.

Faster time controls have favored Stockfish, it's only when you get to 5 min plus 6 seconds that Profound Analysis shines in match play against SF. Ymmv of course. Hash 512M- single thread

Current Results:

Code: Select all

Current date : time (EDST)
Date: 06/09/20 : 13:03:11
Rank Name            Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR 
---------------------------------------------------------------------------------------------------------
   1 Honey-XI-r2      3580   0.0   31   31    97   54.5  56.2   18    6   73  18.6  75.3  3540 
   2 stockfish        3567  12.3   31   31    98   52.5  53.6   18   11   69  18.4  70.4  3544 
   3 Bluefish-XI-r2   3537  30.0   31   31    98   46.5  47.4   10   15   73  10.2  74.5  3554 
   4 corchess         3516  21.7   30   30    97   41.5  42.8    4   18   75   4.1  77.3  3562 
---------------------------------------------------------------------------------------------------------
link:
https://www.dropbox.com/s/po89jovub1qcf ... a.txt?dl=0

games in pgn format:
https://www.dropbox.com/s/vaaqnok1gi8sn ... 1.pgn?dl=0
For completeness, final results.

Code: Select all

Current date : time (EDST)
Date: 06/09/20 : 15:59:25
Rank Name            Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR 
---------------------------------------------------------------------------------------------------------
   1 Honey-XI-r2      3575   0.0   18   18   288  158.5  55.0   46   17  225  16.0  78.1  3542 
   2 stockfish        3560  15.1   19   19   288  149.5  51.9   44   33  211  15.3  73.3  3547 
   3 Bluefish-XI-r2   3540  20.3   19   19   288  138.0  47.9   29   41  218  10.1  75.7  3553 
   4 corchess         3526  13.9   18   18   288  130.0  45.1   21   49  218   7.3  75.7  3558 
--------------------------------------------------------------------------------------------------------
This is better than I would expect and I can attribute that the relatively few games - just 96 with each opponent - 48 positions drawn at random.
I would say +5 Elo gain vs SF at this tc is probably more in line with expectations ( and only at tc 5 min base with 6 sec increment, Elo decreases pretty rapidly as time control is decreased from 5 6. Even +5 Elo may be overly generous as the difference between very similar engines is often exaggerated in testing against itself. They are just too similar. All of the above engines are very similar to each other.

Also, please keep in mind here,that the true heroes here are the members of SF development team, both past and present, as none of this happens without them.
Image