Forum

Of Computer Chess

 
    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 the
    features of the Chess2U forums. To start viewing messages, select the forum that you want to visit from the selection below.

IvanHoe 999946f x64 vs. IvanHoe B46fB x64, 2.800 games, still running 5 5 14

Goto page : 1, 2  Next

View previous topic View next topic Go down  Message [Page 1 of 2]

Hi there,

at first:

IvanHoe B46fB by PeterPan is available.
Some bugfixes for Shredder Classic 4 GUI.
I believe PeterPan set the download link in immortal forum!
Version is tested by myself, all works fine!

at second:

IvanHoe 999946f x64
Shredder Classic 4 configuration file


Code:
[ENGINE]
Name=IvanHoe 999946f x64
Author=KLO
Filename=C:\Chess\Shredder 12\Engines\IvanHoe\v999946f-x64\IvanHoe_999946f-x64.exe
[OPTIONS]
PawnsHash=16
PVHash=4
EvalHash=1
Threads=1
AllowInstantMoveFromHash=false
RobboTripleBaseDirectory=c:\Chess\Shredder 12\Engines\IvanHoe\TripleBases-new
RobboTotalBaseDirectory=c:\Chess\Shredder 12\Engines\IvanHoe\TotalBases-dir
RobboTotalBaseCacheSize=32
TripleHashSize=32
MultiCentiPawnPV=1
TimeEasyFactor=14
TimeEasyFactorPonder=31
TimeOrdinaryFactor=64
TimeAbsolutePercent=24


IvanHoe B46fB x64
Shredder Classic 4 configuration file


Code:
[ENGINE]
Name=IvanHoe B46fB x64
Author=PeterPan
Filename=C:\Chess\Shredder 12\Engines\IvanHoe\vB46fB-x64\IvanHoe_B46fB-x64.exe
[OPTIONS]
PawnsHash=16
PVHash=4
EvalHash=1
TitanicCPUsSizings=1
AllowInstantMoveFromHash=false
TripleRootFolder=c:\Chess\Shredder 12\Engines\IvanHoe\TripleBases-new
LoadTripleBases=Load
TotalRootFolder=c:\Chess\Shredder 12\Engines\IvanHoe\TotalBases-dir
LoadTotalBases=Load
RobboTotalBaseCacheSize=32
TripleHashSize=32
TimeEasyFactor=14
TimeEasyFactorPonder=31
TimeOrdinaryFactor=64
TimeAbsolutePercent=24


I changed for both versions different settings!
I think this one is better for ponder = on and my SWCR conditions with 40 moves in 10 minutes, 40 moves in 10 minutes again and again, without resign mode.

Code:
AllowInstantMoveFromHash=false
PawnsHash=16
PVHash=4
TripleHashSize=32
TimeEasyFactor=14
TimeEasyFactorPonder=31
TimeOrdinaryFactor=64
TimeAbsolutePercent=24


And here are the round robin information:

1. IvanHoe 999946f x64
http://www.amateurschach.de/ftptrigger/ivanhoe_999946f-x64.html
1.400 games : 40 = 35 participants.
I added all what I can and is strong and tested

2. IvanHoe B47fB x64
http://www.amateurschach.de/ftptrigger/ivanhoe_b46fb-x64.html
1.400 games : 40 = 35 participants.
I added the same engines for a better compare to IvanHoe 999946f x64.

My webapges will be updated today too, this need a while.

For a better compare:
Latest tested IvanHoe version in SWCR:

Code:
12 IvanHoe B47cB x64              2940  15  15  1768  71%  2783  39%


If I got not a GUI crash ...
288 games on 4 Q9550 systems can be play, 8 games are still running to the same time.
2.400 : 288 = Test need around 8,5 days!

SWCR conditions:

Code:
#Time control: 40/10 "repeatedly"
#Game average: 40 minutes, moves average = 171,55
#Resign      : OFF
#Ponder      : ON
#Learning    : OFF
#Endgames    : 4-pieces, 32Mb cache
#Opening book: Own PGN random 5.10 book
#GUI        : Shredder Classic 4 / Arena
#OS          : Windows XP Pro. x64 Edition
#Processors  : Intel® Core™2Q, 4xQ9550 2,83GHz
#Cores      : 1 core for each engine
#Hash-Tables : 256Mb 
#Games      : 1.000 per engine minimal


Have fun with IvanHoe ...

Best
Frank

wow...
very nice Frank
great job...thanks!

i'll be watching, am very interested to see the results

Hi Norman,

thanks, also for the all the work you do!

Now, let us have fun with the results in the next 8-10 days.
I hope that all will be run without crashes. Some of the opponents have small problems and I must make correction by hand (lost on time games for an example).

Best
Frank

Frank, Always i see a new good ideas in your mind, you are doing a great work with your rating list for free.
You are a real chess lover. (I like such style by the way.)
Keep up the good work!

Frank wrote:IvanHoe 999946f x64 vs. IvanHoe B46fB x64, 2.800 games, still running
[...]

I think your Ivanhoe 999946f/B46fB test is a waste of time, Just because the two engines have the same strength (from the same source).

Hi there,

I updated the SWCR database. So you can find the first 200 games for both Ivanhoe version in the database.
The playing style between both versions is different.
Can be compare to Rybka 4.0 / Rybka 4.0 Exp. 42 setting or Robbolito 0.09 x64 / Houdini 1.0 x64

IvanHoe 999946f x64 is playing more tacticial chess.
Remis quote 30%, move average 83 (without resign in SWCR)

IvanHoe B46fB x64 is playing more solid chess.
Remis quote 38%, move average 88 (without resign in SWCR)

In this case the final results should be clear. Can be say after the first 248 games, think so.

IvanHoe 999946f x64, better vs. weaker opponents, better for blitz
IvanHoe B46fB x64, better vs. stronger opponents, better for rapid

The results in this constellation will be allways the same!

Interesting is that IvanHoe 999946f x64 could be the first engines which win a 40 game match vs. Houdini 2.0c / 1.5 x64.

Best
Frank

But you can see it in selfwork.
Install both and analyze a position.
Example start position: One engine will play 1. d4 and the other 1. e4

Style is quit different.
Or load the SWCR database (included are the first 200 games) and replay the games, open both versions as analyze engine.

Hi Ahmed,

thanks for your friendly comments.

But I am not sure how long my list will be run.
I am thinking about it to closed my rating list start of next year, or in the second quartal.
I am not sure about it.

Best
Frank

Frank wrote:Hi Ahmed,

thanks for your friendly comments.

But I am not sure how long my list will be run.
I am thinking about it to closed my rating list start of next year, or in the second quartal.
I am not sure about it.

Best
Frank

Why ?

Hi Ahmed,

for different reasons, isn't easy to explain in English for myself. I can't see any new statistical records which are interesting for myself. Only interesting is to visiting one or two of the still running 8 matches on my 4 SWCR quad core systems. I am thinking about it to do other things in computer chess. Furthermore, I am very disappointed about the second open letter by a group of programmers. Wrote a lot about it in TalkChess.

I can't support this second open letter for around 20 reasons I added in TalkChess.

Furthermore, I have a more work with my list as I am thinking before. OK, I think my list is now wonderful organized but this should be not a reason to work a longer time on it.

But I am not sure about it and will again thinking about my decisson after the Oldie-Mix tourney (end of January 2012).

Best
Frank

I do not understand what you talking about... I do not follow such things,sorry.
I play chess ,It's more fun!
But if you want advice from a friend:
Impatience, anger, frustration, are things which happen in your mind. do not let them control your enjoyment of life.

Hi Ahmed,

if you spent so many hour as I do in the past for this hobby have this nothing to do with controls or frustration. But I says at before, this have more as one reasons and isn't easy to explain.

I am sure you knwo what I mean if I am thinking on your mail to myself with the wish that I have to delete the SWCR tested IvanHoe compiles by yourself. I am sure that you have reasons for it and be sure that I have my reasons.

Have a nice day!

Best
Frank

Okay Frank, This is your decision. Just think carefully before taking any step.
Good luck and you have my full support.

Ahmed Kamal wrote:
I play chess ,It's more fun!
But if you want advice from a friend:
Impatience, anger, frustration, are things which happen in your mind. do not let them control your enjoyment of life.

+1

Frank wrote:
But you can see it in selfwork.
Install both and analyze a position.
Example start position: One engine will play 1. d4 and the other 1. e4

Style is quit different.
Or load the SWCR database (included are the first 200 games) and replay the games, open both versions as analyze engine.

The same engine with different parameters, that is all.

Think so!
But this is quit clear.

We have other examples:
Robbolito 0.09 and Houdini 1.0 or Rybka 4 and Rykba 4 Exp Settings.
Stockfish 2.1.1 and 2.1.1 PHQ.

Optimition in detail could be 100 ELO.
Not old, but never we can produced it.

But its interesting to test it, in combination with older engines which such optimation, like the examples I wrote before. Will give the puzzle one detail more how a program have to work for a better performance. In this case the IvanHoe test is very interesting for myself and perhaps for the persons which working in compiling engines or others.

:-)

A bit to find out the secret how to set the good known sources and settings.

Best
Frank

View previous topic View next topic Back to top  Message [Page 1 of 2]

Goto page : 1, 2  Next

Permissions in this forum:
You cannot reply to topics in this forum