A New IvanHoe Version!!! v Rybka 4.1

[geots]

For today's match we have the newest IvanHoe version release. This match features the "46f compile" by Peter Pan. Tomorrow's match will introduce the "46f compile" by KLO- to determine or at least get some idea of the comparisions between them.

With having different match lengths, I have settled on all "50 game matches" run at 40/3 repeating. All engines like this time limit, and when I wake up after 8 hours sleep, it will in general have run 46 or so of the 50 games.

I am using Rybka 4.1 as a baseline, and comparing each engine I run to that baseline. So the elo difference between 2 engines as it shows in a rating list will be determined by how each performs ag. Rybka. Example: "engine A" shows +70 elo in match with Rybka, and "engine B" shows +35 elo difference. That means in the list "engine A" will list as 35 elo stronger than "engine B".

This system is about the most UNSCIENTIFIC a person could come up with, and is full of enough holes to sink the Titanic. Just starting with too few games- it is most likely the way you would tell someone NOT to set up a testing method to form a rating list. BUT, you CAN count on reliable results from each match. That I promise.

So please no replies about how unreliable this idea is, why it is that way, and how to get reliable results. Right now I am just not interested. I am having a hell of a lot of fun- and that is all that counts. Tho I am happy to share my results. This little tiny, insignificant project is for my pleasure , and if anyone wants some kind of idea to at least start with in determining the IvanHoe versions best to download and try- it will give you a good idea about that- if nothing else.


12/9/2011


1CPU/32bit
128MB hash
Bases=NONE (Rybka used the Nalimovs)
Ponder_Learning=OFF
Perfect 12.32 book w/12-move limit
40/3 Repeating
Match=50 games


IvanHoe 46f........+42 - +17/-11/=22 - 28.0/50
Rybka 4.1............-42 - +11/-17/=22 - 22.0/50


Results using the 64bit versions will also be introduced soon to give you a broader picture or idea- an idea of exactly WHAT I haven't a clue.

So many people are wanting the PGNs- I may have to get someone to help in sending them out. Again, my address is in some of these match posts- you just have to look.


Best,

george

[Dr.Wael Deeb]

Thanks George

The new version of IvanHoe seems pretty strong indeed....
Testing it myself regards,
Dr.D

[Tomcass]

This is my first test (300 games) with this strong PeterPan's version. Large pages allowed.

GUI: Fritz 12
i7 9750 3.33 Mhz.
4 cores no HT
Book: Masterbook 3.1
Time control: 2 min + 1 sec / game.

201112IvanB46bPPLP 2011

Ivanhoe B46fb x64(x4)LP - Critter 1.2 64-bit SSE4(x4) 55.0 - 45.0 +21/=68/-11 55.00%
Ivanhoe B46fb x64(x4)LP - Deep Rybka 4.1 SSE42 x64(x4) 64.0 - 36.0 +37/=54/-9 64.00%
Ivanhoe B46fb x64(x4)LP - Houdini 2.0c Pro x64(x4) 39.0 - 61.0 +12/=54/-34 39.00%

I will start repeating this test at a longer Time Control. (60 minutes per game).

Regards,

Tom.

[geots]

This is my first test (300 games) with this strong PeterPan's version. Large pages allowed.

GUI: Fritz 12
i7 870 4 cores no HT
Book: Masterbook 3.1
Time control: 2 min + 1 sec / game.

201112IvanB46bPPLP 2011

Ivanhoe B46fb x64(x4)LP - Critter 1.2 64-bit SSE4(x4) 55.0 - 45.0 +21/=68/-11 55.00%
Ivanhoe B46fb x64(x4)LP - Deep Rybka 4.1 SSE42 x64(x4) 64.0 - 36.0 +37/=54/-9 64.00%
Ivanhoe B46fb x64(x4)LP - Houdini 2.0c Pro x64(x4) 39.0 - 61.0 +12/=54/-34 39.00%

I will start repeating this test at a longer Time Control. (60 minutes per game).

Regards,

Tom.

Hi Tom. Haven't seen your PGNs , but you are evidently getting extremely accurate results. So keep going- you are doing a great job. And your 60 min. matches are very important- as they will give us an idea at more different time limits. You are doing great- keep up the good work.


Best,
george

[Tomcass]

Thanks George. Best regards from Barcelona.

Tom.

[Sylwy]

For today's match we have the newest IvanHoe version release. This match features the "46f compile" by Peter Pan. Tomorrow's match will introduce the "46f compile" by KLO- to determine or at least get some idea of the comparisions between them.
george

Hello Monsieur Georges !

I don't understand why not a test for the last version of Ivan Hood !?!
The h version of this sublime chess engine developped in Eastern Europe and compiled in South Africa (a Mr.Izak compilation) is here:

http://www.spaadyshare.com/4FOXHFARJQ0R/Ivanhoe_46h.rar

Ivan Ivanovici Ivanovski sublime tester regards,
SilvianR

PS:tonight the results of my comparative test of Bagatur 1.2 c & d - original compiles !

[geots]

For today's match we have the newest IvanHoe version release. This match features the "46f compile" by Peter Pan. Tomorrow's match will introduce the "46f compile" by KLO- to determine or at least get some idea of the comparisions between them.
george

Hello Monsieur Georges !

I don't understand why not a test for the last version of Ivan Hood !?!
The h version of this sublime chess engine developped in Eastern Europe and compiled in South Africa (a Mr.Izak compilation) is here:

http://www.spaadyshare.com/4FOXHFARJQ0R/Ivanhoe_46h.rar

Ivan Ivanovici Ivanovski sublime tester regards,
SilvianR

PS:tonight the results of my comparative test of Bagatur 1.2 c & d - original compiles !

Hi Ruxy. The "h versions" are coming next. I have missed doing any testing since my last results as I got a new dual core 64 bit system in. Networking them, loading stuff, and then finding out they will have to replace a video card or hard drive- or both. Waiting on a service man to arrive- I will be testing one of the "h versions" and hopefully posting results tomorrow.


Best,

george

[Sylwy]

For today's match we have the newest IvanHoe version release. This match features the "46f compile" by Peter Pan. Tomorrow's match will introduce the "46f compile" by KLO- to determine or at least get some idea of the comparisions between them.
george

Hello Monsieur Georges !

I don't understand why not a test for the last version of Ivan Hood !?!
The h version of this sublime chess engine developped in Eastern Europe and compiled in South Africa (a Mr.Izak compilation) is here:

http://www.spaadyshare.com/4FOXHFARJQ0R/Ivanhoe_46h.rar

Ivan Ivanovici Ivanovski sublime tester regards,
SilvianR

PS:tonight the results of my comparative test of Bagatur 1.2 c & d - original compiles !

Hi Ruxy. The "h versions" are coming next. I have missed doing any testing since my last results as I got a new dual core 64 bit system in. Networking them, loading stuff, and then finding out they will have to replace a video card or hard drive- or both. Waiting on a service man to arrive- I will be testing one of the "h versions" and hopefully posting results tomorrow.


Best,

george

Hi Maestro George !

From yesterday night this wonderful PP compile is under tests on 2 PCs. Unfortunately I haven't (yet ) a 64 bit OS. So my tests are only 32 bit single and quad !
First results in this weekend !
Here are the KLO compiles of IH 46h version together with an interesting readme.pdf:
http://chess.cygnitec.com/engine/ivanho ... 999946h.7z

Regards,
SilvianR

[geots]

For today's match we have the newest IvanHoe version release. This match features the "46f compile" by Peter Pan. Tomorrow's match will introduce the "46f compile" by KLO- to determine or at least get some idea of the comparisions between them.
george

Hello Monsieur Georges !

I don't understand why not a test for the last version of Ivan Hood !?!
The h version of this sublime chess engine developped in Eastern Europe and compiled in South Africa (a Mr.Izak compilation) is here:

http://www.spaadyshare.com/4FOXHFARJQ0R/Ivanhoe_46h.rar

Ivan Ivanovici Ivanovski sublime tester regards,
SilvianR

PS:tonight the results of my comparative test of Bagatur 1.2 c & d - original compiles !

Hi Ruxy. The "h versions" are coming next. I have missed doing any testing since my last results as I got a new dual core 64 bit system in. Networking them, loading stuff, and then finding out they will have to replace a video card or hard drive- or both. Waiting on a service man to arrive- I will be testing one of the "h versions" and hopefully posting results tomorrow.


Best,

george

Hi Maestro George !

From yesterday night this wonderful PP compile is under tests on 2 PCs. Unfortunately I haven't (yet ) a 64 bit OS. So my tests are only 32 bit single and quad !
First results in this weekend !
Here are the KLO compiles of IH 46h version together with an interesting readme.pdf:
http://chess.cygnitec.com/engine/ivanho ... 999946h.7z

Regards,
SilvianR

Thanks Ruxy, but I have KLO's compile and PeterPan's loaded also. I will run the PP compile now, and then follow with KLO's compile, both ag. Rybka 4.1 . Then I can compare results.


Best,

george

[gerold]

They are all about the same George. I tested them.
Also tested Houdini 1.5 vs. Houdini 1.5-TC. They are also
the same. No improvement in 1.5-TC as reported.

Best,
Gerold.

P.S. Looking for something else to test now.
Until the clones come up with a better version i will test
something else. :_)