SWCR2: Final Result, Gaviota 1.0 AVX x64

[Frank Quisinsky]

Hi Sedat,

100% agree ...
BRAVO !!

And unfortunately, we lost so many talented programmer in the past. Main reason is this topic if I read in detail what older programmer mailed me.

In a Rating list we should try to hold a "Limit" to the topic "clones".

I can understand if a programmer made his engine free availabe with open source and is proud about his own work. But this is a bit danger if users gave "clones" to many Attention.

I am very happy with chess wiki, the work the IsiChess programmer do or the grafic by the Fire programmer. On the other Hand, if you look in detail in stats of the games, you can see also a lot.

Best
Frank

[Sedat Canbaz]

Hi Sedat,

yes, that are the points ...
Wrote a lot about it in the Senpai Thread in CSS Forum.

Exactely the problem ...

Will give you an example:
Senpai have problems vs. Smarthink (could be, not sure after 20 games but I think it's possible ...

Without Smarthink after 400 games Senpai got 10 ELO more.

Now, if Smarthink is open source and three other programmer cloned with different changes ... not 10 ELO ... the result will be 30 ELO.

A lot of more fun we have in creating of rating list Information for the others if we are using complety different engines.

And yes ... I have the same opinion to Houdini.
150 more to Robbolite ... the programmer must have knowledge and create a very strong engine for analyzes and so one. On the other Hand looking in Stockfish Framework. The persons try this one and this one and per random one of the changes is good or not. With the result more knowlege about chess programming. If a Person at home have many cores it should be possible to improve such a code like Robbolite with 150 ELO with normaly knowlegde about programming.

But to used an engine by an other, closed the sources and made it commercial ... I have a problem with it, sorry! If others like such things, not my problem.

You can see in your list in the message before ... it isn't easy to find out 20 different engines.

I am sure Critter in Version 0.9 or 0.8 have nothing to do with IPP. I hope that the Fire programmer can add this Versions of Critter in his test / available grafic.

Best
Frank

Yes Frank,

Strange indeed.... if we are testing their engines (Rybkanians) we are hero, if we are not, then they say 'insult' hehe

Actually Rybkanian authors should be happy with SCCT new rule,
because SCCT is allowing derivatives or clones, but there is one very small point : only +100 Elo (not much )

For example, (on my previous scratch testing) I requested +500 Elo over...

BTW,
Let's say... we started testing all strongest engines (it does not matter clone or derivative) then our top 20 engines will be look like a Rybkanian competition,
I mean approx. 90 % of the top 20 engines will be derivatives or clones...

Is it a good idea ?

And what about rest original engines ?!
Who will test their own work engines ?!


Best,
Sedat

[Sedat Canbaz]

Hi Sedat,

100% agree ...
BRAVO !!

And unfortunately, we lost so many talented programmer in the past. Main reason is this topic if I read in detail what older programmer mailed me.

In a Rating list we should try to hold a "Limit" to the topic "clones".

I can understand if a programmer made his engine free availabe with open source and is proud about his own work. But this is a bit danger if users gave "clones" to many Attention.

I am very happy with chess wiki, the work the IsiChess programmer do or the grafic by the Fire programmer. On the other Hand, if you look in detail in stats of the games, you can see also a lot.

Best
Frank

Thank you dear Frank

I respect too your work and decisions!

Have a nice weekend,
Sedat

[Frank Quisinsky]

Hi Sedat,

it's hard for me to killed Houdini in my list.
But I think its better for myself or I lost again my motivation.

Your rules are fully OK for myself. I will not see rules I created in each list on www. But I would be happy if others in working on a Rating list will thinking about it to give "clones" not so many Attention. It's bad for each list and much more important ... I see that daily ... more fun each one will have with different engines.

To your idea:
Yes why not ...
To added all the clones in a list A and create a list B for different engines. This would be a nice Information for each Reader without so many knowledge we have.

If I am speaking for myself ...
I don't have fun on a list with TOP-20 and 90% clones. To many draw games, to high the move average, to many boring endgames and I must pay for it ... means electric current ... for what if I don't like it. For others?

But if you will see the TOP-20 why not. A really good idea in testing to have List A with clones ... clear for each Reader and List B no clones.

Thinking the Long day about it ...
Critter 0.9 in SWCR ... I am to 95% sure no IPP clone.

Best
Frank

[Sedat Canbaz]

Hi Sedat,

it's hard for me to killed Houdini in my list.
But I think its better for myself or I lost again my motivation.

Your rules are fully OK for myself. I will not see rules I created in each list on www. But I would be happy if others in working on a Rating list will thinking about it to give "clones" not so many Attention. It's bad for each list and much more important ... I see that daily ... more fun each one will have with different engines.

To your idea:
Yes why not ...
To added all the clones in a list A and create a list B for different engines. This would be a nice Information for each Reader without so many knowledge we have.

If I am speaking for myself ...
I don't have fun on a list with TOP-20 and 90% clones. To many draw games, to high the move average, to many boring endgames and I must pay for it ... means electric current ... for what if I don't like it. For others?

But if you will see the TOP-20 why not. A really good idea in testing to have List A with clones ... clear for each Reader and List B no clones.

Thinking the Long day about it ...
Critter 0.9 in SWCR ... I am to 95% sure no IPP clone.

Best
Frank

Hello Frank,

+ 100

Just I'd like to add,
I tested many Rybkanians (I lost many CPU time as you) but now it is time for original engines!

About Houdini in your rating list,
Think a little bit please...Houdini 4 deserves to be as participant in all competitions!


Best,
Sedat

[Frank Quisinsky]

Hi Sedat,

Houdini is really a long time topic for myself.
I see the good work Robert do in improvements.

But fact is he started with ... closed sources, goes commercial.

A long fight with myself and the winner is ...
I am the winner
Houdini is out.

What others do isn't important for myself if I am thinking that my decision is right in this case ... hobby only.

And the next one is Rybka I will delete and I think all other participants are OK. Only the question with Critter ... Each Version after 0.9 is more IPP I am thinking if I am looking in the stats the older SWCR1. Not sure about it, only what I can see in my stats. Must not be right, more information needed.

Hope that a newer Sjeng or Jonny will be available for free or Commercial. I must be deleted because both are not available. I will only add available engines in the future in testing.

Very very hard too ... means to replaced Sjeng and Jonny. Very nice engines.

Best
Frank

[Sedat Canbaz]

Hi Sedat,

Houdini is really a long time topic for myself.
I see the good work Robert do in improvements.

But fact is he started with ... closed sources, goes commercial.

A long fight with myself and the winner is ...
I am the winner
Houdini is out.

What others do isn't important for myself if I am thinking that my decision is right in this case ... hobby only.

And the next one is Rybka I will delete and I think all other participants are OK. Only the question with Critter ... Each Version after 0.9 is more IPP I am thinking if I am looking in the stats the older SWCR1. Not sure about it, only what I can see in my stats. Must not be right, more information needed.

Hope that a newer Sjeng or Jonny will be available for free or Commercial. I must be deleted because both are not available. I will only add available engines in the future in testing.

Very very hard too ... means to replaced Sjeng and Jonny. Very nice engines.

Best
Frank

Hello again Frank

You are planning to disqualify Rybka too ??

Wow...then your tournament will look like WCCC

What about Komodo?
For example, if I follow your strictly rules, I should disqualify Komodo too


Note that I dont claim Komodo is a clone or derivative engine
But I can't see any difference between Rybka and Houdini...

Komodo was free and later become commercial work, same thing like Houdini...

In my opinion,
Komodo also includes some codes/ideas from other engines

Code: Select all

sim version 3
------ Komodo32 1.1 JA (depth: 2, scale: 1.0) ------
79.97  Komodo32 1.2 JA (depth: 2, scale: 1.0)
77.17  Komodo32 1.0 JA (Doch32) (depth: 2, scale: 1.0)
76.72  Doch32 1.3.4 JA (depth: 2, scale: 1.0)
75.81  Doch32 1.2 JA (depth: 2, scale: 1.0)
75.41  Komodo32 1.3 JA (depth: 2, scale: 1.0)
61.63  Strelka R-3-E (Rybka-3-Eval) (depth: 2, scale: 1.0)
58.34  IPPOLIT 0.080a (depth: 2, scale: 1.0)
58.32  Komodo32 2.03 JA (depth: 2, scale: 1.0)
58.32  Belka 2.0.0 (depth: 2, scale: 1.0)
58.19  Strelka 2.0 B (depth: 2, scale: 1.0)
57.89  RobboLito verzione 085F1 w32 (depth: 2, scale: 1.0)
57.84  Murka 3 w32 UCI (depth: 2, scale: 1.0)
57.82  RobboLito 0.085g3 w32 (depth: 2, scale: 1.0)
57.79  RobboLito 0.09 w32 (depth: 2, scale: 1.0)
57.62  IPPOLIT 0.080b w32 (depth: 2, scale: 1.0)
57.05  IvanHoe 9.76b w32 (depth: 2, scale: 1.0)
57.03  Igorrit 0.086v8_w32 (depth: 2, scale: 1.0)
56.98  Vitruvius_1.11C_HEM_x32 (depth: 2, scale: 1.0)
56.72  Naum 4.2 (depth: 2, scale: 1.0)
56.72  Naum 4 (depth: 2, scale: 1.0)
56.71  Naum 4.1 (depth: 2, scale: 1.0)
56.70  Houdini 1.5 w32 (depth: 2, scale: 1.0)
56.66  LEOpard 0.7c (depth: 2, scale: 1.0)
56.58  Strelka 1.0 Beta (depth: 2, scale: 1.0)
56.54  Houdini 1.01 w32 1_CPU (depth: 2, scale: 1.0)
56.53  Houdini 1.02 w32 1_CPU (depth: 2, scale: 1.0)
56.47  Strelka 1.8 UCI (depth: 2, scale: 1.0)
56.36  Houdini 2.0c w32 (depth: 2, scale: 1.0)
56.36  Critter 1.2 32-bit (depth: 2, scale: 1.0)
56.25  Fire 1.31 w32 (depth: 2, scale: 1.0)
56.22  Daydreamer 1.75 JA (depth: 2, scale: 1.0)
56.20  Critter 1.0 32-bit (depth: 2, scale: 1.0)
56.15  Houdini 1.03a w32 1_CPU (depth: 2, scale: 1.0)
56.15  Houdini 1.03 w32 1_CPU (depth: 2, scale: 1.0)
56.12  Fire 1.5 xTreme w32 (depth: 2, scale: 1.0)
56.09  Robodini 1.1 (depth: 2, scale: 1.0)
56.05  IvanHoe 9.46b w32 (depth: 2, scale: 1.0)
55.97  IvanHoe 9.47b w32 (depth: 2, scale: 1.0)
55.91  IvanHoe T0.5.4.1 (depth: 2, scale: 1.0)
55.91  IvanHoe 9.50b w32 (depth: 2, scale: 1.0)
55.90  IvanhoeB50kBX32 (depth: 2, scale: 1.0)
55.85  Fire 2.2 xTreme w32 (depth: 2, scale: 1.0)
55.81  Critter 1.6 32-bit (depth: 2, scale: 1.0)
55.80  Fruit 2.1 (depth: 2, scale: 1.0)
55.74  Vitruvius_1.11H_HEM_x32 (depth: 2, scale: 1.0)
55.74  Strelka 5.5 x32 (depth: 2, scale: 1.0)
55.67  DeepSaros version 4.1.6 (depth: 2, scale: 1.0)
55.66  Fire 3.0 w32 (depth: 2, scale: 1.0)
55.62  Houdini 3 (depth: 2, scale: 1.0)
55.60  Vitruvius version 1.0C (depth: 2, scale: 1.0)
55.58  Komodo32 3 32bit (depth: 2, scale: 1.0)
55.57  Bouquet 1.8 x32 (depth: 2, scale: 1.0)
55.54  DeepSaros.exe ver.2.3c3 (depth: 2, scale: 1.0)
55.51  Houdini 4 (depth: 2, scale: 1.0)
55.43  Vitruvius version 1.0H (depth: 2, scale: 1.0)
55.39  Komodo32 Version 4 (depth: 2, scale: 1.0)
55.09  Critter 1.4 32-bit (depth: 2, scale: 1.0)

Actually there is no any engine which is 100% based on own ideas...

Even many original engines have same opening code,
I mean when we are testing the engines without books, some of them has same or similar in starting opening positions

It is already proven:
SCCT's 3 Type Tournaments:
http://sedatchess.110mb.com/index.php?p=1_32

SCCT Swiss No Book:
http://www.sedatcanbaz.com/chess/?page_id=1113


Why ???
if they are 100 % original... they should play different openings

Plus,
Zappa, Hiarcs, Chiron (under Arena) have same MP bug... Why ???


In other words,
There is no any engine, which is 100% original

And to be honest
For me it more important strength than originality!!

And once more I'd like to say,
The authors of Stockfish,Rybka,Houdini,Komodo,Gull... (like many others) are genius, masters in chess engine programing!!

BTW,
Yes...you are right, chess is our hobby and as I mentioned before,
I will respect your decision in both cases



Good luck my friend,
Sedat

[Sedat Canbaz]

My apologies to the authors of Don

Don engine seems to be very clean and original!

This issue happened due to I 've seen wrongly the matrix crosstable

And here are the right results:

Code: Select all

sim version 3
------ DON 1.0b x64 (time: 100 ms  scale: 1.0) ------
 63.56  Stockfish 070114 64 SSE4.2 (time: 100 ms  scale: 1.0)
 56.36  Stockfish 2.0 JA 64bit (time: 100 ms  scale: 1.0)
 53.08  TwinFish 0.07 (time: 100 ms  scale: 1.0)
 52.33  Rybka 41 (time: 100 ms  scale: 1.0)
 51.61  Houdini 4 x64 (time: 100 ms  scale: 1.0)
 51.34  Fire 3.0 x64 (time: 100 ms  scale: 1.0)
 50.98  Komodo TCECr 64-bit  (time: 100 ms  scale: 1.0)
 50.92  Senpai 1.0 (time: 100 ms  scale: 1.0)
 50.86  DiscoCheck 5.2 (time: 100 ms  scale: 1.0)
 50.80  IvanHoe-Beta 999946h6 x64 Tr (time: 100 ms  scale: 1.0)
 50.49  Bouquet 1.8 x64 (time: 100 ms  scale: 1.0)
 50.46  Rybka 3  (time: 100 ms  scale: 1.0)
 50.18  Equinox 2.02 x64mp (time: 100 ms  scale: 1.0)
 49.96  Komodo64 SSE Version 4 (time: 100 ms  scale: 1.0)
 48.80  Protector 1.6.0 x64 (time: 100 ms  scale: 1.0)
 48.74  Komodo64 1.3 JA (time: 100 ms  scale: 1.0)
 48.69  spark-1.0 (time: 100 ms  scale: 1.0)
 48.45  Hannibal 1.4bx64 (time: 100 ms  scale: 1.0)
 47.90  Deep Shredder 12 x64 (time: 100 ms  scale: 1.0)
 47.66  Chiron 2 64bit (time: 100 ms  scale: 1.0)
 46.48  Spike 1.4 (time: 100 ms  scale: 1.0)
 46.47  Deep HIARCS 14 WCSC (time: 100 ms  scale: 1.0)
 46.09  HIARCS 11.2 MP (time: 100 ms  scale: 1.0)
 44.89  Chess Tiger 2007.1  (time: 100 ms  scale: 1.0)
 44.32  Gaviota v1.0 (time: 100 ms  scale: 1.0)
 43.68  Fruit 2.1 (time: 100 ms  scale: 1.0)
 42.74  Shredder 9 UCI (time: 100 ms  scale: 1.0)
 41.73  Texel 1.03 64-bit (time: 100 ms  scale: 1.0)
 41.05  Ruffian 2.1.0 (time: 100 ms  scale: 1.0)
 38.66  Zappa Mexico II (time: 100 ms  scale: 1.0)
 33.13  Deep Sjeng WC2008 x64 (time: 100 ms  scale: 1.0)

[Frank Quisinsky]

I can see same mistakes again and again in different engines daily. From this point of view we need the strongest only But if you are looking in Detail you can see that we have different engines in playing style. And the most of this group of engines have nothing to do with clones.

Yes Rybka will be out soon and I too if I am ready with the world.

Best
Frank

[Sedat Canbaz]

I can see same mistakes again and again in different engines daily. From this point of view we need the strongest only But if you are looking in Detail you can see that we have different engines in playing style. And the most of this group of engines have nothing to do with clones.

Yes Rybka will be out soon and I too if I am ready with the world.

Best
Frank

Dear Frank,


Its ok...no problem from my side


Btw, I have a new idea...

But however, still I am not sure (for the future) about my next SCCT tournaments

I mean,
Maybe I will buy only those commercial engines which are stronger than free Stockfish !


What is your opinion about my new idea ?

Best,
Sedat