Similarity tests

[kranium]

PS-
If the testers and rating lists would simply put derivatives in a separate list, it would really help
I believe a lot of the anger and animosity from 'original' authors would disappear

And who determines what is a derivative and what isn't.... That is the problem. There is rarely any agreement on this.

I don't see much disagreement Ray...the situation now is clearer than it has ever been, unless you're referring to the endless Rybka debate.
(and that situation is easily resolved: anything newer than Rybka 1.0 can undoubtedly be considered unique and 'original' IMO)

The community has a host of tools and techniques available, and a lot of smart people...
there are many ways to examine an engine today, and a lot of shared knowledge in this area.

Each rating list could adopt a simple 'standard'...ie a set of criteria, requirements that need to be met
(better yet the community could agree on a universal standard)

That's why I applaud Sedat's idea of a combination of similarity testing and ELO improvement...
he's really the 1st to establish a firm set of criteria that derivative authors can aspire to meet, and believe me it won't be easy to do.

Instead of banning/discouraging derivatives, the community should encourage participation and progress,
by sending the signal that it's ok to start with an established codebase (like Stockfish did) if one so chooses.
(kudos to CCRL, CEGT, etc!)

Norm

[Adam Hair]

Sorry, if I bother you, but where can I get the Similarity Detector by Don Daily? The original link from komodochess.com doesn't work anymore.

http://komodochess.com/downloads.htm

[Sedat Canbaz]

Hello Norman,


What a pity that still many chess friends have different (probably wrong) views

Note also that I don't claim that,
All my views are 100% right...of course I can be mistaken in some cases

But I have a little bit experience in chess
And that experience helps me in most cases
And I am doing my best for Computerchess's progress

But, unfortunately there are some people, who are like barrier!!
But SCCT has a solution: + 55% +100 Elo )!
Why they hate a lot of others work ?? i.e who are succeed in our area ?!
They are busy a lot about proving over other engines are clones

Note also,
Without Fruit; Crafty:
- Stockfish, Komodo, Houdini, Gull, Rybka, Chiron etc...would no be so strong as it nowadays!

I am quite sure in that,
- Almost all engines (Crafty, Fruit, Senpai, Zappa, etc...) are not based 100% on own ideas too

And it's not required much work to prove...
Just is needed to test them without opening books
Almost all of them have same or similar opening code!

In my understanding, (if it's 100% original)
Their engines (in case of tests without books) should play different opening moves...!
And I hope to see X engine to start playing, for example B90 )))

So according to my testings,
Mostly of them preferred to play same or similar opening moves (usually up to 4-5 moves)

And after all,
I don't claim that the mentioned engines are not original work...
And I believe that all the mentioned engines belong to the most original ones, but not 100% based on own work and own ideas

Btw,
Do you know why I mainly cancelled my scratch rating ?
-Because my 6th sense said me that some of them are not started 100% from scratch
And I am not going to count which ones, because I am afraid to miss to mention some of them)))
Another reason was that, some of them were too buggy...



Best,
Sedat

[Henk]

PS-
If the testers and rating lists would simply put derivatives in a separate list, it would really help
I believe a lot of the anger and animosity from 'original' authors would disappear

And who determines what is a derivative and what isn't.... That is the problem. There is rarely any agreement on this.

I don't see much disagreement Ray...the situation now is clearer than it has ever been, unless you're referring to the endless Rybka debate.
(and that situation is easily resolved: anything newer than Rybka 1.0 can undoubtedly be considered unique and 'original' IMO)

The community has a host of tools and techniques available, and a lot of smart people...
there are many ways to examine an engine today, and a lot of shared knowledge in this area.

Each rating list could adopt a simple 'standard'...ie a set of criteria, requirements that need to be met
(better yet the community could agree on a universal standard)

That's why I applaud Sedat's idea of a combination of similarity testing and ELO improvement...
he's really the 1st to establish a firm set of criteria that derivative authors can aspire to meet, and believe me it won't be easy to do.

Instead of banning/discouraging derivatives, the community should encourage participation and progress,
by sending the signal that it's ok to start with an established codebase (like Stockfish did) if one so chooses.
(kudos to CCRL, CEGT, etc!)

Norm

If there is a tournament with prices and the winner is a stockfish clone with some minor errors with ELO 2900 and the second place goes to an engine with ELO 2800 where a programmer worked on for twenty years then at least something is not fair.

Might be that clone was created within a month just to pass similarity test.

[kranium]

If there is a tournament with prices and the winner is a stockfish clone with some minor errors with ELO 2900 and the second place goes to an engine with ELO 2800 where a programmer worked on for twenty years then at least something is not fair.

Might be that clone was created within a month just to pass similarity test.

Agreed...unfair,
this scenario sounds like an error in judgement by the tournament director.
(ie: the Stockfish clone should never have been admitted to a prize money tournament of 'original' engines)

[Ralf Müller]

Many thanks!

[Frank Quisinsky]

Hello Sedat,

Sedat:
Firstly Frank started to criticize some of my used participants (Rybka+ Houdini).

Frank:
Never I do that.
I wrote to you "Each one can use what he want".

I wrote "generelly" that I don't like tourneys, engines participants - commercial available - , based on the work by others ... without to see it in a readme file.

Sorry Sedat, that isn't my World and if I am looking in the past ... how many interesting persons lost interest on our Hobby we should go with a good example in front of the others if we have in reality interest on computer chess.

With each celebration of a new clone the community lost more and more people. And that isn't in my interest because I benefit on good works by the others. I can't benefit from the work "cloners" do.

Most important for myself ...
Again and again ...

The readme file.
No readme file or other important Information to an engine ...
No test by myself!

If CEGT or others like to test such engines ...
In Germany we say:

Wenn einer von der Brücke springt, springst Du dann auch?

Oh in my bad English
If anyone jumping from a bridge, do you jumping too?
It seems that in computer chess we have a lot of "Bridge Jumper".

I believe you prefere, Houdini or Rybka. You have a big heart Sedat ... I can read it between the lines. For sure both programmers have a good knowledge to programming. It is agreed upon. But both don't write computer chess history if I can't see any proof about the own works. Playing strength isn't a proof for myself. For that reason that if a programmer are using the work by an other, I can be sure that such a programmer are using the work from a lot of others too.

After all I read in the past to Rybka and Houdini i can't build on other opinion to the topic that both starts with a basic from an other one.

Again ...
It's easy to send an own engine to an other programmer for checking a bit (like that, Critter programmer do that).

I prefere "Gentlemen like" ...
We should protect each honest work and should not say to a programmer ... your work is bad because 30 clones with 300 Elo more are available. Such an information a rating list is building if clones are playing. Or we lost more and more interesting developments.

The programmers of Rybka and Houdini goes _"for me"_ the wrong way. Each programmer, are working on a clone or derivative engine, should have a good readme file. That is very very easy. Children's, 6 years old can write a readme file.

Thanks Fabien ...
Thanks Bob ...
Thanks Tord ...
Thanks ...

That's all!
Should be not a big problem.
If a big problem ... easy for me to build my own opinion!

Best
Frank

[Sedat Canbaz]

Hello Frank,

Let me started with this wise saying:

I don't forgive people because I'm weak, I forgive them
because I am strong enough to know people make mistakes
-Marilyn Monroe


About the current issues,
I fully give you right...and I can understand you very well...

But however,
Everybody can make mistakes...and we should FORGIVE, but of course we should not FORGET !!

For example,
Normally (if I was weak and unprofessional) I should not test Houdini, even Rybka too
Note: both Top engine teams also made injustices over me!!
But anyway.... I will not prefer to go any more deeper over the old stories...))

In shortly my dear friend,
We should not make double standards, and we should be neutral testers!
I kindly recommend you, use my newest SCCT rules...you will be not sorry
And please include again the both white pigeons (Rybka+Houdini) in FCT !

And be sure in this:
In any decision of yours... I will respect you !

And without to not mention this I can't,
You are one the best testers that I have ever seen!



Best Regards,
Sedat

[Frank Quisinsky]

Hi Sedat,

if I included Houdini and Rybka again two much more important engines lost here places.

1. Deuterium
2. Tornado

Never, never, never ...

I am sick to that topic, really.

I have an other idea ...
I delete all engines with more as 2.950 Elo from my list and the topic is over for me. Much more interesting because place 01-21 are only 200-250 Elo different. Much more fun for myself in follow the games at home.

Can be solved very easy!
We have enough Rating lists, others can looking if they search results of clones vs. SF and Komodo. That isn't my World.

Lissen ...
After my qualification tourney my new book will be ready I hope. I will start FCT2 with longer time controls the new book. A new rule ... engines more as 2.950 Elo aren't admitted.

That are much more interesting!
So very nice to see unkown engines for me in the qualification tourney.

Best
Frank

I like Marilyn, must be a sister from Cleopatra!
Each mistake people made are to repair.

Example:
Yes, I am using the sources and ideas from Fabien and made a mistake not to give the information to the othes. Sorry for that, I had money only in my brain and one time in my live I like it to be a super star.

Could be one example!

[Sedat Canbaz]

Hello Frank,

No problem... and I respect your rules)!

And with pleasure I will follow your interesting tournaments as before

Btw, do you remember those days (5-6 months ago)
I mean when I told my story about my racing pigeons ?

In those days,
I was thinking to make SCCT rules: the similarity limit to be +50% + 100 Elo
Then the latest versions like Senpai, Protector etc.
They would not be as SCCT participants, then I would requested +100 Elo to be stronger than Rybka 3

But however,
Later I decided to make SCCT rules easier...


Best,
Sedat

sim version 3
------ Rybka 3 (time: 100 ms scale: 1.0) ------
61.79 RobboLito 0.085g3 w32 (time: 100 ms scale: 1.0)
59.18 Elektro 1.0 (time: 100 ms scale: 1.0)
58.76 Fire 3.0 x64 (time: 100 ms scale: 1.0)
58.45 BlackMamba 2.0 x64 (time: 100 ms scale: 1.0)
57.66 Critter 1.6a 64-bit (time: 100 ms scale: 1.0)
57.65 Equinox 3.20 x64mp (time: 100 ms scale: 1.0)
57.00 Naum 4.6 (time: 100 ms scale: 1.0)
56.98 Murka 3 x64 UCI (time: 100 ms scale: 1.0)
56.48 Houdini 4 x64 (time: 100 ms scale: 1.0)
55.12 Critter 0.90 64-bit SSE4 (time: 100 ms scale: 1.0)
54.87 Rybka 1.0 Beta (time: 100 ms scale: 1.0)
53.93 Fruit 090705 Test Beta (time: 100 ms scale: 1.0)
53.69 Stockfish 2.1 JA 64bit (time: 100 ms scale: 1.0)
53.44 Stockfish 2.1 JA 64bit (time: 50 ms scale: 1.0)
53.29 Strelka 2.0 B (time: 50 ms scale: 1.0)
52.82 Komodo64 2.03 DC (time: 100 ms scale: 1.0)
52.51 Stockfish 1.5 JA 64bit (time: 100 ms scale: 1.0)
52.10 Stockfish 1.7 JA 64bit (time: 100 ms scale: 1.0)
52.05 Senpai 1.0 (time: 100 ms scale: 1.0)
51.69 Stockfish 1.7.1 JA (time: 100 ms scale: 1.0)
51.21 Protector 1.7.0 (time: 100 ms scale: 1.0)
51.08 Komodo 8 64-bit (time: 100 ms scale: 1.0)
50.50 Gull 3 x64 (time: 100 ms scale: 1.0)
50.47 Glaurung 2.2 JA (time: 100 ms scale: 1.0)
50.46 Stockfish 140614 64 SSE4.2 (time: 100 ms scale: 1.0)
49.58 Toga II 3.0 (time: 100 ms scale: 1.0)
49.09 MinkoChess 1.3 x64 (time: 100 ms scale: 1.0)
49.00 Crafty 23.8 x64 (time: 100 ms scale: 1.0)
48.54 Fruit 2.1 (time: 100 ms scale: 1.0)
48.54 Bobcat 3.25 (time: 100 ms scale: 1.0)
47.54 Daydreamer 1.75 JA (time: 100 ms scale: 1.0)
47.29 Octochess revision 5190 (time: 100 ms scale: 1.0)
46.78 TwinFish 0.07 (time: 100 ms scale: 1.0)
46.71 cheng4 0.36a (time: 100 ms scale: 1.0)
46.55 Rodent 1.4 (build 2) (time: 100 ms scale: 1.0)
46.33 Cyrano 0.6b17 (time: 100 ms scale: 1.0)
45.78 Tornado 5.0 x64 SSE4 (time: 100 ms scale: 1.0)
44.32 Vajolet 2.48 (time: 100 ms scale: 1.0)
43.28 EXchess v7.31b x64 (time: 100 ms scale: 1.0)
42.33 Igorrit 0.086v8_x64 (time: 100 ms scale: 1.0)
40.25 Booot 5.2.0(64) (time: 100 ms scale: 1.0)
27.40 Arasan 17.2 (time: 100 ms scale: 1.0)