Movei, The Baron, and TSCP

[Macumba]

I published SCCT Scratch Rating (a few minutes ago):
http://www.sedatcanbaz.com/chess/?page_id=634

Note that so far approx. 2700 games are already played
And the scratch engine database includes 113 games lost on time
That means in every 100 games, there is possibility to see 3-4 games forfeits on time (the number is not too big, honestly I expected much more...)
Those games which are lost on time are mainly belong to Djinn v1.008 (39 games), without those lost games on time, probably Djinn would be in Top 20

Since today, I started testing only the Top 20 scratch engines

Soon I will test a few scratch engines more (Jazz,Milady,Neurone...)

Games and more info coming soon!

Have fun,
Sedat

I see that the baron is included in the scratch engines.
Note that I read that the baron started from tscp.

http://chessprogramming.wikispaces.com/TSCP

Note that I started from some code from tscp but not from some working code(For example the code that I started did not include something of the evaluation function of tscp) so it is not the same for movei.

I do not claim that the version of baron that you test is similiar to tscp
but it may be interesting to know results by the similiarity tool.

Baron is completely original. I suppose this kind of affirmation remembers me about your fantastic move 1. h3!!.

Any real chess programmer, who knows Richard Pijl, not you of course, can affirm without any doubt that new Baron was created from scratch. Your Movei is derived from TSCP because you said it openly here.

I find it so silly, that I suppose Richard will pass this. My goodness, the things that I have to read here!

Note that
I did not say that the new baron include something from tscp but only that baron started from tscp.

From the page that miguel gave:
"Although all of Tom's original code has gone now, it gave me the opportunity to quickly build a working program."

I admit openly that I started from some code of tscp that did not include evaluation function or move generator that it is clearly less than what Richard admitted.

You distort my words when you say that I admit openly that movei is a tscp derivative.

If we define chess code as code that is relevant only for chess and not for other games then
I think that there was never chess code that is common for movei and tscp(unless you consider some names of variables to be chess code).

The common pv array that movei use and also tscp use is not a chess code because it can be used also for other games and I do not consider the way that I store moves to be chess code(originally by struct like tscp and later I replaced it by integer)

I guess that when richard said that all of Tom's original code has gone now he meant only to chess code and I doubt that he replaced every line to see that there is no common lines between the baron and tscp.

No, I don't distort anything. It is you that affirms things without having any base. You said it here that Movei derived from TSCP. There is no distortion. Simple.

As always, you guess wrongly. Richard is an original chess programmer and his code is completely created from scratch, programe that you won't have.

End of story.

[Sedat Canbaz]

First of all,
Many thanks again to anyone for your useful and valuable information

So...I am a Neutral Tester since many years, plus I don't belong to any Chess Engine team...
That's why, I have another new idea regarding solving the current issue...

I think with the new rule, we will stop and solve all injustices over X Engine Programmer

And here is my new idea/rule and I hope you will like it too :

SCCT Scratch Rating's New Rule:
1) X Chess Engine, which is with more than 55 % of similarity will be removed of the current scratch testing
2) X Derivative engine will be allowed of testing, but only in case if the derivative engine is:
- at least 500 Elo points stronger than the free open source engine (for example, derived on TSCP, Fruit, Crafty vs...)
3) Private engines will be not allowed for the current scratch testing

So with the current new rule,
Movei will be as participant and Baron will be left in the current scratch testing


Btw, here are the Elo differences, played in CEGT and CCRL:

Code: Select all

CEGT:                   
The Baron 2.23 w32 	2339
TSCP 1.81e 	        1791

CCRL:
Movei 00.8.438 (10 10 10)	2628
Tscp 1.81	                1706

Thanks for your understanding and best regards,
Sedat Canbaz

[Uri Blass]

I published SCCT Scratch Rating (a few minutes ago):
http://www.sedatcanbaz.com/chess/?page_id=634

Note that so far approx. 2700 games are already played
And the scratch engine database includes 113 games lost on time
That means in every 100 games, there is possibility to see 3-4 games forfeits on time (the number is not too big, honestly I expected much more...)
Those games which are lost on time are mainly belong to Djinn v1.008 (39 games), without those lost games on time, probably Djinn would be in Top 20

Since today, I started testing only the Top 20 scratch engines

Soon I will test a few scratch engines more (Jazz,Milady,Neurone...)

Games and more info coming soon!

Have fun,
Sedat

I see that the baron is included in the scratch engines.
Note that I read that the baron started from tscp.

http://chessprogramming.wikispaces.com/TSCP

Note that I started from some code from tscp but not from some working code(For example the code that I started did not include something of the evaluation function of tscp) so it is not the same for movei.

I do not claim that the version of baron that you test is similiar to tscp
but it may be interesting to know results by the similiarity tool.

Baron is completely original. I suppose this kind of affirmation remembers me about your fantastic move 1. h3!!.

Any real chess programmer, who knows Richard Pijl, not you of course, can affirm without any doubt that new Baron was created from scratch. Your Movei is derived from TSCP because you said it openly here.

I find it so silly, that I suppose Richard will pass this. My goodness, the things that I have to read here!

Note that
I did not say that the new baron include something from tscp but only that baron started from tscp.

From the page that miguel gave:
"Although all of Tom's original code has gone now, it gave me the opportunity to quickly build a working program."

I admit openly that I started from some code of tscp that did not include evaluation function or move generator that it is clearly less than what Richard admitted.

You distort my words when you say that I admit openly that movei is a tscp derivative.

If we define chess code as code that is relevant only for chess and not for other games then
I think that there was never chess code that is common for movei and tscp(unless you consider some names of variables to be chess code).

The common pv array that movei use and also tscp use is not a chess code because it can be used also for other games and I do not consider the way that I store moves to be chess code(originally by struct like tscp and later I replaced it by integer)

I guess that when richard said that all of Tom's original code has gone now he meant only to chess code and I doubt that he replaced every line to see that there is no common lines between the baron and tscp.

No, I don't distort anything. It is you that affirms things without having any base. You said it here that Movei derived from TSCP. There is no distortion. Simple.

As always, you guess wrongly. Richard is an original chess programmer and his code is completely created from scratch, programe that you won't have.

End of story.

Richard did not post here and it may be interesting to read him(I wonder how do you know that I guess wrong and if you asked richard about it).

I am not against the participation of the baron in chess tournaments
and I do not claim that it is not original engine even if it includes few names of consts or variables or arrays or functions from tscp.

My point was only that I think that it is not fair to include the baron and not to include Movei based on Sedat's rules.

After Sedat decided to include both of them
there is no problem from my point of view.

[Sedat Canbaz]

Btw, I plan to add as participants the below engines too (in my latest scratch rating):

Fritz
Chiron
Hannibal
Hiarcs
Sjeng
Shredder
Spike
Junior
SmarThink
Chess Tiger
The King

Note also that for the current scratch test,
I plan to use sometimes one of my i7 980X or i7 970 machines too

Greetings,
Sedat

[Uri Blass]

Btw, I plan to add as participants the below engines too (in my latest scratch rating):

Fritz
Chiron
Hannibal
Hiarcs
Sjeng
Shredder
Spike
Junior
SmarThink
Chess Tiger
The King

Note also that for the current scratch test,
I plan to use sometimes one of my i7 980X or i7 970 machines too

Greetings,
Sedat

I remember reading that Fritz does not pass the similarity tests with the ippo family.

[Sedat Canbaz]

Btw, I plan to add as participants the below engines too (in my latest scratch rating):

Fritz
Chiron
Hannibal
Hiarcs
Sjeng
Shredder
Spike
Junior
SmarThink
Chess Tiger
The King

Note also that for the current scratch test,
I plan to use sometimes one of my i7 980X or i7 970 machines too

Greetings,
Sedat

I remember reading that Fritz does not pass the similarity tests with the ippo family.

Thanks again dear Uri, for letting me know

But are you sure ? can you provide please source (direct link) regarding the similarity of Fritz vs Ippo ?!

[carldaman]

Btw, I plan to add as participants the below engines too (in my latest scratch rating):

Fritz
Chiron
Hannibal
Hiarcs
Sjeng
Shredder
Spike
Junior
SmarThink
Chess Tiger
The King

Note also that for the current scratch test,
I plan to use sometimes one of my i7 980X or i7 970 machines too

Greetings,
Sedat

I remember reading that Fritz does not pass the similarity tests with the ippo family.

Thanks again dear Uri, for letting me know

But are you sure ? can you provide please source (direct link) regarding the similarity of Fritz vs Ippo ?!

There was a thread in Engine Origins:

http://www.talkchess.com/forum/viewtopi ... 44&t=50307

I hope you could include Fritz 10, as that was the last fully original Fritz (from what I gather).

[Sedat Canbaz]

Btw, I plan to add as participants the below engines too (in my latest scratch rating):

Fritz
Chiron
Hannibal
Hiarcs
Sjeng
Shredder
Spike
Junior
SmarThink
Chess Tiger
The King

Note also that for the current scratch test,
I plan to use sometimes one of my i7 980X or i7 970 machines too

Greetings,
Sedat

I remember reading that Fritz does not pass the similarity tests with the ippo family.

Thanks again dear Uri, for letting me know

But are you sure ? can you provide please source (direct link) regarding the similarity of Fritz vs Ippo ?!

There was a thread in Engine Origins:

http://www.talkchess.com/forum/viewtopi ... 44&t=50307

I hope you could include Fritz 10, as that was the last fully original Fritz (from what I gather).

Thanks for the links dear Carl

Tons of data... unfortunately I have no much free time to read all comments...but as far as I noticed some people believe that DF14 derive from Ippo family engines, strange indeed...

Actually I am planning to test Fritz 13 (I don't have Fritz 14)

Of course, I have no any info about what is the similarity percentage comparing F13 or F14 with the rest of the engines, but however I expect Fritz 13 to be less than 55%

Fritz 10 belongs to 2006 year (instead of testing Fritz 10, there are newer releases...)
http://en.wikipedia.org/wiki/Fritz_%28chess%29

In any case, I would pick Fritz 12 or Fritz 13 (F12 was released in 2009, F13 was released in 2011)

If Deep Fritz 13 was derive from Ippo family, then DF13 would be ranked at similar points as the rest Ippo engines, for example I already tested it before:
http://www.sedatcanbaz.com/chess/?page_id=137

Btw, I noticed also that Ippo family engines are ranking mainly in Elo strength between Rybka 3 and Houdini 1.5a

And I hope to see Ippo family engines to be on the similar strenght as Stockfish 070114 and Houdini 4

In other words... I expect DF13 to be clean and original work, of course not sure about the latest DF14 release...

Best,
Sedat

[carldaman]

No problem, Sedat

Many people believe that starting with Fritz 11, Fritz was modeled after Rybka, or a close derivative thereof (not necessarily the latest or best Rybka). I can vouch that Fritz 10 and Fritz 11 have vastly different playing styles. From what I've seen with my own eyes and what I've read in CC forums, I can only consider (Deep) Fritz 10 by Frans Morsch as being original.

The new Fritz 14 may be based on Pandix (private engine) and shows a strong similarity to the Ippo-family. The thread in Engine Origins is probably worth reading, as it contains some interesting information.

Regards,
CL

[Sedat Canbaz]

No problem, Sedat

Many people believe that starting with Fritz 11, Fritz was modeled after Rybka, or a close derivative thereof (not necessarily the latest or best Rybka). I can vouch that Fritz 10 and Fritz 11 have vastly different playing styles. From what I've seen with my own eyes and what I've read in CC forums, I can only consider (Deep) Fritz 10 by Frans Morsch as being original.

The new Fritz 14 may be based on Pandix (private engine) and shows a strong similarity to the Ippo-family. The thread in Engine Origins is probably worth reading, as it contains some interesting information.

Regards,
CL

Hello Carl,

Of course I can be wrong too

As you know, nearly 1.5 years I am out of chess... and in that period of time I notice it a lot of things have been changed ))

I think in a few days it will be more clear about the current issue...

Btw, the current scratch list is very very clean... and I am very happy too that I started testing only the original ones



Best,
Sedat