A new engine from Russia Strelka rated abobe 2700

[Isaksen]

Hi Thorsten

In both of these tournaments I have seen Chess Tiger produce some very nice results.
I wonder why CCRL team can't make it work under their conditions.
Under their tests Chess Tiger 2007.1 is clearly behing the 2007 version, that is why I am happy with 2007.1. performance so far.
It clearly shows that Christophe did manage to improve his engine.
I just wish that CCRL team has such results as well.
_______________________
Friendly regards from Brian

[mclane]

i guess tiger has a unique algorithm to find out he is at home running on my machines

[Tony Thomas]

i guess tiger has a unique algorithm to find out he is at home running on my machines

Tiger performs very good on my machines as well. Provided that I do not have Fritz, Junior, Zap or the latest version of List, if the other rating lists get a result as good as mine, then it should be somewhere in top 10-14 place. I checked the other rating lists, and that is exactly were tiger is placed in single CPU category. I didnt get better results for the new version of Tiger, but it is well within the error bars, so it might gain points as time goes on from more games.

Code: Select all

7 Chess Tiger 2007               : 2728   52  52   132    61.7 %   2645   26.5 %
11 ChessTiger2007.1 UCI           : 2715   59  58   104    60.1 %   2644   26.0 %

[Rubinus]

If I were You I wouldn't bet too much money on it.
See the difference between Rybka 1.0 beta and Strelka. You can see that they are using almost same usage of memory. My guess is that it's probably a Rybka clone.
hxxxxxx://foto.radikal.ru/f.aspx?i=8f49b92302524863b0f507422b3a0330

your link does not work
It works only after replacing hxxxxxx by http

Note that I do not know if it is a clone but it seems that the author want people to believe that it is a clone(otherwise I see no reason for the choice of the name that end with "ka" and why the 1 beta in his website).

Uri

[tiger]

Hi Thorsten

In both of these tournaments I have seen Chess Tiger produce some very nice results.
I wonder why CCRL team can't make it work under their conditions.
Under their tests Chess Tiger 2007.1 is clearly behing the 2007 version, that is why I am happy with 2007.1. performance so far.
It clearly shows that Christophe did manage to improve his engine.
I just wish that CCRL team has such results as well.
_______________________
Friendly regards from Brian

I do not understand your comment.

Chess Tiger 2007.1 is ahead of Chess Tiger 2007 in the CCRL 40/4, and given the error bars it is highly probable that the result would not change with a larger number of games.

In CCRL 40/40, not enough games have been played so far.


// Christophe

[Spock]

That's correct, at 40/40 the game count is so small the result should be ignored for the moment

[Graham Banks]

I do not understand your comment.

Chess Tiger 2007.1 is ahead of Chess Tiger 2007 in the CCRL 40/4, and given the error bars it is highly probable that the result would not change with a larger number of games.

In CCRL 40/40, not enough games have been played so far.


// Christophe

There should be quite a few more games for Chess Tiger 2007.1 in our 40/40 list over the coming weeks Christophe as we have some gauntlets in progress.

Regards, Graham.

[tiger]

I do not understand your comment.

Chess Tiger 2007.1 is ahead of Chess Tiger 2007 in the CCRL 40/4, and given the error bars it is highly probable that the result would not change with a larger number of games.

In CCRL 40/40, not enough games have been played so far.


// Christophe

There should be quite a few more games for Chess Tiger 2007.1 in our 40/40 list over the coming weeks Christophe as we have some gauntlets in progress.

Regards, Graham.

Great! I was wondering what was going on.


// Christophe

[mclane]

more results with strelka and other new engines:

Code: Select all

    Motor                 Punkte   Lo  Ry  Fr  Sp  Hi  To  Pr  Ch  St  Fr  Na  Gl  Co  Co    S-B
01: LoopMP 12.32          16,5/26 ··· ==  0=  =1= 1=  =0  10  01  0   11  =1= 1   111 1    167,25
02: Rybka v2.3.lk.w32     15,5/23 ==  ··· 10  10  11  ==  11  =?  1=  0=  11  ==  1   1    169,50
03: Fruit-061115a [002]   15,0/22 1=  01  ··· ==  10  11  1   =   =   ==  =1  11  ==  1    162,50
04: Spike1.2              14,5/26 =0= 01  ==  ··· =0  ==  =11 =   0   11  =1  ==  1=  ==   153,00
05: Hiarcs11.1UCI         13,5/23 0=  00  01  =1  ··· 1=  =0  =   1   =1  1=  10  11  1    136,50
06: TogaII1.3x4 [default] 12,0/23 =1  ==  00  ==  0=  ··· =1  0   0   ==  =1  11  10  1    126,75
07: ProDeo 1.2            11,0/24 01  00  0   =00 =1  =0  ··· 0   0   10  011 =   111 1    104,00
08: ChessTiger2007.1 UCI  10,5/16 10  =?  =   =   =   1   1   ··· ==  =   =1  1   =   1    113,25
09: Strelka1.0beta        10,5/16 1   0=  =   1   0   1   1   ==  ··· =   10  1   1   1    111,75
10: Fruit Leiden2006      10,0/23 00  1=  ==  00  =0  ==  01  =   =   ··· 00  10  11  1    101,50
11: Naum 2.0              9,5/28  =0= 00  =0  =0  0=  =0  100 =0  01  11  ··· 10  01  00   103,25
12: Glaurung1.2.1-32bit   7,0/22  0   ==  00  ==  01  00  =   0   0   01  01  ··· =0  01    76,25
13: Colossus2006f         5,5/24  000 0   ==  0=  00  01  000 =   0   00  10  =1  ··· 0     59,50
14: Colossus2007a         5,0/16  0   0   0   ==  0   0   0   0   0   0   11  10  1   ···   46,00

156 Partien von 182 gespielt
Name des Turniers: november
Ort/ Land: SCW, Germany
Spielstufe: Turnier 40/20

AMD Sempron 3000+, pb off

Code: Select all

    Motor                 Punkte   Ry  Hi  Lo  St  To  To  Sp  Ch  Fr  De  Na  Ph  Co  No    S-B
01: Rybka v2.3.lk.x64     24,5/31 ··· 11  10= 1=  11  =0= 1=  =1= 111 111 1=  11  11  =1   336,75
02: Hiarcs11.1UCI         21,5/32 00  ··· 011 00  111 01= ==  =11 110 11= =1  =1  11  11   278,00
03: LoopMP 12.32          18,0/31 01= 100 ··· 10  0=1 === 10  0=1 01  11  ==  =1  =1  11   247,75
04: Strelka1.0beta        17,5/26 0=  11  01  ··· 10  =1  ==  0=  01  11  11  1=  1=  11   232,00
05: TogaII1.2.1a          17,5/31 00  000 1=0 01  ··· ==  ==  11= 010 =1= 11  11  =1  11   212,50
06: TogaII1.3x4 [default] 17,0/31 =1= 10= === =0  ==  ··· ==  011 10= =1  =0  10  1=  1=   250,50
07: Spike1.2              15,5/26 0=  ==  01  ==  ==  ==  ··· 01  =1  10  11  0=  11  11   197,75
08: ChessTiger2007.1 UCI  15,0/33 =0= =00 1=0 1=  00= 100 10  ··· 10  ==0 ==  1=  =1  011  199,00
09: Fruit-061115a [001]   14,5/31 000 001 10  10  101 01= =0  01  ··· =10 10  ==  11  ==   194,75
10: DeepSjeng25           13,0/31 000 00= 00  00  =0= =0  01  ==1 =01 ··· 1=  11  01  11   153,75
11: Naum2.0               10,5/26 0=  =0  ==  00  00  =1  00  ==  01  0=  ··· =0  11  11   125,75
12: Pharaon               10,5/26 00  =0  =0  0=  00  01  1=  0=  ==  00  =1  ··· 11  1=   122,50
13: Colossus2007a         5,0/26  00  00  =0  0=  =0  0=  00  =0  00  10  00  00  ··· =1    61,50
14: Now2404               4,0/27  =0  00  00  00  00  0=  00  100 ==  00  00  0=  =0  ···   58,00

204 Partien von 364 gespielt
Name des Turniers: Ct2007.1
Ort/ Land: SCW, Deutschland
Spielstufe: Turnier 40/20

and

Code: Select all

    Motor                Punkte  Ry Hi Ch De To Sp St Gl Ph Ka Co No    S-B
01: Rybka v2.3.lk.x64    15,0/18 ·· 10 =1 11 =1 11 11 11 =  =  1  1   138,75
02: Hiarcs11.1UCI        13,0/18 01 ·· 1  =  1= =  1= 1  == 1= =1 11  106,50
03: ChessTiger2007.1 UCI 12,0/19 =0 0  ·· 10 =  1= == 1  11 1= 11 01   93,00
04: DeepSjeng25          11,5/19 00 =  01 ·· =1 1= == 0= 1  11 1  1=   93,00
05: TogaII1.3x4          11,5/19 =0 0= =  =0 ·· 01 1  == 1= 11 11 1    89,25
06: Spike1.2             9,5/18  00 =  0= 0= 10 ·· =1 11 1  0  =  11   73,50
07: Strelka1.0beta       9,5/18  00 0= == == 0  =0 ·· =1 1  1  11 1    72,50
08: Glaurung121-EM64T    7,5/19  00 0  0  1= == 00 =0 ·· 0= =1 =  11   55,00
09: Pharaon              7,0/18  =  == 00 0  0= 0  0  1= ·· 01 10 1=   53,25
10: Kaissa_18_a          7,0/19  =  0= 0= 00 00 1  0  =0 10 ·· 01 11   50,25
11: Colossus2007a        5,0/18  0  =0 00 0  00 =  00 =  01 10 ·· 1=   32,75
12: Now2404              2,5/19  0  00 10 0= 0  00 0  00 0= 00 0= ··   23,75

111 Partien von 132 gespielt
Name des Turniers: test
Ort/ Land: SCW, Germany
Spielstufe: Turnier 40/20

both on INTEL 6300, 1 core, pb off

Code: Select all

   Motor                 Punkte  Hi De To St Ch Na No Co    S-B
1: Hiarcs11.1UCI         10,5/13 ·· =  0= 11 =1 11 11 11   53,00
2: DeepSjeng25           8,5/13  =  ·· == 10 10 1= 11 =1   43,50
3: TogaII1.3x4 [default] 8,0/13  1= == ·· 0  =0 10 11 11   42,00
4: Strelka 1.0b          7,5/13  00 01 1  ·· =1 =1 == =1   40,00
5: ChessTiger2007.1 UCI  6,5/13  =0 01 =1 =0 ·· 0  11 ==   36,50
6: Naum 2.0              5,5/12  00 0= 01 =0 1  ·· 1= 1    28,25
7: Now2404               2,5/13  00 00 00 == 00 0= ·· 1    12,25
8: Colossus2007a         2,0/12  00 =0 00 =0 == 0  0  ··   14,50

51 Partien von 140 gespielt
Name des Turniers: Mai
Ort/ Land: ORION8, Germany
Spielstufe: Turnier 40/40

AMD Athlon 2400+, pb off

-----------------------
Not only Strelka is doing a good job in those tournaments, i am also
surprised how strong the new Sjeng2.5 is.

More games soon.