Is it time to make chess tiger free engine

[Tony Thomas]

@Tony:

Last time i checked the new version did beat CT2007. But i'll let you figure that out

No way, I am afraid that Sjeng isnt that close to Chesstiger in strength.

[Uri Blass]

@Tony:

Last time i checked the new version did beat CT2007. But i'll let you figure that out

No way, I am afraid that Sjeng isnt that close to Chesstiger in strength.

Sjeng seems to be at similiar strength to tiger based on rating lists

from CCRL:

Deep Sjeng 2.5 1CPU 2852 +128 −120 65.0% −91.6 40.0% 20
51.2%
Chess Tiger 2007 2791 +17 −17 44.4% +37.6 40.2% 1172
55.6%

from CEGT:

90 Chess Tiger 2007 2726 15 15 1260 50.2 % 2724 38.7 %
91 Deep Sjeng 2.5 2CPU 2726 22 22 640 46.6 % 2750 36.2 %
92 Chess Tiger 2007.1 2724 17 17 977 46.5 % 2749 38.6 %
94 Deep Sjeng 2.5 1CPU 2721 38 38 196 47.4 % 2739 39.8 %
101 Chess Tiger 2007 Gambit 2706 41 41 176 41.8 % 2764 36.9 %

[Tony Thomas]

@Tony:

Last time i checked the new version did beat CT2007. But i'll let you figure that out

No way, I am afraid that Sjeng isnt that close to Chesstiger in strength.

Sjeng seems to be at similiar strength to tiger based on rating lists

from CCRL:

Deep Sjeng 2.5 1CPU 2852 +128 −120 65.0% −91.6 40.0% 20
51.2%
Chess Tiger 2007 2791 +17 −17 44.4% +37.6 40.2% 1172
55.6%

from CEGT:

90 Chess Tiger 2007 2726 15 15 1260 50.2 % 2724 38.7 %
91 Deep Sjeng 2.5 2CPU 2726 22 22 640 46.6 % 2750 36.2 %
92 Chess Tiger 2007.1 2724 17 17 977 46.5 % 2749 38.6 %
94 Deep Sjeng 2.5 1CPU 2721 38 38 196 47.4 % 2739 39.8 %
101 Chess Tiger 2007 Gambit 2706 41 41 176 41.8 % 2764 36.9 %

It is possible that Sjeng performs stronger at X moves in Y minutes time control. In 1m+1sec the performance wasnt that convincing but strong enough for me to say it is indeed a decent improvement.

[chessfurby]

Well some peeps think that chess == bullet. For your sake i hope that you do not think that way about other things Quickies can be fun but the long battles are the real challenge
Nevertheless, Sjeng never was (and prolly never will be?) so good in blitz. However its nice that you peeps enjoy the improved version and have fun with it. I think GC will be happy about this.

Cheers

[Tony Thomas]

Well some peeps think that chess == bullet. For your sake i hope that you do not think that way about other things Quickies can be fun but the long battles are the real challenge
Nevertheless, Sjeng never was (and prolly never will be?) so good in blitz. However its nice that you peeps enjoy the improved version and have fun with it. I think GC will be happy about this.

Cheers

I have not yet seen any engine that gained more than 50(Zapchess I believe) points from longer time controls. I will wait for more results from CCRL and CGET before I say anything else. That quickie comment was hilarious by the way.

[Uri Blass]

Well some peeps think that chess == bullet. For your sake i hope that you do not think that way about other things Quickies can be fun but the long battles are the real challenge
Nevertheless, Sjeng never was (and prolly never will be?) so good in blitz. However its nice that you peeps enjoy the improved version and have fun with it. I think GC will be happy about this.

Cheers

I have not yet seen any engine that gained more than 50(Zapchess I believe) points from longer time controls. I will wait for more results from CCRL and CGET before I say anything else. That quickie comment was hilarious by the way.

I think that the difference between your time control and ccrl 40/40 may be bigger relative to the difference between ccrl 40/4 and ccrl 40/40

As far as I remember the program that earns the most from longer time control is the baron1.7.0 and it earned almost 100 elo when you compare ccrl 40/40 and ccrl 40/4

Uri

[Tony Thomas]

Well some peeps think that chess == bullet. For your sake i hope that you do not think that way about other things Quickies can be fun but the long battles are the real challenge
Nevertheless, Sjeng never was (and prolly never will be?) so good in blitz. However its nice that you peeps enjoy the improved version and have fun with it. I think GC will be happy about this.

Cheers

I have not yet seen any engine that gained more than 50(Zapchess I believe) points from longer time controls. I will wait for more results from CCRL and CGET before I say anything else. That quickie comment was hilarious by the way.

I think that the difference between your time control and ccrl 40/40 may be bigger relative to the difference between ccrl 40/4 and ccrl 40/40

As far as I remember the program that earns the most from longer time control is the baron1.7.0 and it earned almost 100 elo when you compare ccrl 40/40 and ccrl 40/4

Uri

I never even checked Baron, and I started comparing after I saw some of the comparisons that you posted.

[chessfurby]

Ouch that hurt sooo much knowing it came from you. Ouuuuuuuch the pain ...

Let's just end it here and now. I have no interest in wasting more time with this. Maybe i misread or misunderstood your post. But clearly I wasn't the only one who was upset by your comments.

[eriq]

Your comments clearly lacked understanding and as far as others thinking like you, well there are too many people that love to jump to conclusions. Just to many fools and not enough schools in this world, I don't concern myself with any of them, that would be beneath me.

[eriq]

Great! which version would you be willing to release for free? I have not played with anything more current than ct14 that came with CA. and the latest palm version. I don't think anyone would care if it was an older version for the simply fact that not many masters would survive ct 14 in linux anyway. Or ct 15 gambit it's crazy, I love it.