Has the Sphinx Comet CXG-902 really an elo of 1800?

[paulwise3]

It seems I forgot to mention that there is a translation of the above mentioned document in Hein's database, see http://www.schaakcomputers.nl/hein_veld ... 04-1996%20[A-2465]%20CXG%20-%20Sphinx%20Comet%20(Dutch%20to%20English%20translation).pdf

(Because of all the strange characters in this url it does not seem to work. But copy and paste does work)

Also I forgot to mention that I finally got the "real" 1800 elo Comet, which is the Krypton Comet, a pocket chess computer, clone of the Krypton Challenge.

So I guess we can now close this thread

Aging and forgetting regards,
Paul

[Fernando]

Recently I am playing some matches (30 secs/move) at home, one of them is Novag Carnelian II - CXG Sphinx Comet. The results confirm the difference in strength between the two, assuming it is indeed the same program as the Enterprise S. So far the Comet scored a lot of losses and a few draws by move repetition. As I am playing a match of 20 games, both machines show their strengths and weaknesses in full .
One thing in particular struck me: it seems that the Comet does ponder, as it gives a lot of immediate answers during the game. That is something that is nowhere documented! Very interesting.

Pondering regards,
Paul

Perhaps it could play better WITHOUT the ponder function. No joke
I even once wrote an article about how pondering more harm than benefit weak programs. There is a math reason for that, but I cannot explain it here; for that you should have a master in tensorial calculus, as me...

Lying regards
Fern

[paulwise3]

Recently I am playing some matches (30 secs/move) at home, one of them is Novag Carnelian II - CXG Sphinx Comet. The results confirm the difference in strength between the two, assuming it is indeed the same program as the Enterprise S. So far the Comet scored a lot of losses and a few draws by move repetition. As I am playing a match of 20 games, both machines show their strengths and weaknesses in full .
One thing in particular struck me: it seems that the Comet does ponder, as it gives a lot of immediate answers during the game. That is something that is nowhere documented! Very interesting.

Pondering regards,
Paul

Perhaps it could play better WITHOUT the ponder function. No joke
I even once wrote an article about how pondering more harm than benefit weak programs. There is a math reason for that, but I cannot explain it here; for that you should have a master in tensorial calculus, as me...

Lying regards
Fern

Well you could show us your article...?
Anyway, too much pondering is not good for anyone,

Could get depressive from pondering regards
Paul

[Fernando]

Recently I am playing some matches (30 secs/move) at home, one of them is Novag Carnelian II - CXG Sphinx Comet. The results confirm the difference in strength between the two, assuming it is indeed the same program as the Enterprise S. So far the Comet scored a lot of losses and a few draws by move repetition. As I am playing a match of 20 games, both machines show their strengths and weaknesses in full .
One thing in particular struck me: it seems that the Comet does ponder, as it gives a lot of immediate answers during the game. That is something that is nowhere documented! Very interesting.

Pondering regards,
Paul

Perhaps it could play better WITHOUT the ponder function. No joke
I even once wrote an article about how pondering more harm than benefit weak programs. There is a math reason for that, but I cannot explain it here; for that you should have a master in tensorial calculus, as me...

Lying regards
Fern

Well you could show us your article...?
Anyway, too much pondering is not good for anyone,

Could get depressive from pondering regards
Paul

Thar article orbited around VERY SIMPLE COMMON SENSE IDEA: if from the beginning you know your way of examining something i far from perfect, each extra iteration cann but add more imperfection with just a margin of maybe-better moves in the middle of a forest of bad ones.