Repeat performance issues with dedicated units?

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Just realized we can do this anyway. For training purposes I used to try to deduce the correct move from the remaining choices if I had picked an incorrect or equally viable move. Since the digital unit requires no training, we will simply input the correct move and proceed to the next array of choices. Come to think of it, this is probably how the analog unit was designed to be used anyway.

:P Breathing a sigh of relief regards,
John
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:Just realized we can do this anyway. For training purposes I used to try to deduce the correct move from the remaining choices if I had picked an incorrect or equally viable move. Since the digital unit requires no training, we will simply input the correct move and proceed to the next array of choices. Come to think of it, this is probably how the analog unit was designed to be used anyway.

:P Breathing a sigh of relief regards,
John
Hi John
if the dedicated chooses an incorrect move then as you say we just enter the correct move and proceed from there
you then record its incorrect score ...etcetc

however...suppose the Dedicated chooses a move ...not ..on the move list?
this happened with Rev II-Hiarcs on move 12 in the test game where it wanted to play 12..0-0-0 and it was not a move choice

the NEXT Best feature might work but not every computer has this feature
and even here it can come up with a move not on the list again

for this case I am thinking that I would have to enter every move choice that is on the list into the dedicated computer and see which one gets the higher score by the dedicated unit
the move with the highest score would be its choice

Thinking Out Loud Regards
Steve
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Steve B wrote:
Cyberchess wrote:Just realized we can do this anyway. For training purposes I used to try to deduce the correct move from the remaining choices if I had picked an incorrect or equally viable move. Since the digital unit requires no training, we will simply input the correct move and proceed to the next array of choices. Come to think of it, this is probably how the analog unit was designed to be used anyway.

:P Breathing a sigh of relief regards,
John
Hi John
if the dedicated chooses an incorrect move then as you say we just enter the correct move and proceed from there
you then record its incorrect score ...etcetc

however...suppose the Dedicated chooses a move ...not ..on the move list?
this happened with Rev II-Hiarcs on move 12 in the test game where it wanted to play 12..0-0-0 and it was not a move choice

the NEXT Best feature might work but not every computer has this feature
and even here it can come up with a move not on the list again

for this case I am thinking that I would have to enter every move choice that is on the list into the dedicated computer and see which one gets the higher score by the dedicated unit
the move with the highest score would be its choice

Thinking Out Loud Regards
Steve
Ah, good point. There were alarms blaring in my head but I attributed them to the wrong situation. The human player has a much easier task as analog chess is akin to a multiple choice exam. The digital unit, on the other hand, doesn’t know this, and will not automatically limit its search to the 6 candidate moves.

Having to enter each of the candidate moves in by hand and compare evaluations would be a tedious, time consuming chore, so an alternative might be to simply make no score changes, input the correct move and continue from there.

This reminds me of how in the second Kasparov vs. Deep Blue match, Garry was concerned about the possibility of certain moves being hand pruned by one of the IBM chess consultants. It’s too bad our digital dedicated units lack a provision for hand pruning.

Digital schach lament regards,
John
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:
Having to enter each of the candidate moves in by hand and compare evaluations would be a tedious, time consuming chore, so an alternative might be to simply make no score changes, input the correct move and continue from there.
OK
assuming we don't have too many instances where the dedicated computer chooses an unlisted move.. I think this would work



CyberInStorage Regards
Steve
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

mrcdoc wrote:Hi John,

Great!
Now that you mentioned it, I do remember too: there was a version for Apple that was based on a database issued from the CybercChess of Norbert Mikun. Iit makes sense that MIkun was developing several versions for each platform at this period... and Apple was one of the most popular platforms. I will add this on the ChessEval website. Thanks :)

I was looking to the different packages of games and found a confusing situation. Maybe everyone could give a list of his packages.
I am wondering what are the packages present in the software versions of CyberChess. Maybe this is a way to get the full package collection :)

Kind regards,

Maurice
Hey, The Dead Sea Scrolls weren’t excavated until 1952. Never give up hope that these will turn up someday.

The Holy Grail of analog chess regards,
John
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Steve B wrote:OK
assuming we don't have too many instances where the dedicated computer chooses an unlisted move.. I think this would work



CyberInStorage Regards
Steve
We may have to make an asterisk on these moves and later critically analyze them to determine if the digital unit found an equally viable alternative overlooked by Mr. Mikun, or if it simply failed to find the best move.

No easy way out regards,
John
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:
Steve B wrote:OK
assuming we don't have too many instances where the dedicated computer chooses an unlisted move.. I think this would work



CyberInStorage Regards
Steve
We may have to make an asterisk on these moves and later critically analyze them to determine if the digital unit found an equally viable alternative overlooked by Mr. Mikun, or if it simply failed to find the best move.

No easy way out regards,
John
Well lets see if we get a significant amount of "unlisted" moves
a couple of "unlisted's" I think would not invalidate the overall score of the game
if we get into 1/2 a dozen or so unlisted moves then we can go with a "critical analysis" approach although I would hate to use any subjective criteria to effect the score

Hammering Out The Match Terms Regards
Steve
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Steve B wrote: Well lets see if we get a significant amount of "unlisted" moves
a couple of "unlisted's" I think would not invalidate the overall score of the game
if we get into 1/2 a dozen or so unlisted moves then we can go with a "critical analysis" approach although I would hate to use any subjective criteria to effect the score

Hammering Out The Match Terms Regards
Steve
Greetings Steve:

This is precisely why I recommend the exclusive use of the Pro Modules for all games with the digital units. In stark contrast to the Amateur Modules, the Pro Modules never try to tempt the player with fish-bait moves that the Selective Search feature on the digitals would dismiss out of hand. Pretty much everything they offer looks reasonable, with only slight qualitative differences in most cases. Higher weighted are the brilliancies, hard to see tactical shots and moves based on long term planning, positional motifs, etc.

While I wouldn’t expect any of the late model digital units to fall out of the loop a half dozen times in a single game, positions will likely arise whereupon the digital unit will embark on an eccentric plan that commits one or more positional concessions because the digital calculates that it will win a pawn and erroneously evaluates the resulting position as something like +0.89, when the resulting position should in actuality evaluate to about -1.76 when additional factors are taken into consideration.

This is to be expected regards,
John
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:
This is precisely why I recommend the exclusive use of the Pro Modules for all games with the digital units.

Sounds Good to me
only major detail left to iron out is the actual dedicated computer that will be put to the "CyberChess Challenge"
no rush of course given that your Cyberator is squirreled away under lock and key

the choice is entirely your's .. although I think Ricardo's suggestion of a computer rated in the 2000 Elo range is a good one..unless you think a computer rated 2000-2100 Elo is too weak for a Pro Module game??
if not...
I am thinking of the Novag Citrine...
1)will retain the position in memory(a MUST ..so I don't have to re-enter the moves each day)
2)has a NEXT BEST feature..should we need to employ it for the "unlisted move "issue
(actually it is a RANDOM feature ..where the computer will choose from a list of moves almost equal in evaluation)
3)is a late model computer(released in 2006)
4)perhaps the most important reason..
haven't used the Citrine in a long time so its a good excuse to dust it off and fire it up

more here:
http://www.schach-computer.info/wiki/in ... ag_Citrine

Narrowing The Field Regards
Steve
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Steve B wrote: Sounds Good to me
only major detail left to iron out is the actual dedicated computer that will be put to the "CyberChess Challenge"
no rush of course given that your Cyberator is squirreled away under lock and key

the choice is entirely your's .. although I think Ricardo's suggestion of a computer rated in the 2000 Elo range is a good one..unless you think a computer rated 2000-2100 Elo is too weak for a Pro Module game??
if not...
I am thinking of the Novag Citrine...
1)will retain the position in memory(a MUST ..so I don't have to re-enter the moves each day)
2)has a NEXT BEST feature..should we need to employ it for the "unlisted move "issue
(actually it is a RANDOM feature ..where the computer will choose from a list of moves almost equal in evaluation)
3)is a late model computer(released in 2006)
4)perhaps the most important reason..
haven't used the Citrine in a long time so its a good excuse to dust it off and fire it up

more here:
http://www.schach-computer.info/wiki/in ... ag_Citrine

Narrowing The Field Regards
Steve
Steve,

The Cyberator accepts the challenge from the 2000 – 2100 ELO rated Novag Citrine unit, but be forewarned: if the Novag Citrine should prove unworthy of the challenge, it will henceforth be denounced, stripped of its batteries and unceremoniously entombed in the dark space formerly occupied by the Cyberator unit. You see, Steve, the analog Cyberator unit is waiting for the opportune moment to arise and take its rightful place as the king of the schachcomputer universe!

It’s good that you came clean and admitted beforehand that the [Next Best] feature is akin to a random move generator. The naming of this feature is a bit misleading, as I was under the impression that it actually forced the digital unit to play the move with the next highest evaluation.

There exists prestige and honor in garnering analog points from the Cyberator unit, and said points will neither be accrued nor forfeited based solely on chance. I hereby decree that all opening books and [Next Best] move generators be barred from these proceedings.

Layin’ down the law regards,
John
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:
The Cyberator accepts the challenge from the 2000 – 2100 ELO rated Novag Citrine unit, but be forewarned: if the Novag Citrine should prove unworthy of the challenge, it will henceforth be denounced, stripped of its batteries
OK
CyberChess Pro Vs. Novag Citrine it is

Citrine scoffs at the punishment of having its batteries stripped ..given it only works on AC power(no batteries)

Bring It Regards
Steve
User avatar
ricard60
Senior Member
Posts: 1285
Joined: Thu Aug 09, 2007 2:46 pm
Location: Puerto Ordaz

Post by ricard60 »

Great challange citrine is 2052 elo in the last schachcomputer.info active rating list now how it would be the terms of the challange?

1) number of games?
2) what time will be used for each game? cyberches is a manual dedicated chess computer maybe one minute for each move for citrine and then just wait until cyberchess responds? This is so each games are not so long and others can watch.
3) Where and when on the internet it will be played so at least i want to watch some games.

The arbiter regards :D
Ricardo
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

ricard60 wrote:Great challange citrine is 2052 elo in the last schachcomputer.info active rating list now how it would be the terms of the challange?

1) number of games?
2) what time will be used for each game? cyberches is a manual dedicated chess computer maybe one minute for each move for citrine and then just wait until cyberchess responds? This is so each games are not so long and others can watch.
3) Where and when on the internet it will be played so at least i want to watch some games.

The arbiter regards :D
Ricardo
Hi Ricardo
one game
Citrine will be thinking for 3 min avg. per move
Cyberchess is not really playing a game
it will rate the moves made by the Citrine
John can explain this better then me...

Above My PayGrade Regards
Steve
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Cyberchess wrote:
I hereby decree that all opening books and [Next Best] move generators be barred from these proceedings.
it just now occurred to me that there is no way for me to turn OFF the opening book of the Citrine..that is not an option
the book is not very big though ...only about 24k positions
nothing like the book in the Rev ll-Hiarcs which is in excess of 3 Million positions
my guess is that Citrine will come out of book very early on

Possible Match Breaker Regards
Steve
User avatar
Cyberchess
Full Member
Posts: 658
Joined: Wed Jan 08, 2014 6:10 pm

Post by Cyberchess »

Steve B wrote:
Cyberchess wrote:
I hereby decree that all opening books and [Next Best] move generators be barred from these proceedings.
it just now occurred to me that there is no way for me to turn OFF the opening book of the Citrine..that is not an option
the book is not very big though ...only about 24k positions
nothing like the book in the Rev ll-Hiarcs which is in excess of 3 Million positions
my guess is that Citrine will come out of book very early on

Possible Match Breaker Regards
Steve
I was wondering about this myself. On some of the units that I have owned, entering [Setup Mode], clearing the board and inputting a position by hand seemed to have the effect of having the computer commence calculating without checking the position against positions it had stored in the opening book. Would this work on the Citrine since there is apparently no way to disable the book on this model? Bear in mind that in the Cyber universe, the contest doesn’t begin until the first diagrammed position is reached.

Hoping the game’s still on regards,
John
Post Reply