Playing Ron Nelson Programs, Today

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
klute

Re: Playing Ron Nelson Programs, Today

Post by klute » Mon Apr 09, 2007 10:35 pm

Hi Fernando

I'm the one who's been doing the extensive 40/2 testing including the 166 GM games mentioned so far.

I agree with you in the sense that I think both Morsch and Kittinger can do more with an H8 and 32K than Nelson, but nevertheless the GM is a pretty decent effort.

Cameron

klute

Re: Playing Ron Nelson Programs, Today

Post by klute » Mon Apr 09, 2007 10:48 pm

Hi Steve

Essentially that's what I do, but no machine is ever forced to play a line it could never play of its own volition. In GM's case, that means no 1. c4, 1. Nf3 or 1. f4 for instance. There are no duplicate games in my database.

It's important to note that while GM only plays 1. e4 or 1. d4 on its own, it does have quite a wide repertoire from move 3 onwards in 1. e4 e5 lines (after playing 2. Nf3 here 100% of the time) and also in other 1. e4 and 1. d4 openings - this variety is sufficient for playing fairly extensive matches with no repeat games.

Cameron

klute

Re: Playing Ron Nelson Programs, Today

Post by klute » Mon Apr 09, 2007 11:11 pm

Hi Steve

See my post further down, but for the record and others reading, the GM of its own volition plays only 1. e4 and 1. d4 with the former favoured by a ratio of about 9 to 1.

In reply to 1. e4, the GM of its own volition plays 1...e5, 1...c5 and 1...c6 in fairly balanced ratios.

After 1. e4 e5, the GM as White of its own volition invariably plays 2. Nf3 but after that point plays many lines of its own volition. The GM will also play a wide range of lines of its own volition against Sicilians etc.

As Black after 1. d4, GM plays only 1...d5 or 1...Nf6 of its own volition but from that point will play many lines of its own volition.

Another way to look at it - the GM has little variety as either colour in the first few moves but from that point on almost anything goes.

It's not as limited in openings variety as say a Par Excellence, Savant II / Royale or the original Conchess.

Cameron

Steve B

Re: Playing Ron Nelson Programs, Today

Post by Steve B » Mon Apr 09, 2007 11:35 pm

klute wrote:Hi Steve

Essentially that's what I do, but no machine is ever forced to play a line it could never play of its own volition. In GM's case, that means no 1. c4, 1. Nf3 or 1. f4 for instance. There are no duplicate games in my database.

It's important to note that while GM only plays 1. e4 or 1. d4 on its own, it does have quite a wide repertoire from move 3 onwards in 1. e4 e5 lines (after playing 2. Nf3 here 100% of the time) and also in other 1. e4 and 1. d4 openings - this variety is sufficient for playing fairly extensive matches with no repeat games.

Cameron
hi Cameron
thats what i imagine you had to be doing
again i agree 100% that some intervention is necessary or you would not have even a dozen different games

thanks for clarifying things for me

Regards From The Blue Mountains
Steve

klute

Re: Playing Ron Nelson Programs, Today

Post by klute » Mon Apr 09, 2007 11:53 pm

Hi Steve

After doing so much 40/2 testing with the GM and many other classic and modern machines, I have no doubt that my high 1700s rating for GM is an accurate assessment, using the Mephisto MM IV's traditional and proven 1904 rating as a reference point. For those who subscribe to the SSDF rating deflation move of 1996 (I don't) then the GM would only be in the high 1600s.

It's very clear to me that the GM doesn't quite have what it takes to crack 1800. It's virtually line-ball with the best of the 1984/85 machines like the Elegance, Super Connie and Turbostar - these can't quite crack the 1800 mark either if people are totally objective.

Cameron

Steve B

Re: Playing Ron Nelson Programs, Today

Post by Steve B » Tue Apr 10, 2007 12:18 am

klute wrote:Hi Steve

After doing so much 40/2 testing with the GM and many other classic and modern machines, I have no doubt that my high 1700s rating for GM is an accurate assessment, using the Mephisto MM IV's traditional and proven 1904 rating as a reference point. For those who subscribe to the SSDF rating deflation move of 1996 (I don't) then the GM would only be in the high 1600s.

It's very clear to me that the GM doesn't quite have what it takes to crack 1800. It's virtually line-ball with the best of the 1984/85 machines like the Elegance, Super Connie and Turbostar - these can't quite crack the 1800 mark either if people are totally objective.

Cameron

Well as i have written in this very thread.. Hallsworth Of Selective Search also shows a rating of 1794 for the GM so your rating system is SPOT ON as far as i am concerned

i dont remember your current collection exactly ..you and Jon should think about getting one of the Novag Opal Series computers
they also rate in the mid-1700 Range.. are nice and small pressure sensory units and for Jon they are made in plastic
:wink:

i have all four in the series but there is not all that much difference between them as far as i can tell..just got the Star Opal and i will play some games against the EAG V6 which is about 250 points stronger
i love seeing these plastic units getting CRUSHED by Wood

The Fidelity Decorator Sends His Regards
Steve

User avatar
fern
Posts: 8755
Joined: Sun Feb 26, 2006 3:07 pm

Re: Playing Ron Nelson Programs, Today

Post by fern » Tue Apr 10, 2007 12:49 am

Surely is. I have assigned it the mission -very difficult - to teach me how to play against a powerful patzer. Besides is a pretty objet to see on your desk.

Grandmaster regards
Fernando

klute

Re: Playing Ron Nelson Programs, Today

Post by klute » Tue Apr 10, 2007 1:02 am

Hi Steve

I am pleased the Decorator continues to provide service and pleasure - it is surely the finest breakfast tray in all the land.

I bought a Star Opal to discover why it has a claimed extra 20 ELO points - that's marketing for you. Amazingly enough, it's drawn a 10 game match with Express 16K and even has a slight plus score against Advanced Star Chess so far.

These results are of course garbage - the Star Opal has simply been very lucky so far and will come unstuck. Advanced Star Chess is very serious now and will not tolerate being usurped by the interloper.

Now my good man, I have a mission for you. You are to deliver unto me a boxed, minty and working Destiny Prodigy ("S" version naturally) - my Counter Gambit is getting lonely.

Otherwise, a Holz ARB with 4.0 module and factory carry bag will have to suffice I suppose - a 3.0 module if you really must.

Peg Sensory Regards

Cameron

Steve B

Re: Playing Ron Nelson Programs, Today

Post by Steve B » Tue Apr 10, 2007 1:28 am

klute wrote: Now my good man, I have a mission for you. You are to deliver unto me a boxed, minty and working Destiny Prodigy ("S" version naturally) - my Counter Gambit is getting lonely.
oddly enough... even as we speak ..i have in my Doubles collection.. a Destiny Prodigy(Not Mini-Master)..it aint mint though.. has all of the pieces and dust cover..no box and you will have to make due with a copy of the manual(English)
the keyboard move input works fine but the peg sensory move input is a bit iffy and not all of the squares recognize the moves any longer
so to play a complete game you will have to go with all keyboard move input or you can use the peg input and if you hit a dead square you can switch to keyboard
let me know if your interested and i will give it a run through

Elite Champion (blank Plaque) sends his regards
Steve

JonP01

Re: Playing Ron Nelson Programs, Today

Post by JonP01 » Tue Apr 10, 2007 2:14 am

Steve B wrote:you and Jon should think about getting one of the Novag Opal Series computers
they also rate in the mid-1700 Range.. are nice and small pressure sensory units and for Jon they are made in plastic

Hi Steve,

I am afraid from my point of view that not even glorious, molded plastic can save the Opal / Carnelian program. As I say, it can't even win against the Excalibur Chess station, which isn't even capable of thinking for more than about 2 minues 25 seconds per move and is Excalibur's "cheapest" current chess program. I know I am probably the most scathing of the Novag 16K program amognst the three of us, but I grew tired of the 1, 2 and 3 ply oversights such as blundering pieces, walking straight into forks, etc (although I do agree with Cameron that these programs could conceivably go a bit better against humans of comparable strength than they might machines of comparable strength). In actual fact, the Novag 16K programs play rather like a handicapped PC engine. But for me to say that is still pretty insulting to the Novag program. I have theorised that perhaps these programs could benefit to some degree from having a best move / random function. I have noticed that they don't always repeat the same blunders on takebacks, meaning the machine must be picking amongst several moves. That and perhaps the extensive openings library is taking away too much from the rest of the program. Some authors are not as good as others in working with low memory. There is no comparison, for instance, between an original 2 Mhz Constellation and the 16K Carnelian II / Star Opal. I have no doubt which machine is superior (and by a big margin). I would much rather have seen the latest 16K programs with a smaller library and more room for the engine code.

What amazes me is the shear galaxy of strength difference between these Novag 16K programs and the current crop of Novag 32K programs. I think I mentioned before, the only way to get an even game between the Novag 32K programs and the 16K programs is to give the 16K program 3 minutes per move to the 32K at 4 seconds per move with pondering turned off. And that is assuming the 16K program can abstain from it's chronic blunder-fests and not throw anything away. I think the Novag Jade II and Amber were miniature classics, augmented as much by their entertaining playing style as their finely sculptured plastic appearance. Nevertheless, given my experience with Opals, Agates and Carnelians,

Plastic is good but not always Regards

Jonathan

Post Reply