Final, definitive Post about Nelson....

[Fernando]

LOL?
Why?
Which is the funny part of that?

Because it would have to be a Fidelity chess computer to test against Excalibur computers. Even for us to compare Alexandra with Spraklen you will have to test Fidelity computers to explore the theory that Nelson took Spraklen with him to Excalibur.

Alexandra came out in 2003. It is more of a recent computer.

Best regards

G


Looks like we have confusion. What I try to mean is that I have felt some Spracklen spirit in those two machines I mentioned. That's the reason I said long ago that in my opinion Nelson could have taken something from there. More exactly, I was thinking in Champion line, which includes Champion proper, Fidelity 9 and Super 9, Septennial...

OK I will include them in the tests and lets see what comes out of it.

Great. Maybe I am fully mistaken. But even if it is so in this case, i consider improbable that Nelson could NOT have used some good stuff he had in his hands. I presume the Spracklen code was Fidelity property and then, later, Excalibur property.

fern

[spacious_mind]

LOL?
Why?
Which is the funny part of that?

Because it would have to be a Fidelity chess computer to test against Excalibur computers. Even for us to compare Alexandra with Spraklen you will have to test Fidelity computers to explore the theory that Nelson took Spraklen with him to Excalibur.

Alexandra came out in 2003. It is more of a recent computer.

Best regards

G


Looks like we have confusion. What I try to mean is that I have felt some Spracklen spirit in those two machines I mentioned. That's the reason I said long ago that in my opinion Nelson could have taken something from there. More exactly, I was thinking in Champion line, which includes Champion proper, Fidelity 9 and Super 9, Septennial...

OK I will include them in the tests and lets see what comes out of it.

Great. Maybe I am fully mistaken. But even if it is so in this case, i consider improbable that Nelson could NOT have used some good stuff he had in his hands. I presume the Spracklen code was Fidelity property and then, later, Excalibur property.

fern

Hi Fernando,

That is where it becomes dangerous. Because Fidelity property belongs to Saitek and not Excalibur. He couldn't just take it. It is like stealing.

Best regards

[Fernando]

LOL?
Why?
Which is the funny part of that?

Because it would have to be a Fidelity chess computer to test against Excalibur computers. Even for us to compare Alexandra with Spraklen you will have to test Fidelity computers to explore the theory that Nelson took Spraklen with him to Excalibur.

Alexandra came out in 2003. It is more of a recent computer.

Best regards

G


Looks like we have confusion. What I try to mean is that I have felt some Spracklen spirit in those two machines I mentioned. That's the reason I said long ago that in my opinion Nelson could have taken something from there. More exactly, I was thinking in Champion line, which includes Champion proper, Fidelity 9 and Super 9, Septennial...

OK I will include them in the tests and lets see what comes out of it.

Great. Maybe I am fully mistaken. But even if it is so in this case, i consider improbable that Nelson could NOT have used some good stuff he had in his hands. I presume the Spracklen code was Fidelity property and then, later, Excalibur property.

fern

Hi Fernando,

That is where it becomes dangerous. Because Fidelity property belongs to Saitek and not Excalibur. He couldn't just take it. It is like stealing.

Best regards

I did not know that, BUT equally possible is Nelson was using some extra ideas.. I do not imagine Nelson just copying -paste the Sprackeln code. You just take a line of code or two, the kind of things that clone makers does or even less, and you have a better program.

[spacious_mind]

LOL?
Why?
Which is the funny part of that?

Because it would have to be a Fidelity chess computer to test against Excalibur computers. Even for us to compare Alexandra with Spraklen you will have to test Fidelity computers to explore the theory that Nelson took Spraklen with him to Excalibur.

Alexandra came out in 2003. It is more of a recent computer.

Best regards

G


Looks like we have confusion. What I try to mean is that I have felt some Spracklen spirit in those two machines I mentioned. That's the reason I said long ago that in my opinion Nelson could have taken something from there. More exactly, I was thinking in Champion line, which includes Champion proper, Fidelity 9 and Super 9, Septennial...

OK I will include them in the tests and lets see what comes out of it.

Great. Maybe I am fully mistaken. But even if it is so in this case, i consider improbable that Nelson could NOT have used some good stuff he had in his hands. I presume the Spracklen code was Fidelity property and then, later, Excalibur property.

fern

Hi Fernando,

That is where it becomes dangerous. Because Fidelity property belongs to Saitek and not Excalibur. He couldn't just take it. It is like stealing.

Best regards

I did not know that, BUT equally possible is Nelson was using some extra ideas.. I do not imagine Nelson just copying -paste the Sprackeln code. You just take a line of code or two, the kind of things that clone makers does or even less, and you have a better program.

Yes but that is what you as a proponent of Spraklen being used is insinuating. You are insinuating that he took and used Spracklen. I don't know I am not a programmer but I doubt you would have all that programming in your head. The papers belong to Mephisto later Saitek they bought it. As far as I know he had to leave Mephisto because of lack of progress after they bought Fidelity.

I cannot use work documents when I leave a job and take them to another company. That is illegal.

I prefer my Horvath theory.

Best regards

[Steve B]

Yes but that is what you as a proponent of Spraklen being used is insinuating. You are insinuating that he took and used Spracklen.

I don't think so
we don't know what agreements Excalibur could have had in place with Saitek to license and use any of the old Fidelity code
it is simply not the case that the only way Nelson while programming for Excalibur could have used Spracklen code in his programs is by stealing it

even using your theory about Horvath Excalibur would have had to license the right to use the program in their computers

Totally Legal Regards
Steve

[spacious_mind]

Yes but that is what you as a proponent of Spraklen being used is insinuating. You are insinuating that he took and used Spracklen.

I don't think so
we don't know what agreements Excalibur could have had in place with Saitek to license and use any of the old Fidelity code
it is simply not the case that the only way Nelson while programming for Excalibur could have used Spracklen code in his programs is by stealing it

even using your theory about Horvath Excalibur would have had to license the right to use the program in their computers

Totally Legal Regards
Steve

That's different to stating:

in my opinion Nelson could have taken something from there

And correct about Horvath, but we know that they had a License already with Mirage.

I just don't agree with "took" you know that Mephisto/Saitek would have watched Excalibur like a Hawk to the extend that although Novag, CXG & Krypton were in the range absolutely nothing remotely Mephisto looking was in the range. There was a lot of bad blood there with Mephisto.

Best regards

[klute]

Hi Cameron,

I am just curious, are you indicating that Grandmaster will post a lot of different moves because of Spracklen's Attack Tables to Igor and Mirage? I have played tests with Igor at level 22 and Mirage at level 23 (to account for the 2 MHz speed difference) and Mirage and Igor post 100% identical moves.

I have a Grandmaster and Grandmaster Platinum and not tried any tests yet (their bulk takes up too much table space and therefore are a pain ). If Grandmaster were the same then the same would apply to Mirage and Igor as well and quite likely to Ivan as well? Correct?

Best regards

Hi Nick

I was partly hesitant to post about the Excalibur GM program employing Spracklen Attack Tables, simply because of all the passion the Excalibur / Ron Nelson threads appear to create. Some people will say it's entirely plausible and others will argue it couldn't possibly be.

However, I decided it's important for all of us to impart what we know or what we've been told from primary sources. Ron Nelson himself claimed this quite some years back now. Steve B or Monsieur Plastique might also perhaps recall this. I feel it's quite a significant piece of information to ponder (whatever people decide to make of it) for at least two reasons:

1) While claiming primary authorship of the GM program, Nelson also credited use of techniques from other programmers.

2) Nelson mentioned a very specific contribution (Attack Tables) from named programmers (the Spracklens.)

Now on the practical side, I've played almost 2,500 games between about 25 classic models, all on tournament level (40 moves in 2 hours) because I'm a traditionalist. It took a huge amount of time, but I personally dislike faster time controls for testing chess computers.

About 300 of those games were played by Excalibur GM and about 220 of those games were played by Fidelity Elegance, ideal comparison machines for this topic.

Interestingly, after all this work I derived a rating of 1772 for GM and 1770 for Elegance, using the traditionally accepted rating of 1904 for the Mephisto MM IV as a benchmark machine. The GM and Elegance of course have different playing styles, but I do feel a GM without Spracklen Attack Tables would be about 100 ELO weaker, much like the typical mid range Excalibur programs - food for thought.

So I have a large database of personal work and findings, but it wouldn't really be comparable to other kinds of tests, both because of my use of much longer time controls and because of pondering.

Cameron

[klute]

Well, you proved what I say respect Spracklen stuff on later materiel...

Hi Fernando

Well, I think I agree with you more than not in your efforts to try and summarise a quite complex subject.

Cameron

[klute]

Legend has it that it was Nelsons actual voice that was used
Metallic Regards
Steve

Well Steve

That would align well with my theory that Ron Nelson has been half man / half chess computer since the late 1970s, fueled by a radioactive 4 MHz Z80 CPU.

I might elaborate on the reasons for my theory if there is sufficient interest.

Truth is stranger than fiction Regards

[Steve B]

Ron Nelson himself claimed this quite some years back now. Steve B or Monsieur Plastique might also perhaps recall this

i certainly remember this because he said he was the programmer of the GM.. to me personally..as i have mentioned here many times now...
dont remember him discussing the attack tables.. at least not in my conversations with him..

Primary First Person Account Regards
Steve

[klute]

i certinly remember this becsuse he calimed that to me perosnaslly
as i have mentioned here many times now
Steve

Hi Steve

Ok, if you're referring to the Spracklen Attack Tables aspect specifically, apologies if that's already been mentioned in other forum threads.

On the wrong side of 50 now but still sharp as a tack most of the time Regards

[Steve B]

i certinly remember this becsuse he calimed that to me perosnaslly
as i have mentioned here many times now
Steve

Hi Steve

Ok, if you're referring to the Spracklen Attack Tables aspect specifically, apologies if that's already been mentioned in other forum threads.

On the wrong side of 50 now but still sharp as a tack most of the time Regards

not sure what you mean Cameron
all i meant to say was that Nelson personally told me he was the programmer for the GM
i dont recall him telling me about the attack tables
not saying he didnt mention that to you or others.. just not to me
unless i forgot about it.. but i dont think i did

anyway...dont be surprised but as you will soon see some of the folks here put no weight what-so-ever on any of the claims Nelson made
if you read some of the many threads here regarding Nelson(and you literally have to search back at least a year or so) you will see he has been called an exaggerator, forgetful and even worse
i will add.. these folks are well intentioned.. knowledgeable collectors.. but they simply put no weight on his statements
i of course believe every word he told me especially in light of the evidence which now seems to be growing and growing from multiple independent sources

Just Preparing You Regards
Steve

[klute]

i dont recall him telling me about the attack tables
not saying he didnt mention that to to you or others just not to me
unless i forget about that but i dont think i did
Steve

Hi Steve

Well, the Spracklen Attack Tables aspect for Excalibur GM is something that always stuck in my mind, simply because it was so specific. Perhaps I wrote to Nelson myself at some point to clarify something.

There's nothing untoward in a primary programmer employing contributions or ideas from others, as long as they're acknowledged and legally made. In this regard pretty much all the classic chess programmers and companies were professional, with very few exceptions.

A similar in principle Fidelity example from the early days is Ed English claiming (I use the word "claiming" simply as a precaution here - this one is pretty cut and dried in my opinion) to have doubled the alpha-beta algorithm search speed between Chess Challengers 10 and 7. Yet no-one would ever dispute that Ron Nelson is "the programmer" of Chess Challenger 7.

And yes, I've read through most of the many Excalibur / Ron Nelson threads and I know they're thick and heated more often than not.

Where's my tin hat Regards

[spacious_mind]

Hi Cameron,

I am just curious, are you indicating that Grandmaster will post a lot of different moves because of Spracklen's Attack Tables to Igor and Mirage? I have played tests with Igor at level 22 and Mirage at level 23 (to account for the 2 MHz speed difference) and Mirage and Igor post 100% identical moves.

I have a Grandmaster and Grandmaster Platinum and not tried any tests yet (their bulk takes up too much table space and therefore are a pain ). If Grandmaster were the same then the same would apply to Mirage and Igor as well and quite likely to Ivan as well? Correct?

Best regards

Hi Nick

I was partly hesitant to post about the Excalibur GM program employing Spracklen Attack Tables, simply because of all the passion the Excalibur / Ron Nelson threads appear to create. Some people will say it's entirely plausible and others will argue it couldn't possibly be.

However, I decided it's important for all of us to impart what we know or what we've been told from primary sources. Ron Nelson himself claimed this quite some years back now. Steve B or Monsieur Plastique might also perhaps recall this. I feel it's quite a significant piece of information to ponder (whatever people decide to make of it) for at least two reasons:

1) While claiming primary authorship of the GM program, Nelson also credited use of techniques from other programmers.

2) Nelson mentioned a very specific contribution (Attack Tables) from named programmers (the Spracklens.)

Now on the practical side, I've played almost 2,500 games between about 25 classic models, all on tournament level (40 moves in 2 hours) because I'm a traditionalist. It took a huge amount of time, but I personally dislike faster time controls for testing chess computers.

About 300 of those games were played by Excalibur GM and about 220 of those games were played by Fidelity Elegance, ideal comparison machines for this topic.

Interestingly, after all this work I derived a rating of 1772 for GM and 1770 for Elegance, using the traditionally accepted rating of 1904 for the Mephisto MM IV as a benchmark machine. The GM and Elegance of course have different playing styles, but I do feel a GM without Spracklen Attack Tables would be about 100 ELO weaker, much like the typical mid range Excalibur programs - food for thought.

So I have a large database of personal work and findings, but it wouldn't really be comparable to other kinds of tests, both because of my use of much longer time controls and because of pondering.

Cameron

Hi Cameron,

Thanks for the clarification. As time allows I will continue testing all the computers through all the tests and of course Spracklen as well. I am not hung up on any particular programmer. I just want to get to the bottom of this someday once and for all

I believe there might be a possibility to narrow it down to a few suspects by playing through several test games and identifying specific moves that are only repeatedly unique to the same few computers.

But to do that takes time to play every computer and every author and have enough test games to rate and compare and identify unique abnormalities.

So it is going to be a long and patient process, but worth it since I will also have ratings and comparisons for every computer (except for the ones that Steve and others have and I don't have But hey they of course could also repeat the same tests too)

Patience is a virtue.

Regards

[spacious_mind]

Hi Cameron,

Sorry I should have asked you earlier but while I am sitting at computer testing Super 9 Deluxe, I am pondering together with my computer and my curiosity to ask you some questions got the better of me. (Other people please feel free to participate, lets call it a Logic test!)

I am just curious how you would explain these abnormalities with Nelson claiming authorship. (I am trying to get my mind working through some logic and tie them to actual facts as they exist)

1) Lets assume that at least the timelines are correct and that Mirage came from Krypton and by nature of absolute fact, Igor plays the same moves of Mirage and therefore these two programs have to come from the same chess program author. (taking Ivan out of the discussion albeit it also comes from the same source). If we identify that Grandmaster is generally the same but has a few moves that differ to Igor (just like Ivan/Igor differ), then in your opinion this move difference would account for the inclusion of Spracklen Attack Tables in the Grandmaster as stated by Nelson himself correct?

2) Now lets assume that this is in fact the case. This would explain a differentiation between Grandmaster and Igor (assuming it exists) correct?

But the original author whoever he might be, would still be who? It would mean that Nelson modified per his statement Grandmaster by including Spracklen Attack Tables?

I am asking because you are being very specific with the Grandmaster statement and you are not including the other discussed computers into your post. I have not personally tested Grandmaster yet, therefore I don't know if deviations exist but your post indicates a GM difference to the others correct. Yes?

3) If Nelson as stated by Nelson did in fact include Spracklen Attack Tables into Grandmaster, would it not also stand to reason that given that the programming for the levels, the evaluation (albeit reversed from black to white perspective) Manual etc., would also not be his original authorship as these can be traced directly back to Krypton and even before that to CXG. The author of all these programming functions could not possibly be Nelson himself, because Legend stems from 1992 and he was at that time directly employed by Mephisto. Would you agree with this reasonable conclusion?

4) So now we have in summary the following that could not be categorized as Nelson:

- Spracklen Attack Tables (Spracklen) for GM
- Programmed Levels (CXG/Krypton)
- Evaluation function (CXG/Krypton - albeit reversed)
- Teach Modes (Krypton)
- User Manual (Krypton)
- Ivan Housing (Krypton)

Would you agree this deductive conclusion?

5) Based specifically for GM, Igor, Mirage & Ivan would it not also be very reasonable to say that the original peripheral programing authorship of all of the things listed in point 4 could not possibly be Nelson's? Particularly, since we know as a fact that Mirage was licensed to Excalibur by Krypton and therefore all of the above was fully accessible for study and future modifications and improvements? Would this be a reasonable and logical conclusion?

Which circles back to chess program author? Who was that? Can we say it was Nelson for the below specific 4 computers in order of existence?:

1) Mirage
2) Ivan
3) Igor
4) Grandmaster


6) Might it be possible that Nelson's words are being incorrectly quoted or misinterpreted by mistaking programming or end product author (there is no doubt that he single handedly put most of the pieces together at Excalibur, especially in later years. At the beginning NO since Excalibur started as a Distributor) with the word chess engine author? In his statement did he mean "I am the author of the chess engine"? or did he mean "I am the Author/Creator of the finished product"? As in the case of GM?

Our pursuit is the author of the chess engine itself and your additional insight just adds more weight that Nelsons skill was being the Puppet Master, his skill was in putting all the pieces together and that means he was not necessarily the writer or the builder of every single piece that makes the computer a finished product.

In a sense yes absolutely you can call him the author of the finished product. But the engine itself remains a topic of doubt and you added to the doubt with your additional insight Our disagreement and heated discussions are around Chess Engine Authorship and that is still I think a question mark.

The statement "Modified by Nelson", now that could be a reasonable statement in the case of GM (but only if it actually shows differences to Mirage).

Hoping for a calm opening minded discussion and participation.

Perhaps we should be giving credits to creators as well as chess engine author, but unfortunately for most every other computer and manufacturer we don't really know who the "creator/visionary" was.

With Nelson we seem to be continuously mixing one with the other.

Best regards