World Championship Winning Computer Chess Software Program & Downloads for Chess Databases, Analysis and Play on PC, Mac, iPad and iPhone — Visit: Hiarcs.com
bataais wrote:I still think Triomphe, Counter Gambit and 301XL could well be early Morsch, the mate-in-2 tests I made strongly support my theory.
Solving a mate-in-2 problem in 3 moves tells something about the search algorithm, Depth First or some form of selective deepening. Mephisto Europa (6301, 16K, Morsch) and 301XL (6301, 4K) play the exact same moves which lead to a mate-in-3. Isn't this interesting?
One can't expect too much of 4K ROM and 128Bytes of RAM, even if it's a Morsch. Triomphe is running on 4Mhz (1Mhz, clock divison by 4), so it's clearly weaker than Counter Gambit or Yeno 301XL, both running at double speed.
Your theory suggests that 301XL and Counter Gambit might be from a same author. There is no such theory that Morsch ever created a program that plays at around 1100-1200 as in Triomphe or 1200-1400 as in Counter Gambit. Therefore if you are associating Yeno 301XL with Counter Gambit then you are categorizing Yeno 301XL as a very weak program. These programs are too weak for Morsch regardless of an early Morsch that you see in other CXG's. Its an apples to oranges comparison. You have Frans Morsch in a CXG 40 plus as around the weakest Morsch program and Europa from 1987. That is as weak as Morsch gets on same hardware which is a difference of 400 ELO.
bataais wrote:We will never know, unless David Levy or the original programmer speaks out.
What about my test method?
Your rating test is inconclusive as there might be plenty of programs that come to a same conclusion.
Try a Colditz Test or do my Spacious Mind test that will actually give you a score and exact move by move difference played over 5 games that you can compare move by move. Also include Europa so you can exactly compare to a weakest known Morsch.
(Besides, next to Triomphe there is M1, Morsch 1, Model 1, Mark 1 ? )
So if it were a Morsch, then it would be his second program, just after completing Mephisto Mondial, which was released in September 1985.
Best,
Michael
LOL that could be my last name.... totally meaningless.
You do realize that the play quality is day and night between a Triomphe and Mondial?
It is humanly impossible for Morsch to create a program as weak as what you are suggesting. Just like Ed Schroeder he was of a new generation that had already surpassed what you see in a Triomphe by the time he was engaged as manufacturer's programmer. None of these manufacturers needed a new programmer to come along and give them yet another weak program. They all already had plenty of those to choose from that would have cost them nothing.
ps I did test the Triomphe a while back and its performance sits nicely between two Danielsen computers just where it belongs from a strength perspective propping up the table with the rest of the bottom feeders
bataais wrote:We will never know, unless David Levy or the original programmer speaks out.
Perhaps Bryan Whitby can track Mark Taylor down?
hes usually good at tracking these guys down for an interview
here is a post of his from the CCC 20 years ago... http://www.stmintz.com/ccc/index.php?id=11870
Well the Timeline shows a ? next to the Triomphe which means that the date is uncertain. The only concrete bit of evidence I have is a guarantee card which shows my Triomphe was sold by a French retailer on 25th April 1986. Which is well before Hein's date of February 1987.
As is often the case this chess computer arrived in different countries at different times making a date harder to pin down. Although the Triomphe shares components with the Master it looks like there were at least two production runs.
The closest hardware specification to the Triomphe comes in the form of the CXG Enterprise S, CXG Star Chess and Mephisto Merlin 4K which all have the 6301V1 chip, 4K ROM and 128 bytes RAM. They are all currently attributed to Kaare Danielsen. The Enterprise S and Star Chess are late 1984/early 1985. The Counter Gambit has a 63B0V1 chip so is probably later?
Mike Watters wrote:
Triomphe now changed on the Timeline to 1985?
While i have you...
i notice your timeline shows a release date for the Scisys President Chess of Oct.1982(no question mark)
my manual for the President Chess shows a copyright date of 1983
i guess its possible they first released them for sale in 1982 and then again in 1983 and revised the manual ...can i ask your source for an Oct 1982 date?
Steve B wrote:
While i have you...
i notice your timeline shows a release date for the Scisys President Chess of Oct.1982(no question mark)
my manual for the President Chess shows a copyright date of 1983
i guess its possible they first released them for sale in 1982 and then again in 1983 and revised the manual ...can i ask your source for an Oct 1982 date?
Digging Down Deep Regards
Steve
Fair point. Think the Oct.1982 is an old entry based on a SciSys announcement. 1983 it is.
Steve B wrote:
While i have you...
i notice your timeline shows a release date for the Scisys President Chess of Oct.1982(no question mark)
my manual for the President Chess shows a copyright date of 1983
i guess its possible they first released them for sale in 1982 and then again in 1983 and revised the manual ...can i ask your source for an Oct 1982 date?
Digging Down Deep Regards
Steve
Fair point. Think the Oct.1982 is an old entry based on a SciSys announcement. 1983 it is.
OK Mike
true scholarly work your Timeline
i refer to it often
What is so intriguing with Danielsen is that he had those 4 or 5 play styles in his programs which all when tested perform close yet tantalizingly different. It would be a simple matter of just picking one of the styles turn the rest off and you would almost not have a clue whose or what program it is.
But I guess most people would not know that unless they have played and tested enough of them with all his different play styles.
It is because of examples as this that you end up having odd computers where the programmer remains a mystery. Strength wise fits into someone's programming category perfectly yet hard to pin down exactly.
This of course leads to discussions about other programmers being the author for some of them, and of course some suggestions are very exotic and far fetched but it keeps discussions lively.
Pity it wasn't a few years later because Ron Nelson could have been thrown into the mix. Especially since he was a master of redesign. Now that would have been fun to discuss. Or maybe he did do some secret work for CXG as well at around that time.
bataais wrote:We will never know, unless David Levy or the original programmer speaks out.
What about my test method?
Hi Michael,
I am game if you are. Below is a link for 5 spreadsheets. You set up your Yeno and I will set up my Counter Gambit and lets first play through these tests to see if Yeno 301 XL and Counter Gambit repeat the same moves.
If you are interested or anyone else then then download the spreadsheet and start with Game 1. Lets see how they score and this can be compared against Triomphe which I had already done a couple of years ago.
I do have Yeno 301XL, Enterprise S, Star Chess and Computachess III so its not a big deal for me to add one of these later but I prefer not to at the moment as I am more into completing the tournament that I am playing at the moment.
So volunteers are welcome.
Tests should all be 30 seconds per move ie:
Chess King Counter Gambit - Level 6
Yeno 301XL - Level 6
CXG Star Chess - Level 7
As quick examples.
If Yeno 301 XL and Counter Gambit are the same then more than 90% of their moves should be identical.