Millenium new release of chess computer

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
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Yarc wrote: What I find interesting, as mentioned in my earlier post, is that the Cosmos hardware seems to be much slower than that of the MCG. I wonder how well the Cosmos would play with MCG's hardware, unless I'm missing something about the Cosmos's hardware. Despite the advances in Computer chess, the MCG is still only drawing (near enough!) with the Cosmos. I think this shows just how good the earlier machines could be. I wonder what version two of the MCG will be like if it comes to fruition?
Your findings are very interesting, as indeed the hardware of the GK2100/Cosmos is slower. As a Morsch fan I am glad about it ;-).
But then there is also the possibility that the MCG has trouble with Cosmos's playing style. Against other machines the MCG may score much better, as it's current rating indicates.
Yarc wrote: I wonder what version two of the MCG will be like if it comes to fruition?
Yeah! We have serious indications that another strong Millennium will come up! I hope it will be before next Christmas :-P

Christmas present regards,
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

paulwise3 wrote: But then there is also the possibility that the MCG has trouble with Cosmos's playing style. Against other machines the MCG may score much better, as it's current rating indicates.
Paul
I had considered that, a real test would be a tournament against several chess engines at the same time control. I have wondered whether to do another match pitting the MCG against my Star Diamond, but I really think the SD would win. It's chess rating is probably about 130 ELO higher than the Cosmos, so it might not be a fair match.
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Yarc wrote:
paulwise3 wrote: But then there is also the possibility that the MCG has trouble with Cosmos's playing style. Against other machines the MCG may score much better, as it's current rating indicates.
Paul
I had considered that, a real test would be a tournament against several chess engines at the same time control. I have wondered whether to do another match pitting the MCG against my Star Diamond, but I really think the SD would win. It's chess rating is probably about 130 ELO higher than the Cosmos, so it might not be a fair match.
I am playing a match via internet with a german chessfriend, my MCG against his Star Diamond. Four games at 30 minutes/game resulted in a 3-1 win for the SD. We are now running a match at 30 secs/move, after two games 1-1.
So you might be surprised! ;)

Regards, Paul
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

paulwise3 wrote:
I am playing a match via internet with a german chessfriend, my MCG against his Star Diamond. Four games at 30 minutes/game resulted in a 3-1 win for the SD. We are now running a match at 30 secs/move, after two games 1-1.
So you might be surprised! ;)

Regards, Paul
I will be very interested in the outcome. However, based on the 3-1 win for the SD, if this win/loss ratio continued for a 10 or 100 game match, I believe it would make the ELO ratings correct for both machines. Taking nothing away from the MCG, your matches show that it's capable of taking games against the stronger machine even at a faster time control.

What I should do now is pit the Cosmos against the SD.

Thanks for the info, will you publish the games?
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

I will publish them ok.
But there are already a number of games published for the MCG on this site, and among them the 5 games it played in the worldfamous
2015 1ST CHESS FRIENDS OF THE WORLD tournament, for the final table see http://hiarcs.net/forums/viewtopic.php? ... &start=120, somewhere at the end of the page. All the games are also published in the same thread. You will see that it scored a respectable 3 points in a five round swiss tournament.

Proud MCG owner regards ;-)
Paul
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

paulwise3 wrote:I will publish them ok.
But there are already a number of games published for the MCG on this site, and among them the 5 games it played in the worldfamous
2015 1ST CHESS FRIENDS OF THE WORLD tournament, for the final table see http://hiarcs.net/forums/viewtopic.php? ... &start=120, somewhere at the end of the page. All the games are also published in the same thread. You will see that it scored a respectable 3 points in a five round swiss tournament.

Proud MCG owner regards ;-)
Paul
Paul, thanks for the info.

I remember reading from the computer chess magazine 'Selective Search' that some dedicated machines will only play at their best when set to tournament levels. The Star Diamond was one such machine identified as having this trait. I don't know if this was a factor with the machines used in the tournament as they were all set to 30 sec average level. I unfortunately don't own any of those nice machines to verify if this is true. Nonetheless, I see the MCG has a rating of 2226, which is much higher than the rating I achieved for the match against the Cosmos. Maybe I should try the 30S AV level. I also think I will play a match between the MCG and the SD. If the MCG rating is 2226 then it stands a good chance of winning against the SD.

Despite the curious performance against the Cosmos, I still like my MCG.
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

I think the novag (kittinger) machines are also good in blitzgames, like for instance the 30 min./game level. They seem to better identify the positions that need more time to investigate. I noticed that too with the SD, the Beluga and the Emerald.
2024 Special thread: viewtopic.php?f=3&t=12741
2024 Special results and standings: https://schaakcomputers.nl/paul_w/Tourn ... 25_06.html
If I am mistaken, it must be caused by a horizon effect...
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

paulwise3 wrote:I think the novag (kittinger) machines are also good in blitzgames, like for instance the 30 min./game level. They seem to better identify the positions that need more time to investigate. I noticed that too with the SD, the Beluga and the Emerald.
I read once, possibly in Selective Search, that the SD has a great deal of knowledge programmed into it. I don't know if it actually has more than other machines, but if so, could account for your observation that its able to find the lines that need more analysis at quicker time controls.

I think its amazing what the programmers achieved with these dedicated machines.
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

I have just started a match between my Star diamond and the MCG at the slow time control of 40 moves in 2 hours. So far they have played two games and the MCG easily won both games against the SD in around 40 moves or less! It announced mate in 7 in the second game! This seems to be in direct contrast with the games against the Cosmos, especially when you consider the SD has a rating of around 2180 and the Cosmos is around 2020.

It just goes to show that to get true ratings many games have to be played with many opponents.

I'll carry on with this match, 8 games to go. I'm wondering if the SD stands a chance.
Larry
Senior Member
Posts: 2269
Joined: Wed Aug 01, 2007 8:42 am
Location: Gosford, NSW Australia

Post by Larry »

Yarc wrote:I have just started a match between my Star diamond and the MCG at the slow time control of 40 moves in 2 hours. So far they have played two games and the MCG easily won both games against the SD in around 40 moves or less! It announced mate in 7 in the second game! This seems to be in direct contrast with the games against the Cosmos, especially when you consider the SD has a rating of around 2180 and the Cosmos is around 2020.

It just goes to show that to get true ratings many games have to be played with many opponents.

I'll carry on with this match, 8 games to go. I'm wondering if the SD stands a chance.
Look forward to the results. Don't expect MCG to trounce SD.
L
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Larry wrote: Look forward to the results. Don't expect MCG to trounce SD.
L
I wont, it's way too soon to make a judgement and I'm hoping the SD will come back, it's a nice machine. I'm also glad the MCG is probably not going to be trounced, but time will tell.

I will give the results of the match, and as usual it's going to take a little time.

I discovered a curiosity with the MCG's game clock. I had suspended a game (with the SD to move) by turning both machines off. Sometime later the MCG turned itself on announcing loss on time, presumably for the SD!? This never happened during the 10 game match with the Cosmos. Am I missing something here? If I turn the machines off, surely their clocks are paused or should be? The only difference from the match with the Cosmos is that I was powering the MCG with the mains adaptor and not batteries!? Surely, this should make no difference?
the mighty grob
Member
Posts: 4
Joined: Sun Feb 14, 2016 6:38 am

mcg "technique"

Post by the mighty grob »

Hi,
on a setting of game in 60 MCG quite easily beat Novag star diamond and mephisto atlanta I have only played one game against each and didn`t record moves. Seemed to play with "technique" winning without really trying. Rather like my Berlin (has won lots of games using very little of its allotted time!)has done on many occasions, however a game in 30 against the portable expert (cosmos) was drawn.
Just discovered and joined these forums and will record games in future.
I have the following:-
Mephisto London 68000,Atlanta, Master,Cougar(expert @ 50% speed up - could be interesting!? although Selective search gives them both the same rating) Expert, Talking chess academy, Nigel short.
Novag star diamond
A few good matches in there! God we dedicated fans have been starved. :D
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Post by Yarc »

Yarc wrote:
Larry wrote: Look forward to the results. Don't expect MCG to trounce SD.
L
I wont, it's way too soon to make a judgement and I'm hoping the SD will come back, it's a nice machine. I'm also glad the MCG is probably not going to be trounced, but time will tell.

I will give the results of the match, and as usual it's going to take a little time.

I discovered a curiosity with the MCG's game clock. I had suspended a game (with the SD to move) by turning both machines off. Sometime later the MCG turned itself on announcing loss on time, presumably for the SD!? This never happened during the 10 game match with the Cosmos. Am I missing something here? If I turn the machines off, surely their clocks are paused or should be? The only difference from the match with the Cosmos is that I was powering the MCG with the mains adaptor and not batteries!? Surely, this should make no difference?
Just completed game 3 of my 10 game match MCG v Star diamond. It's another win for the MCG in 65 moves. That's 3 wins in succession for the new machine. This game was closer but the MCG had the edge in the end game having better placed pawns. Come on SD, time to win some games!
User avatar
Yarc
Senior Member
Posts: 1051
Joined: Sat Apr 11, 2009 8:13 am
Location: United Kingdom

Re: mcg "technique"

Post by Yarc »

the mighty grob wrote:Hi,
on a setting of game in 60 MCG quite easily beat Novag star diamond and mephisto atlanta I have only played one game against each and didn`t record moves. Seemed to play with "technique" winning without really trying. Rather like my Berlin (has won lots of games using very little of its allotted time!)has done on many occasions, however a game in 30 against the portable expert (cosmos) was drawn.
Just discovered and joined these forums and will record games in future.
I have the following:-
Mephisto London 68000,Atlanta, Master,Cougar(expert @ 50% speed up - could be interesting!? although Selective search gives them both the same rating) Expert, Talking chess academy, Nigel short.
Novag star diamond
A few good matches in there! God we dedicated fans have been starved. :D
I don't know what it is about the Cosmos, but it seems to be very good against some stronger and much faster opposition. All these dedicated machines suffer far more from the horizon effect than say Stockfish on a super fast PC. Maybe it only takes a poor positional assessment and several moves later they can be in trouble irrespective of their speed. This is why several games (min 10) should be played before a slightly more accurate result can be realised.

I haven't pitted the Cosmos against the Star diamond yet, that could be interesting considering the 10 game match I played between the MCG and Cosmos was almost a draw. The MCG won the last game! In my current match with the star diamond at 40 moves in 2 hours, the MCG trounced the SD in the first two games and won the third in an end game. I think the difference in chess rating between these two machines (about 39 ELO going by the result of the 2015 1st Chess friends of the world match discussed earlier) should make this a much closer match. But the SD is not doing great at the moment.
User avatar
Dave C
Member
Posts: 460
Joined: Fri Jul 17, 2015 1:50 am
Location: SoCal, US

Re: mcg "technique"

Post by Dave C »

the mighty grob wrote:Hi,
Just discovered and joined these forums and will record games in future.
...we dedicated fans have been starved. :D
Welcome Grob - glad you have joined the forum. You have a very solid collection of strong chess computers. I'm looking forward to seeing the results of your matches.

Dave
Post Reply