High Performer Chess Computer Tournament

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 »

Hi Helge,

Maybe you missed my reply-mail to Michael and you.
I can play Saturday early in the morning, or sunday evening at a time of your choice.

Best 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
blaubaer
Full Member
Posts: 935
Joined: Thu Jul 28, 2011 12:53 pm
Location: Bavaria, the centre of Mysticum
Contact:

Post by blaubaer »

Hi Paul,

I'll contact Helge!

65/66 Regards, Michael
User avatar
blaubaer
Full Member
Posts: 935
Joined: Thu Jul 28, 2011 12:53 pm
Location: Bavaria, the centre of Mysticum
Contact:

Post by blaubaer »

Hi,

the last game of this tournament

game 8-3 Millennium ChessGenPro/Paul - Montreux/Helge

will be celebrated next Saturday at 7h UTC+2h!

You're welcome to watch it on Lichess.org!

66/66 Regards, Michael
:lol:
Larry
Senior Member
Posts: 2269
Joined: Wed Aug 01, 2007 8:42 am
Location: Gosford, NSW Australia

Post by Larry »

blaubaer wrote:Hi,

the last game of this tournament

game 8-3 Millennium ChessGenPro/Paul - Montreux/Helge

will be celebrated next Saturday at 7h UTC+2h!

You're welcome to watch it on Lichess.org!

66/66 Regards, Michael
:lol:
Sorry but I'm puzzled at the time thing. Is the game played
in Berlin?....or at least in the same time zone as Berlin? If
not, then in the time zone of which major city will it be played?
Thanks in advance...Larry
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi Larry,

Yes, it is Berlin and Amsterdam time. I am puzzled also, leave it up to Michael to explain ;-). The winter/summertime changes make it even more complicated...

The game will be played on lichess.org. I am playing in Eindhoven, Helge from somewhere in Germany...

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
blaubaer
Full Member
Posts: 935
Joined: Thu Jul 28, 2011 12:53 pm
Location: Bavaria, the centre of Mysticum
Contact:

Post by blaubaer »

Hi Larry,
Larry wrote:
blaubaer wrote:Hi,

the last game of this tournament

game 8-3 Millennium ChessGenPro/Paul - Montreux/Helge

will be celebrated next Saturday at 7h UTC+2h!

You're welcome to watch it on Lichess.org!

66/66 Regards, Michael
:lol:
Sorry but I'm puzzled at the time thing. Is the game played
in Berlin?....or at least in the same time zone as Berlin? If
not, then in the time zone of which major city will it be played?
Thanks in advance...Larry
it's at 7h UTC+2 - ok, it's written rather small, but I stated it in the above post.

Australia is UTC+8, +9, +10, so it's 6, 7, 8 h later for you - good timing for a dessert after lunch, instead of a cigar...

Regards, Michael :D
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Post by paulwise3 »

Hi all,

The last game of this tournament has been played. Finally the MCG Pro managed to win a game! ;-) And even left the last position in the ranking...

In an Open Catalan the Montreux tried too long to keep it's gambit pawn, and came in a bad position, which the MCG Pro cleanly played to a win.

[Event "Hiarcs 2017-1"]
[Site "https://lichess.org/4Q0ssxuX"]
[Date "2017.07.01"]
[Round "8.3"]
[White "Millennium ChessGenius Pro"]
[Black "Mephisto Montreux"]
[Result "1-0"]
[ECO "E04"]
[Annotator "lichess.org"]
[PlyCount "106"]
[EventDate "2017.??.??"]

1. d4 d5 2. Nf3 e6 {* last bookmove} 3. g3 Nf6 4. c4 dxc4 5. Bg2 {
E04 Catalan Opening: Open Defense} Nc6 6. O-O {* last bookmove} Bd7 7. Nbd2 Na5
8. Qc2 Bb5 9. Rd1 Bb4 10. Ne5 c5 11. dxc5 Qc7 12. Ndxc4 Bxc5 13. Be3 Bb4 14.
Rac1 Nxc4 15. Nxc4 O-O 16. Qb3 Qe7 17. a3 Bxc4 18. Rxc4 Bd6 19. Qxb7 Qxb7 20.
Bxb7 Rab8 21. Bf3 Be7 22. Rc7 Rfd8 23. Rxd8+ Bxd8 24. Rxa7 g6 25. b4 Bb6 26.
Bxb6 Rxb6 27. Kf1 Kf8 28. e3 Nd5 29. Bxd5 exd5 30. Ke2 h6 31. Ra5 Ke7 32. Rxd5
Ke6 33. Ra5 Rb8 34. Ra7 h5 35. h4 Kf6 36. Kd2 Ke6 37. Kc3 Rc8+ 38. Kd3 Rd8+ 39.
Kc2 Rb8 40. Kb3 Rd8 41. b5 Rb8 42. a4 f6 43. Rg7 Kf5 44. f3 g5 45. e4+ Ke6 46.
hxg5 fxg5 47. Rxg5 Rf8 48. f4 h4 49. Rf5 Rh8 50. gxh4 Rxh4 51. a5 Rh8 52. b6
Kd7 53. a6 Kc6 {Black resigns.} 1-0

Here is the final position where the Montreux resigned.

[fen]7r/8/PPk5/5R2/4PP2/1K6/8/8 w - - 1 54[/fen]

Michael, thanx for organising this tournament!

Best 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
blaubaer
Full Member
Posts: 935
Joined: Thu Jul 28, 2011 12:53 pm
Location: Bavaria, the centre of Mysticum
Contact:

Post by blaubaer »

Hi,

mission accomplished - game over - we finished the tournament! :lol:

Finally MCGPro kicked London68000 to the last position!

And the winner is: Mysticum Aristarch4.50 - congratulations!

Congrats to Resurrection Ruffian as the second winner!

Risc25002MB and Reflection Vancouver are both on position 8!

Thanks for your outstanding discipline - 66 games within 13 weeks...

See you next year!

Image

Nail me to the cross, Steve, Regards, Michael :P
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

Thanks all! I've really enjoyed following along with the games.

I'm surprised how poorly the MCG Pro performed.

I'll play a few games of my MCG Pro against the MCG (similar to Lars).

First game played yesterday. MCG destroyed the Pro, which played very defensively. Results will be posted in Tourney forum.

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

Post by paulwise3 »

fourthirty wrote:Thanks all! I've really enjoyed following along with the games.

I'm surprised how poorly the MCG Pro performed.

I'll play a few games of my MCG Pro against the MCG (similar to Lars).

First game played yesterday. MCG destroyed the Pro, which played very defensively. Results will be posted in Tourney forum.

Greg
Hi Greg,

It could have been just a bad opening choice of the MCG Pro?
It should be the same program, but running at almost 3 times faster speed.

Best 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
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

paulwise3 wrote:
Hi Greg,

It could have been just a bad opening choice of the MCG Pro?
It should be the same program, but running at almost 3 times faster speed.

Best regards,
Paul
Hi Paul,

Good point. I did analyze the game using Deep HIARCS on my computer (30 seconds per move), and HIARCS had the MCG at a very slight +0.10 advantage over the Pro after both units were out of book.

In the tournament that Lars is currently running, the MCG Pro has a 2:1 win ratio over the MCG (8.5 - 4.5). I'm not sure if this is the expected result with the difference in clock speed and ELO between the two machines?

Besides the clock speed, the only other difference would be the opening books. The MCG books are well defined - the MCG book is still somewhat of a mystery to me.

If I recall - you played the MCG Pro with the HIARCS book in the tourney, and Steffen utilized the London book.

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

Post by paulwise3 »

Hi Greg,

The MCG book should be the same as book 1 of the MCG Pro.
Book 2 is the Hiarcs book (= the default chosen book).

With several computers you can select a certain part of the opening book, such as passive, active, and tournament. With the MCG's that is not possible. So you are not sure that a certain opening in the book fits the playing style of the program. As for the difference in speed, it is often assumed that doubling the speed gives an elo raise of 50 - 70 points. But also, to be able to calculate one ply deeper, one generally needs a speed accelaration with factor 6. But it just depends how close you are to a next ply at a given level. For instance, the Excalibur Grandmaster tuned at double speed has an amazing raise of 127 points at active level (30 secs/move) on the wiki-elo-liste, whereas at the tournament level the raise is only a few points.
As the MCG Pro has larger hashtables, this could specially in the endgame give a bigger advantage.
I also played a few games MCG - MCG Pro. One game which the MCG won with white, I replayed the opening with the MCG Pro playing white, and it became a draw... So looking a bit deeper does not always give better results...

Puzzling 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
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

paulwise3 wrote:Hi Greg,

The MCG book should be the same as book 1 of the MCG Pro.
Book 2 is the Hiarcs book (= the default chosen book).

With several computers you can select a certain part of the opening book, such as passive, active, and tournament. With the MCG's that is not possible. So you are not sure that a certain opening in the book fits the playing style of the program. As for the difference in speed, it is often assumed that doubling the speed gives an elo raise of 50 - 70 points. But also, to be able to calculate one ply deeper, one generally needs a speed accelaration with factor 6. But it just depends how close you are to a next ply at a given level. For instance, the Excalibur Grandmaster tuned at double speed has an amazing raise of 127 points at active level (30 secs/move) on the wiki-elo-liste, whereas at the tournament level the raise is only a few points.
As the MCG Pro has larger hashtables, this could specially in the endgame give a bigger advantage.
I also played a few games MCG - MCG Pro. One game which the MCG won with white, I replayed the opening with the MCG Pro playing white, and it became a draw... So looking a bit deeper does not always give better results...

Puzzling regards,
Paul
Hi Paul,

Very helpful information regarding speed vs plys. I just finished games 2 and 3 - both resulted in fairly boring play that ended in draws from threefold repetition (they wanted to end the pain). During the middle game - I was observing the calculations of each machine. If the computer used the full analysis time - the MCG was typically at 6/7 plys at 30 seconds, and the MCG Pro between 7/8 plys (but not always). So, with the Pro's increase in speed, it appears that it sometimes could push one ply deeper (with equivalent material on the board).

I'm not sure if the MCG opening book is the same as the MCG Pro book 1. According to the specs, the MCG books has "> 20.000 Opening positions", and the Pro has the Classic London Book "with approximately 57,000 positions".

In game 2, the MCG (black) was "out of book" in move 10:

[fen]r1bqnrk1/ppp3pp/2np4/4pP2/2P5/2PP1NP1/P4PBP/R1BQ1RK1 w - - 0 10[/fen]

Yet the MCG Pro wasn't out of book until three moves later at move 13:

[fen]1r1q1rk1/ppp3pp/2npbn2/4p3/2P2P1N/2PP2P1/P5BP/1RBQ1RK1 w - - 1 14[/fen]

I'll continue to watch this in future games...

Deep-Plyple regards,
Greg
steffen
Full Member
Posts: 500
Joined: Sun Sep 27, 2015 12:30 pm

Post by steffen »

Hello together,

I had a lot of fun playing this tournament.
I would love to play it once more.
Any interest?

Best regards,
Steffen
donkeylane
Full Member
Posts: 679
Joined: Mon Aug 29, 2016 8:31 pm
Location: Cheshunt, Hertfordshire, UK

Post by donkeylane »

Hi Steffen,it would be interesting to see another high performer tournament, I have some rather curious results between two of the participants, namely the MCGE and R30, in the match I posted, the result was 6.5-5.5 in favour of the MCGE with the R30 on active setting, then I had 7-3 in favour of the MCGE with the R30 on normal style, and in the last couple of weeks, I have 7.5-3.5 in favour of the R30 again on active style setting. Surprising that the style settings have produced such a discrepancy. Regards Steve C.
Post Reply