Atlanta bug question

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.
User avatar
Dave C
Member
Posts: 460
Joined: Fri Jul 17, 2015 1:50 am
Location: SoCal, US

TTTT

Post by Dave C »

Hello 49,

I can hardly type because I am laughing sooo hard.. :lol: Your descriptions of the terrible play of both the Atlanta and M4-5 is priceless: "The endgame played by these two imbeciles now enters the “Top Ten Tabletop Turds” of all time."

Sometimes I play games that surprise me as being strong...and then I play a game that is a complete mess. My opponent's responses are key to these mixed results. It seems that top level computers like the Atlanta and Mach4-50 should play better than they did in your tournament. :shock:

Thanks for the commentary along with the games.

Dave
Larry
Senior Member
Posts: 2269
Joined: Wed Aug 01, 2007 8:42 am
Location: Gosford, NSW Australia

Re: TTTT

Post by Larry »

Dave C wrote:Hello 49,

I can hardly type because I am laughing sooo hard.. :lol: Your descriptions of the terrible play of both the Atlanta and M4-5 is priceless: "The endgame played by these two imbeciles now enters the “Top Ten Tabletop Turds” of all time."

Sometimes I play games that surprise me as being strong...and then I play a game that is a complete mess. My opponent's responses are key to these mixed results. It seems that top level computers like the Atlanta and Mach4-50 should play better than they did in your tournament. :shock:

Thanks for the commentary along with the games.

Dave
Although I haven't done it for a long time, it can be fun to run through
a game between two supposedly strong tabletops with eg, Hiarcs, and
watch the evaluations swing wildly. Positional and tactical errors are
being made routinely by both machines. The blunders would make
a grown man cry. And to think people used to trade their weeks paypacket
for a plastic pressure sensory chess computer.
L
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Re: TTTT

Post by Steve B »

Larry wrote:
Dave C wrote:Hello 49,

I can hardly type because I am laughing sooo hard.. :lol: Your descriptions of the terrible play of both the Atlanta and M4-5 is priceless: "The endgame played by these two imbeciles now enters the “Top Ten Tabletop Turds” of all time."

Sometimes I play games that surprise me as being strong...and then I play a game that is a complete mess. My opponent's responses are key to these mixed results. It seems that top level computers like the Atlanta and Mach4-50 should play better than they did in your tournament. :shock:

Thanks for the commentary along with the games.

Dave
Although I haven't done it for a long time, it can be fun to run through
a game between two supposedly strong tabletops with eg, Hiarcs, and
watch the evaluations swing wildly. Positional and tactical errors are
being made routinely by both machines. The blunders would make
a grown man cry. And to think people used to trade their weeks paypacket
for a plastic pressure sensory chess computer.
L
that's not fair..subjecting strong dedicated computers game play to strong PC engine analysis (running on far more powerful hardware)and then arm chair quarterbacking and yuking it up over their mistakes
how many of us..right now can defeat any of the dedicated computers mocked by you guys at normal or even blitz time controls?
my.guess...less then 5%(if that even)

also...try taking games played by GM's under tournament conditions and subject those games to Hiarcs analysis (or other powerful PC engines)
you will see the blunders fill up the screen as well

Fairness For The Old Chess Computers Regards
Steve
User avatar
paulwise3
Senior Member
Posts: 1505
Joined: Tue Jan 06, 2015 10:56 am
Location: Eindhoven, Netherlands

Re: TTTT

Post by paulwise3 »

Steve B wrote:
that's not fair..subjecting strong dedicated computers game play to strong PC engine analysis (running on far more powerful hardware)and then arm chair quarterbacking and yuking it up over their mistakes
how many of us..right now can defeat any of the dedicated computers mocked by you guys at normal or even blitz time controls?
my.guess...less then 5%(if that even)

also...try taking games played by GM's under tournament conditions and subject those games to Hiarcs analysis (or other powerful PC engines)
you will see the blunders fill up the screen as well

Fairness For The Old Chess Computers Regards
Steve
Quite right Steve. I see this also in our user tournaments: powerful engines showing a +6.0 advantage in he middlegame, and then the machine "screws" the position and loses. The operators hardly see the advantage themselves, but still complain about their machine losing a winning position...

We would have a very hard time even reaching the endgame phase against a Mach4 or Atlanta.

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...
Larry
Senior Member
Posts: 2269
Joined: Wed Aug 01, 2007 8:42 am
Location: Gosford, NSW Australia

Post by Larry »

Yes guys, but please forgive me. I wasn't meaning to speak badly
of the dedicated's, just pointing out the huge gulf in strength between
them and the engines. I actually avoid the new age wooden units
that have an engine in them. To me they are in a different league
and it's impossible to enjoy playing them knowing that I'm just
going to be squashed like a bug. The joy of the dedicated's comes
from knowing I'm in with a chance.
L
User avatar
Dave C
Member
Posts: 460
Joined: Fri Jul 17, 2015 1:50 am
Location: SoCal, US

Apology Offered

Post by Dave C »

You are right on Steve....I would be thrilled with a 5% success rate against chess computers. :)
Dave
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Well now that you guys were making fun of the Atlanta's play...
that Chinese Scam Artist is selling them for $20 a pop(shipping included)

https://www.ebay.com/itm/Schachcomputer ... SwxzdaCxjY

Grifter Regards
Steve
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

Steve B wrote:Well now that you guys were making fun of the Atlanta's play...
that Chinese Scam Artist is selling them for $20 a pop(shipping included)

https://www.ebay.com/itm/Schachcomputer ... SwxzdaCxjY

Grifter Regards
Steve
Just so we are all perfectly clear:
I do not, as you have asserted several times, make fun of my Atlanta.
It is an excellent and highly prized chess computer that I own.

It is however, frustrating to see the Atlanta calculate some very low level patzer play, knowing that Atlanta can play this type of game (from a 2014 tournament):


Game 36 Last game for Tasc, the champion! And…Atlanta WINS!

Finally, on the very last game, Tasc loses a game…to the former champ Atlanta!: fine game by Atlanta. D14 QGD semi-Slav, Chigorin defense in book for 11 moves. Tasc calculates 12. Qxd4 (book is Nxd4). Not a disaster as Stockfish still has the position even. Stockfish “curiously” calls for 16. e5 which quickly loses that pawn, although Stockfish isn’t too worried??? By move 25, Tasc is down -1.5 with no apparent bad moves and nothing to show for the pawn sacrifice…WHY THE F, CALL FOR 16. e5 THEN, JIT? On the other side, Atlanta is cautiously maneuvering to take eventual advantage of the extra Queenside pawn. Now, Tasc did start outplaying Atlanta and slowly whittled away at that lead. With weak move 33. .. Qc6?, the game is virtually a draw, thanks to the Tasc advanced pawn on c5. Both sides then maneuvered pieces, but played several “non-optimal” moves (Atlanta calculated slightly better and by move 38 had a -1.1 eval lead). Then Tasc disaster: 38 Qc3?? This single move can be considered: GAME OVER! Atlanta will win (38. Qb4 keeps Tasc in the game). From this point, Atlanta made (I believe I counted) ELEVEN STRAIGHT Stockfish recommended moves; Tasc was finished off, tout de suit. Tasc SHOULD have courteously resigned at move 43, but has been programmed to waste the user’s time until a forced “Mate in 7 or 8” is found.

[Date "Dec 22, 2014"]
[White "Tasc"]
[Black "Atlanta"]
[Result "0-1"]

1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Nf3 c6 5. e3 Nbd7 6. Bd3 Bd6 7. O-O O-O 8. e4
dxc4 9. Bxc4 e5 10. Bg5 Qe7 11. Re1 exd4 12. Qxd4 Ne5 13. Nxe5 Bxe5 14. Qe3 Qb4
15. Bxf6 Bxf6 16. e5 Bxe5 17. Qxe5 Qxc4 18. Rad1 Be6 19. b3 Qa6 20. Qg5 h6 21.
Qc5 Rfd8 22. Qe7 Re8 23. Qb4 Rad8 24. Qf4 Rxd1 25. Rxd1 Qa5 26. Qe3 Kf8 27. Qd4
c5 28. Qd3 a6 29. Ne4 c4 30. bxc4 Qxa2 31. Nd6 Rd8 32. c5 Qa4 33. Rb1 Qc6 34.
Qa3 Rd7 35. Re1 b5 36. Re5 Rd8 37. f3 Rb8 38. Qc3 b4 39. Qd3 Kg8 40. Nf5 Qc7
41. Nd6 b3 42. Qb1 b2 43. Kf2 Ba2 44. Qxa2 b1=Q 45. Qd5 Rb2+ 46. Kg3 Qg6+ 47.
Kf4 Qxg2 48. Qxf7+ Qxf7+ 49. Nxf7 Kxf7 50. h4 Qh2+ 51. Ke4 Resigns 0-1
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Four.nine wrote:
It is an excellent and highly prized chess computer that I own.

Here we can agree

Finding Common Ground Regards
Steve
Four.nine
Full Member
Posts: 911
Joined: Tue Mar 21, 2017 7:30 pm

Post by Four.nine »

Now that all appropriate genteel pleasantries have been thoroughly exchanged concerning the Atlanta, I would actually like to get back to the original thread subject.
Now... I forget, what was that?
Oh, yes: "Atlanta bug question".
I still would like to get a confirmation of the occasional E1-G1 lockup, i. e.

One other Atlanta quirk:

I have seen Atlanta suddenly go into "STOP" mode, when pressing E1 to castle.
Don't have a move sequence saved, but this has happened maybe 3-4 times in about 100 games.

Maybe somebody knows a guy, who knows someone, who knows somebody else, who has seen this.
And if this does happen to somebody: just press "GO".
Post Reply