Man vs machine - how to beat chess engine

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
pocopito
Posts: 238
Joined: Tue Jul 12, 2011 11:31 am
Contact:

Re: Man vs machine - how to beat chess engine

Post by pocopito » Fri Jun 08, 2012 10:40 am

On the other hand I guess games in which you lost but got a superior position could be interesting too (specially if you lost due to a blunder)
Two first meanings of the dutch word "leren":
1. leren [vc] (learn, larn, acquire) acquire or gain knowledge or skills.
2. leren [v] (teach, learn, instruct) impart skills or knowledge to.

User avatar
Kingghidorah
Posts: 224
Joined: Thu Nov 10, 2011 4:23 pm
Location: CT,USA

Re: Man vs machine - how to beat chess engine

Post by Kingghidorah » Fri Jun 08, 2012 12:04 pm

I went over the games quickly and it actually showed where Miroslav could've won quicker in some games. I did notice though that Critter 1.4 x64 version made better moves where it lost with him playing 1.4 x32 but all in all it looked like he beat the programs fair and square.
En passant,

Lonnie

"Never be bullied into silence. Never allow yourself to be made a victim. Accept no one's definition of your life; define yourself."

Harvey Fierstein

zullil
Posts: 5844
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Man vs machine - how to beat chess engine

Post by zullil » Fri Jun 08, 2012 1:12 pm

overlord wrote:Just check this positions more times. You will find that at the same depth engines doesn´t play always the same move...
Yes, this is true for parallel searches. In the search below, Critter does consider c4 for an instant. But the move is rejected at depth 14, in less than 1 second. It seems very odd that Critter with 3 threads and 256MB hash would play c4 on move 7 of a 5-minute game.

Not sure what is going on, but I suspect the engines you are defeating are much weaker than they should be.

Code: Select all

Critter v1.4 64-bit SSE4, by Richard Vida
setoption name Threads value 3
setoption name Hash value 256
setboard r1bqk2r/pp2bppp/2n1pn2/2pp4/3P4/2PBPN2/PP1N1PPP/R1BQ1RK1 b kq - 2 7
info string Opening book: "book.cbk" - 57853 entries [native]
info string GTB Init OK (5 piece set)
info string Using 3 thread(s), 256 MB hash
go infinite
 2/ 2  00:00       163  163000  -0.09 O-O Ng5 Bd7 Ndf3
 3/ 4  00:00       824  824000  +0.07 O-O Ne5 Nxe5 dxe5
 4/ 6  00:00      3195 1065000  +0.12 O-O b3 cxd4 exd4 Qc7
 5/ 6  00:00      3976  994000  +0.10 O-O b3 cxd4 exd4 Bd6 h3
 6/ 8  00:00      5203 1300750  +0.07 O-O b3 cxd4 exd4 Bd6 Ng5 h6
 7/10  00:00      9673 1209125  +0.19 O-O b3 Bd7 Bb2 cxd4 exd4 Qc7 Ng5 h6
 8/11  00:00     13593 1359300  +0.19 O-O b3 Bd7 Bb2 cxd4 exd4 Qc7 Ng5 h6
 9/13  00:00     30072 1670666  +0.16 O-O dxc5 Bxc5 b4 Bd6 b5 Ne5 Nxe5 Bxe5 Bb2 Qc7
10/14+ 00:00     82562 1965761  +0.25 O-O dxc5 Bxc5 b4 Bd6 b5 Ne5 Nxe5 Bxe5 Bb2 Qc7
10/14  00:00    117926 1902032  +0.13 c4 Bc2 O-O e4 b5 e5 Nd7 b3 h6 Bb2
11/20- 00:00    214946 2027792  +0.05 c4 Bc2 O-O e4 b5 e5 Nd7 b3 h6 Bb2
11/20  00:00    229958 2071693  +0.03 c4 Bc2 O-O e4 b5 e5 Nd7 b3 h6 Bb2 Bb7 bxc4 bxc4 Rb1 Qc7
12/25  00:00    484248 2373764  +0.07 c4 Bc2 O-O e4 Qc7 e5 Nd7 b3 b5 bxc4 bxc4 Rb1 f6 exf6 Nxf6 Ng5
13/25  00:00    621309 2408174  +0.05 c4 Bc2 O-O e4 Qc7 e5 Nd7 b3 b5 Bb2 Bb7 Qe2 h6 a4 a6
14/30  00:01   3317539 3168614  +0.01 O-O dxc5 Bxc5 b4 Bd6 b5 Ne7 c4 a6 cxd5 Nexd5 bxa6 b5 Bb2 Bxa6 Ne4 b4 Nxd6 Bxd3 Qxd3 Qxd6 Ne5
15/30  00:01   3924071 3232348  +0.05 O-O dxc5 Bxc5 b4 Bd6 b5 Ne7 c4 a6 Qc2 axb5 c5 Bb8 Bb2 Nf5 e4 dxe4 Nxe4 Nxe4 Bxe4 Qc7
16/32- 00:04  14487905 3557059  -0.03 O-O dxc5 Bxc5 b4 Bd6 b5 Ne7 c4 a6 Qc2 axb5 c5 Bb8 Bb2 Nf5 e4 dxe4 Nxe4 Nxe4 Bxe4 Qc7
16/34  00:04  17256617 3587654  +0.00 O-O dxc5 Bxc5 b4 Bd6 b5 e5 bxc6 e4 cxb7 Bxb7 Bb5 a6 Ba4 exf3 Nxf3 Ne4 Bb2 Rb8 Qd4 Bc5 Qe5 Bd6 Qd4
17/34  00:05  19002738 3601049  +0.00 O-O dxc5 Bxc5 b4 Bd6 b5 e5 bxc6 e4 cxb7 Bxb7 Bb5 a6 Ba4 exf3 Nxf3 Ne4 Bb2 Rb8 Qd4 Bc5 Qe5 Bd6 Qd4
18/38  00:11  42297517 3718136  -0.05 O-O dxc5 Bxc5 b4 Bd6 b5 Ne5 Nxe5 Bxe5 Bb2 a6 a4 h6 Qc2 Qc7 f4 Bd6 c4 Bc5 Bd4 Bxd4 exd4 Bd7 bxa6 Rxa6 c5
18/38  00:13  52535261 3767588  +0.00 b6 Ne5 Nxe5 dxe5 Nd7 f4 c4 Bc2 Qc7 e4 Bb7 exd5 Bxd5 Qh5 g6 Qh6 Bf8 Qh4 Be7 Qh6
19/41  00:26 103280757 3871092  +0.05 b6 Ne5 Nxe5 dxe5 Nd7 f4 c4 Bc2 Qc7 e4 Bb7 exd5 Bxd5 Nf3 O-O Be3 Nc5 Qe2 Rad8 Rad1 f5 Nd4 Bh4 Nf3 Qe7 Nxh4 Qxh4 Bxc5 bxc5
20/42  00:36 141721672 3907299  +0.03 b6 Ne5 Nxe5 dxe5 Nd7 f4 c4 Be2 O-O b3 b5 Qc2 Qa5 Nf3 Nc5 b4 Qa4 Qxa4 Nxa4 Bd2 a5 a3 Bd7 Nd4 g6 g3 axb4 axb4 Kg7
21/47  00:58 234880370 4004439  +0.03 b6 Ne5 Nxe5 dxe5 Nd7 Nf3 c4 Bc2 Qc7 e4 Bb7 exd5 Bxd5 Bf4 Qb7 Re1 b5 a4 b4 Qe2 Rd8 Rad1 O-O Be4 Bxe4 Qxe4 Qxe4 Rxe4
?
bestmove b6 ponder Ne5

time: 100646 nodes: 406994939 evals: 288255878 knps: 4043
phash: 87.71% evalcache: 10.60% lazy: 39.75% eg_recog: 55
splits: 9300

Sedat Canbaz
Posts: 3018
Joined: Thu Mar 09, 2006 10:58 am
Location: Antalya/Turkey
Contact:

Re: Man vs machine - how to beat chess engine

Post by Sedat Canbaz » Fri Jun 08, 2012 1:34 pm

Dear Kvicala Miroslav,

Just my two cents over this issue

Wow,wow,wow...it seems:
-You are Hero
-You are Star
-You are Genius

But however, i challenge you to play with my engine
My Conditions:i7 980X;10 games;5 min;Houdini 2.0c;Perfect 16 book;LIVE (against any cheating)
In case of any win by you or even ended with a draw,you will get a prize 5000 USD (by me)
And in case of 10 wins by Houdini against you,then you need to pay me 5000 USD

Are you ready for a such 'Challenge' ?

I hope you will agree...


Best Regards,
Sedat

User avatar
Kingghidorah
Posts: 224
Joined: Thu Nov 10, 2011 4:23 pm
Location: CT,USA

Re: Man vs machine - how to beat chess engine

Post by Kingghidorah » Fri Jun 08, 2012 1:58 pm

Sedat Canbaz wrote:Dear Kvicala Miroslav,

Just my two cents over this issue

Wow,wow,wow...it seems:
-You are Hero
-You are Star
-You are Genius

But however, i challenge you to play with my engine
My Conditions:i7 980X;10 games;5 min;Houdini 2.0c;Perfect 16 book;LIVE (against any cheating)
In case of any win by you or even ended with a draw,you will get a prize 5000 USD (by me)
And in case of 10 wins by Houdini against you,then you need to pay me 5000 USD

Are you ready for a such 'Challenge' ?

I hope you will agree...


Best Regards,
Sedat
Now that is a "Put your money where your mouth is" type of statement if I ever saw one!
En passant,

Lonnie

"Never be bullied into silence. Never allow yourself to be made a victim. Accept no one's definition of your life; define yourself."

Harvey Fierstein

User avatar
Don
Posts: 5106
Joined: Tue Apr 29, 2008 2:27 pm

Re: Man vs machine - how to beat chess engine

Post by Don » Fri Jun 08, 2012 3:58 pm

I don't have any reason to dispute your findings here - but I know that it's possible to beat a chess engine, it's just not likely, especially on a modern CPU. There is a pretty big difference in a powerful desktop system and a mobile device so I'm much less impressed by that.

There is also the issue of how to verify this - you add at least 100 ELO to yourself every time you are allowed to take back a move. It's not that we don't trust you but still a certain amount of verification is needed when the claim is incredible.

I know that it's possible to learn to beat a computer if you simply play it over and over again using the same plan each time and the same level and settings. Although the games will vary, if you stick with it you will learn how to handle all the ways it will vary and learn by trial an error. If the level and settings are the same you will only get a reasonably small variety of different "pathways" you must learn to beat it. To avoid big opening book variety you just find sound openings that are somewhat off-beat so that you get just a few openings.

I used to do this with the fidelity machines, but of course I could beat them once in a while without doing this. But I found a way to beat it every time which required learning only a few variations and figuring out what it takes to win. I think modern programs vary more than the old programs and it will no doubt be a lot more difficult also to find a way to win but it should be possible. Once I was a piece up I didn't memorize anything after that as I could just beat it on my own but that may not be so easy with a super grandmaster strength program.

overlord wrote:Many people thinks that at the present time chess engines are too strong to make some good results against them. I would like to decline those myths. Even very average player (FIDE ELO 2000 and higher) may be able to get a lot draws and some wins againts the strongest engines. I am very average player but I can beat most of engines on my Samsung 9220 and even on my old quad (Q6600). I must note that the question is not how strong are you but how can you adapt your play style against engines. That is the reason that average player with a lot of experience with engines may get better results than GM. Before I post some games, I would like to define general rules which must be respected to achieve good results.
1. Opening choice - there is no chance to play any Sicilian defence or openings like Moscow system, human is simply smashed by calculation power of engine. Very good opening choice anainst engine is Russian defence (especially system with Jc3 Jxc3), Caro-kann (system with g4 attack), many lines in French defence (strategy is much more important than tactics here), against some wekaker engines even Evans gambit can be used. But the weakest opening for engines is king´s indian defence (human can win with both colours - if computer has something like general book 8-9 moves).
2. To keep comfortable draw it is necessary to have symetrical structure of pawns on both sides and to keep at least 6 pawns on chessboard (each side).
3. Right now engines are absolutly superior to human in tactics late middle game (especially Houdini is really king of the hill) and endgame are also at very solid level (with table bases). On the other hand every engine is limited by it´s horizon. That is why opening lines with pawns sacrifice and clear attacking strategy are so unpleasant for engines.
4. Maybe it will sound little bit strange but thera are hihger chances for human playing 5 minute blitz than long games - it is also associated with horizon which I discussed earlier.

Before someone will accuse me that I was playing against engine level 1800, here are engines settings:

For Samsung 9220 (2 cores 1,4 GHz, full strength, ponder on,Droidfish GUI)
For Q6600 (3 cores for Houdini and all engines that support more than 1 core, 256 MB hash, ponder on, default ponder) - someone may tell that increasing contempt will reduce my chances rapidly but on the other hand original setting is the strongest and engine doesn´t know who is his opponent. When Houdini is playing on playchess it also doesn´t change it´s contept playing weaker engines or slower PC).

Well, let´s see some games:

[Event "5 min. blitz"]
[Site "?"]
[Date "2012.06.03"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Houdini 1.5a x64"]
[Result "1/2-1/2"]
[ECO "C42"]
[WhiteElo "2035"]
[Annotator ",Microsoft"]
[PlyCount "254"]
[TimeControl "300"]

{319MB, Mirekbook.ctg, Vlastník-PC} 1. e4 {0} e5 {B/0 0} 2. Nf3 {1} Nf6 {B/0 0}
3. Nxe5 {1} d6 {B/0 0} 4. Nf3 {1} Nxe4 {B/0 0} 5. Nc3 {1} Nxc3 {B/0 0} 6. dxc3
{1} Be7 {B/0 0} 7. Be3 {1} Nc6 {B/0 0} 8. Qd2 {1} O-O {B/0 0} 9. O-O-O {2} Ne5
{B/0 0} 10. Nxe5 {2} dxe5 {0.14/17 4} 11. Bc4 {(Qxd8) 4} Qxd2+ {0.09/18 10} 12.
Rxd2 {1} Bf5 {0.13/19 14} 13. Rhd1 {3} a6 {0.08/18 12} 14. f3 {(Bd3) 14} Rad8 {
0.03/19 9} 15. a3 {(Bd3) 12} Rxd2 {0.00/19 9} 16. Rxd2 {(Bxd2) 1} Re8 {
-0.01/19 6} 17. Bd5 {(Ba2) 1} Bc8 {-0.03/20 7} 18. c4 {(b4) 10} c6 {-0.09/19 5}
19. Be4 {1} f5 {-0.13/19 13} 20. Bd3 {2} c5 {-0.13/19 6} 21. f4 {(Kd1) 12} e4 {
-0.21/21 4} 22. Be2 {(Bf1) 1} Kf7 {-0.25/21 5} 23. g3 {(Bf1) 2} b6 {-0.24/21 10
} 24. h4 {(c3) 3} Ke6 {-0.31/18 4} 25. a4 {(h5) 7} a5 {-0.34/18 5} 26. c3 {6}
Rg8 {-0.32/19 2} 27. Kc2 {9} g6 {-0.32/22 0} 28. b3 {2} Bf6 {-0.32/19 2} 29.
Bd1 {(Rd1) 7} Bd7 {-0.34/20 5} 30. Be2 {(Rh2) 2} h6 {-0.38/21 11} 31. Bd1 {
(Rd5) 12} g5 {-0.39/21 5} 32. hxg5 {12} hxg5 {-0.39/24 0} 33. Rh2 {(Bh5) 2} g4
{-0.39/21 4} 34. Be2 {(Rh6) 5} Bg7 {-0.40/22 5} 35. Bf1 {(Bf2) 5} Be8 {
-0.40/23 3} 36. Be2 {1} Bg6 {-0.40/23 2} 37. Bf1 {(Rh1) 1} Rd8 {-0.40/23 3} 38.
Be2 {(Rh1) 4} Rb8 {-0.40/24 3} 39. Bf1 {1} Rg8 {-0.35/24 21} 40. Be2 {(Rh1) 2}
Rc8 {-0.40/24 3} 41. Bf1 {(Rh1) 1} Rc7 {-0.40/22 6} 42. Be2 {(Rd2) 1} Be8 {
-0.40/23 2} 43. Bf1 {(Bc1) 1} Rc8 {-0.40/23 4} 44. Be2 {(Bc1) 3} Rd8 {
-0.40/23 2} 45. Bf1 {(Rh7) 7} Kf6 {-0.40/23 2} 46. Be2 {(Rd2) 1} Bf7 {
-0.40/23 2} 47. Bf1 {1} Rb8 {-0.40/22 1} 48. Be2 {1} Ke6 {-0.40/23 1} 49. Bf1 {
(Rh1) 1} Be8 {-0.40/23 2} 50. Be2 {(Rh7) 1} Rc8 {-0.40/23 2} 51. Bf1 {1} Rd8 {
-0.40/23 1} 52. Be2 {(Rd2) 1} Bc6 {-0.40/22 2} 53. Bf1 {(Rh7) 1} Rc8 {
-0.40/22 2} 54. Be2 {(Kd2) 0} Rc7 {-0.40/22 2} 55. Bf1 {1} Rb7 {-0.40/23 1} 56.
Be2 {(Rd2) 1} Rd7 {-0.40/24 4} 57. Bf1 {0} Re7 {-0.40/24 3} 58. Be2 {(Rd2) 0}
Rf7 {-0.40/22 2} 59. Bf1 {(Bd1) 0} Rc7 {-0.40/23 2} 60. Be2 {(Re2) 1} Ra7 {
-0.37/19 2} 61. Bf1 {(Bc1) 1} Bd7 {-0.37/19 1} 62. Be2 {(Bc1) 1} Be8 {
-0.37/19 1} 63. Bf1 {(Rh1) 1} Rb7 {-0.32/19 8} 64. Be2 {(Bc1) 1} Bg6 {
-0.32/18 1} 65. Bf1 {(Rh1) 1} Bf6 {-0.30/19 1} 66. Be2 {(Re2) 1} Be8 {
-0.28/18 2} 67. Bf1 {(Bc1) 1} Rb8 {-0.28/18 3} 68. Be2 {(Rh7) 1} Kf7 {
-0.27/17 1} 69. Bf1 {(Rh7+) 1} Bc6 {-0.24/17 1} 70. Be2 {(Bg1) 1} Kg6 {
-0.24/16 1} 71. Bf1 {1} Rc8 {-0.21/16 1} 72. Be2 {(Re2) 1} Rd8 {-0.19/17 1} 73.
Bf1 {1} Bg7 {-0.19/18 0} 74. Be2 {(Rd2) 1} Kf7 {-0.17/16 1} 75. Bf1 {1} Ke6 {
-0.15/16 1} 76. Be2 {(Rh7) 1} Rh8 {-0.11/16 2} 77. Rxh8 {(Rg2) 1} Bxh8 {
-0.18/16 0} 78. Bf1 {(Bf2) 1} Bf6 {-0.17/22 1} 79. Be2 {(Kd2) 1} Bd7 {
-0.17/23 1} 80. Bf1 {(Bf2) 1} Bg7 {-0.17/24 1} 81. Bg2 {(Kd2) 3} Be8 {
-0.17/22 1} 82. Bf1 {(Kd2) 1} Bf7 {-0.17/24 1} 83. Bg2 {(Bf2) 1} Kd6 {
-0.17/23 1} 84. Bf1 {(Kd2) 1} Be6 {-0.17/24 1} 85. Bg2 {(Kd2) 1} Kc7 {
-0.17/23 1} 86. Bf1 {1} Bd7 {-0.17/24 0} 87. Bg2 {(Kd2) 1} Be8 {-0.17/23 1} 88.
Bf1 {(Kd2) 1} Bf8 {-0.17/25 1} 89. Bg2 {(Kd2) 1} Kc6 {-0.17/24 1} 90. Bf1 {
(Kd2) 1} Bf7 {-0.17/25 1} 91. Bg2 {(Kd2) 1} Bg8 {-0.17/24 1} 92. Bf1 {1} Bg7 {
-0.17/24 0} 93. Bg2 {(Kd2) 1} Bh7 {-0.17/24 1} 94. Bf1 {(Kd2) 0} Bg6 {
-0.17/26 1} 95. Bg2 {(Kd2) 1} Bh8 {-0.17/25 1} 96. Bf1 {(Kd2) 1} Be8 {
-0.17/26 1} 97. Bg2 {(Kd2) 1} Bf6 {-0.17/24 1} 98. Bf1 {(Bf2) 1} Bf7 {
-0.17/25 1} 99. Bg2 {(Kd2) 1} Be6 {-0.17/24 1} 100. Bf1 {(Kd2) 1} Kd6 {
-0.17/24 1} 101. Bg2 {(Kd2) 1} Bc8 {-0.17/23 1} 102. Bf1 {(Kd2) 1} Be7 {
-0.17/24 0} 103. Bg2 {(Kd2) 0} Bb7 {-0.17/23 1} 104. Bf1 {(Kd2) 0} Bf8 {
-0.17/24 1} 105. Bg2 {(Kd2) 0} Ke6 {-0.17/23 0} 106. Bf1 {(Kd2) 1} Bc6 {
-0.17/23 1} 107. Bg2 {(Kd2) 0} Be8 {-0.17/22 0} 108. Bf1 {(Kd2) 1} Kd6 {
-0.17/22 0} 109. Bg2 {(Kd2) 0} Be7 {-0.17/23 0} 110. Bf1 {0} Bf7 {-0.17/25 0}
111. Bg2 {(Kd2) 1} Bf6 {-0.17/23 0} 112. Bf1 {(Kd2) 1} Bg7 {-0.09/22 3} 113.
Bg2 {(Bc1) 1} Bh8 {-0.09/20 1} 114. Bf1 {(Bc1) 1} Kc6 {-0.09/22 1} 115. Bg2 {0}
Kc7 {-0.09/22 0} 116. Bf1 {(Bc1) 1} Bg8 {-0.09/20 1} 117. Bg2 {1} Kd6 {
-0.09/22 0} 118. Bf1 {(Bc1) 1} Kd7 {-0.04/20 2} 119. Bg2 {1} Kc7 {0.00/18 0}
120. Bf1 {(Bc1) 1} Bg7 {0.00/16 0} 121. Bg2 {1} Bf7 {0.00/19 0} 122. Bf1 {
(Bc1) 1} Bf8 {0.00/15 0} 123. Bg2 {(Kd2) 1} Kd6 {0.00/16 0} 124. Bf1 {(Kd1) 1}
Bg6 {0.00/19 0} 125. Bg2 {1} Bh5 {0.00/24 0} 126. Bf1 {1} Bg7 {0.00/28 0} 127.
Bg2 {(Bc1) 1} Bf8 {0.00/27 0} 1/2-1/2

[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.05.03"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Critter 1.4 32-bit"]
[Result "1-0"]
[ECO "D05"]
[WhiteElo "2028"]
[PlyCount "75"]

1. d4 d5 2. Nf3 Nf6 3. e3 e6 4. Bd3 Be7 5. O-O c5 6. c3 Nc6 7. Nbd2 c4 8. Bb1
b5 9. a3 O-O 10. e4 Nd7 11. Re1 Re8 12. e5 Bb7 13. Nf1 Nf8 14. g4 a5 15. Ng3 a4
16. g5 Na5 17. Bc2 Nb3 18. Bxb3 axb3 19. Bd2 Qd7 20. Qe2 Qc8 21. Nh5 Ng6 22.
Nf6+ gxf6 23. gxf6 Bf8 24. Ng5 h6 25. Qh5 hxg5 26. Bxg5 Be7 27. Bh6 Bc5 28. Re3
Re7 29. Rg3 Rd7 30. Kh1 Rxa3 31. bxa3 Bf8 32. Rag1 Bg7 33. Bxg7 Qf8 34. Bxf8 b4
35. Rxg6+ fxg6 36. Rxg6+ Kxf8 37. Qh8+ Kf7 38. Rg7# 1-0

[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.05.26"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Critter 1.4 32-bit"]
[Result "1-0"]
[ECO "C51"]
[WhiteElo "2035"]
[PlyCount "75"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Bd6 6. O-O Nf6 7. Re1 O-O 8. d4
Rb8 9. Bg5 h6 10. Bh4 b5 11. Bb3 Qe7 12. Nbd2 Re8 13. Nf1 Qd8 14. Ng3 exd4 15.
cxd4 Bxg3 16. hxg3 g5 17. Nxg5 hxg5 18. Bxg5 d6 19. Qd3 Kg7 20. Rad1 Qe7 21. f4
Nb4 22. Qb1 Qf8 23. Bxf6+ Kxf6 24. e5+ Kg7 25. f5 c5 26. Qe4 d5 27. Qg4+ Kh7
28. Kf2 Qg7 29. Rh1+ Kg8 30. Qf4 cxd4 31. f6 Rb6 32. fxg7 Bd7 33. Qf6 Rxf6+ 34.
exf6 Bh3 35. Rxh3 Nd3+ 36. Rxd3 Re2+ 37. Kxe2 a6 38. Rh8# 1-0

[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.05.27"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Cuckoo chess"]
[Result "1-0"]
[ECO "C51"]
[WhiteElo "2035"]
[PlyCount "59"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. b4 Bxb4 5. c3 Bd6 6. O-O a6 7. d4 Nf6 8. Re1
O-O 9. Bg5 Qe7 10. Nbd2 b5 11. Bb3 Bb7 12. Nf1 h6 13. Bh4 g5 14. Nxg5 hxg5 15.
Bxg5 exd4 16. cxd4 Nxd4 17. Qxd4 Be5 18. Qe3 Bxa1 19. Qg3 Be5 20. Bf4+ Ng4 21.
Qxg4+ Bg7 22. e5 Qb4 23. Re3 Rfb8 24. Rg3 Qf8 25. Bh6 Be4 26. Bxg7 d5 27. Bxd5
Qc5 28. Bh6+ Bg6 29. Qxg6+ Kh8 30. Qg7# 1-0

[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.05.28"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish 2.2.2"]
[Result "1-0"]
[ECO "C42"]
[WhiteElo "2035"]
[PlyCount "123"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. Nc3 Nxc3 6. dxc3 Be7 7. Be3 O-O
8. Qd2 c5 9. O-O-O Be6 10. Kb1 h6 11. Bxh6 gxh6 12. Qxh6 Bf5 13. Rd5 Bg6 14.
Bd3 Bxd3 15. Rh5 f6 16. cxd3 Kf7 17. Nh4 Ke8 18. Nf5 Nc6 19. Re1 Kd7 20. Qe3
Re8 21. Rh7 Kc7 22. Nxe7 Rxe7 23. Rxe7+ Nxe7 24. Qxe7+ Qxe7 25. Rxe7+ Kb6 26.
h4 Rf8 27. g4 Rh8 28. h5 f5 29. f3 fxg4 30. fxg4 Rg8 31. Re4 Kc6 32. c4 Kd7 33.
Kc2 Rh8 34. Kd2 Rg8 35. Ke3 Rh8 36. Kf4 a5 37. Kg5 d5 38. cxd5 Kd6 39. h6 b5
40. Kg6 Kxd5 41. Rf4 Ke5 42. Rf5+ Kd4 43. Kg7 Rd8 44. h7 a4 45. h8=Q Rxh8 46.
Kxh8 a3 47. bxa3 c4 48. dxc4 bxc4 49. Rf4+ Kd3 50. Rxc4 Kxc4 51. g5 Kd5 52. g6
Kc4 53. g7 Kb5 54. g8=Q Kb6 55. Qc4 Kb7 56. Kg7 Kb6 57. Kf7 Kb7 58. Ke7 Kb6 59.
Kd7 Ka5 60. Qb4+ Ka6 61. Kc7 Ka7 62. Qb7# 1-0

[Event "10 min. game"]
[Site "Microsoft"]
[Date "2012.06.06"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish 2.2.2"]
[Result "1-0"]
[ECO "C02"]
[WhiteElo "2035"]
[PlyCount "173"]

1. e4 e6 2. d4 d5 3. e5 c5 4. c3 Nc6 5. Nf3 Qb6 6. a3 c4 7. Be2 Na5 8. Nbd2 Bd7
9. O-O Ne7 10. Qc2 Nf5 11. Rb1 Be7 12. Re1 O-O 13. Nf1 Nb3 14. Bg5 Bxg5 15.
Nxg5 Qd8 16. f4 h6 17. Nf3 Ba4 18. Bd1 Nc5 19. Qd2 Bxd1 20. Rexd1 Ne4 21. Qe2
Rc8 22. N1d2 Nxd2 23. Rxd2 Qb6 24. g4 Ne7 25. Qf2 Qb3 26. Nh4 b5 27. f5 g5 28.
f6 Nc6 29. Qe3 Kh7 30. Nf3 Rg8 31. Kf2 Rg6 32. Rh1 a5 33. h4 b4 34. axb4 axb4
35. hxg5 bxc3 36. bxc3 Rh8 37. Rxh6+ Rxh6 38. gxh6 Qb1 39. Kg2 Qe4 40. Qxe4+
dxe4 41. Ng5+ Kg6 42. Nxe4 Rxh6 43. Nd6 Rh4 44. Kg3 Rh8 45. Nxc4 Rh1 46. Ne3
Rh8 47. Kf4 Nb8 48. c4 Na6 49. d5 Nc5 50. d6 Ra8 51. Nf1 Ra3 52. Ng3 Rc3 53.
Rd4 Nd7 54. Ne4 Rc1 55. c5 Kh6 56. Rd3 Kg6 57. Rc3 Rxc3 58. Nxc3 Nxc5 59. Na4
Nd7 60. Nb2 Kh7 61. Nd3 Kg8 62. Ke4 Kf8 63. Kd4 Kg8 64. g5 Kh7 65. Nf4 Kh8 66.
g6 Kg8 67. g7 Kh7 68. Nd3 Kg8 69. Nc5 Nb8 70. Na6 Nc6+ 71. Ke4 Kh7 72. d7 Nd8
73. Nc7 Nb7 74. Kd4 Na5 75. Kc5 Kg6 76. g8=Q+ Kf5 77. Qxf7 Ke4 78. Qxe6 Nb7+
79. Kc4 Nd8 80. Qd5+ Ke3 81. Qd4+ Ke2 82. Nd5 Nc6 83. Qe3+ Kf1 84. Nf4 Nxe5+
85. Kc3 Nd3 86. Qe2+ Kg1 87. Qg2# 1-0

[Event "30 min. game"]
[Site "Microsoft"]
[Date "2012.03.29"]
[Round "?"]
[White "Critter 1.4 32-bit"]
[Black "Kvicala, Miroslav"]
[Result "0-1"]
[ECO "E97"]
[BlackElo "2028"]
[PlyCount "136"]

1. Nf3 Nf6 2. c4 g6 3. Nc3 Bg7 4. d4 O-O 5. e4 d6 6. Be2 e5 7. O-O Nc6 8. d5
Ne7 9. h3 Nd7 10. Bg5 h6 11. Be3 f5 12. Rc1 f4 13. Bd2 g5 14. c5 Nf6 15. Qb3
Ng6 16. Bd3 Rf7 17. c6 b6 18. Nh2 h5 19. Rce1 g4 20. hxg4 hxg4 21. Nb5 Ne8 22.
g3 f3 23. Rd1 Bf6 24. Qc3 Qe7 25. Rc1 Rh7 26. Nxf3 gxf3 27. Bc2 Bg4 28. Bd1 Nh4
29. Na3 Qg7 30. Nc4 Bg5 31. Re1 Qh6 32. gxh4 Qxh4 33. Bxf3 Qh2+ 34. Kf1 Bh4 35.
Ke2 Qxf2+ 36. Kd3 Bxf3 37. Kc2 Rf7 38. Kb3 Nf6 39. Rf1 Nxe4 40. Rxf2 Nxc3 41.
Rg1+ Kf8 42. Rxf3 Rxf3 43. Bxc3 Re8 44. Nd2 Rg3 45. Rh1 Bg5 46. Ne4 Rg4 47.
Rh8+ Ke7 48. Rh7+ Kd8 49. Nf2 Rg2 50. Ne4 Be7 51. Kc4 a6 52. Bd2 Reg8 53. Kd3
R2g4 54. Rh2 Rh4 55. Re2 Rgg4 56. a4 Ke8 57. a5 b5 58. b3 Kf7 59. Re3 Kg6 60.
Re1 Rh3+ 61. Be3 Bg5 62. Nxg5 Rxg5 63. Rf1 Rgg3 64. Re1 Rxe3+ 65. Rxe3 Rxe3+
66. Kxe3 Kf5 67. Kd3 Kf4 68. Ke2 Ke4 0-1



[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.06.02"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish 2.2.2"]
[Result "1/2-1/2"]
[ECO "B19"]
[WhiteElo "2035"]
[PlyCount "76"]

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. h4 h6 7. Nf3 Nd7 8. h5
Bh7 9. Bd3 Bxd3 10. Qxd3 e6 11. Bd2 Ngf6 12. O-O-O Qc7 13. Kb1 Bd6 14. Ne4 Nxe4
15. Qxe4 O-O 16. g4 Nf6 17. Qe2 Nh7 18. g5 Nxg5 19. Nxg5 hxg5 20. Bxg5 Bf4 21.
Rdg1 Bxg5 22. Rxg5 Qf4 23. Rhg1 Qxd4 24. Rxg7+ Qxg7 25. Rxg7+ Kxg7 26. Qe5+ Kh6
27. Qf6+ Kxh5 28. f4 Kg4 29. Qg5+ Kf3 30. Qh5+ Kf2 31. Qh4+ Ke2 32. Qg4+ Ke3
33. Qg3+ Kd2 34. Qf2+ Kd1 35. Qf1+ Kd2 36. Qf2+ Kd1 37. Qf1+ Kd2 38. Qf2+ Kd1
1/2-1/2

and many more similar results...
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.

overlord
Posts: 198
Joined: Sun Jun 03, 2012 4:46 pm
Location: Trinec, Czech Republic

Re: Man vs machine - how to beat chess engine

Post by overlord » Fri Jun 08, 2012 7:47 pm

I don´t know exactly the statistics but it will be something like 1 win - 5 draw - 20 lost (all played with white). Playing with black pieces I won only two games and lost maybe 200...

This looks maybe suspicious...nut human have one big advantage over computer evolution. If I find out that some opening or line is not suitable, I am looking for something different. If I find better opening, I chech the lines that engines play, if I lose the game I can learn from my mistakes and next time I play better ...it is something like Hiarcs auto-learning function...if you let play Hiarcs many thousand games against Houdini (few openings or lines) it will improve it´s performance and in some point it will be able to beat Houdini. I use the same procedure but of course every human can learn and improve much faster...it is no magic...everybody can make some wins and draws too...

One more example from today...played on smartphone...

[Event "5 min. game"]
[Site "Microsoft"]
[Date "2012.06.08"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Komodo 3"]
[Result "1/2-1/2"]
[WhiteElo "2035"]

{256MB, Mirekbook.ctg} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6.
h4 h6 7. Nf3 Nd7 8. h5 Bh7 9. Bd3 Bxd3 10. Qxd3 e6 11. Bd2 Ngf6 12. O-O-O Bd6
13. Ne4 Nxe4 14. Qxe4 Nf6 15. Qe2 Qc7 16. Ne5 O-O 17. Kb1 Rfd8 18. g4 c5 19. g5
cxd4 20. gxf6 Bxe5 21. fxg7 d3 22. cxd3 Bxb2 23. Kxb2 Qb6+ 24. Ka1 Qd4+ 25. Kb1
Qb6+ 26. Ka1 Qd4+ 27. Kb1 Qb6+ 28. Ka1 1/2-1/2

overlord
Posts: 198
Joined: Sun Jun 03, 2012 4:46 pm
Location: Trinec, Czech Republic

Re: Man vs machine - how to beat chess engine

Post by overlord » Fri Jun 08, 2012 7:54 pm

OK :) So here are some lost games... :( Generally said playing black against computer is very tough...

[Event "Blitz:10'"]
[Site "Microsoft"]
[Date "2012.04.20"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Critter 1.4 32-bit"]
[Result "0-1"]
[ECO "E97"]
[WhiteElo "2028"]
[PlyCount "224"]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5
Nd4 9. Be3 Re8 10. Nxd4 exd4 11. Bxd4 Nxe4 12. Bxg7 Kxg7 13. Nxe4 Rxe4 14. Bd3
Re8 15. Re1 Rxe1+ 16. Qxe1 Qf6 17. Qd2 Qd4 18. Rd1 Bd7 19. Be2 Qe5 20. f4 Qf6
21. Qd4 Qxd4+ 22. Rxd4 Bf5 23. Kf2 Bb1 24. a3 Re8 25. g3 a5 26. Bf3 Rb8 27. a4
Re8 28. b3 h5 29. h4 Bc2 30. Bd1 Bf5 31. Bf3 Re7 32. Rd2 Kf6 33. Rd4 Bc2 34.
Bd1 Bb1 35. Bf3 Bf5 36. Rd2 Kg7 37. Rd4 Kh6 38. Rd2 b6 39. Rd4 Kg7 40. Rd2 Kf6
41. Rd4 Re8 42. Rd2 Kg7 43. Rd4 Kh7 44. Rd2 Re7 45. Rd4 Kh8 46. Rd2 Kg7 47. Rd4
Kf8 48. Rd2 Re8 49. Rd4 Ke7 50. Rd2 Kd8 51. Rd4 Kd7 52. Rd2 Re7 53. Rd4 Bh3 54.
Rd2 Ke8 55. Rd4 Bf5 56. Rd2 Kd8 57. Rd4 Re8 58. Rd2 Kd7 59. Rd4 Kc8 60. Rd2 Re7
61. Rd4 Kb7 62. Rd2 Bh3 63. Rd4 Kc8 64. Rd2 Bd7 65. Rd4 Bf5 66. Rd2 Re8 67. Rd4
Bh3 68. Rd1 Kd8 69. Rd4 Bd7 70. Rd2 Re7 71. Rd4 Bf5 72. Rd2 Ke8 73. Rd4 Bc8 74.
Rd2 Kf8 75. Rd4 Bd7 76. Rd2 Re8 77. Rd4 Bf5 78. Rd2 Kg8 79. Rd4 Bc2 80. Bd1 Bb1
81. Bf3 Kh7 82. Rd2 Bf5 83. Rd4 Bh3 84. Rd2 Kh8 85. Rd4 Kg7 86. Rd2 f6 87. Rd4
Bf5 88. Rd2 Re7 89. Rd4 Kh6 90. Rd2 g5 91. hxg5+ fxg5 92. fxg5+ Kxg5 93. Rd4
Rf7 94. Kg2 Rf8 95. Rd2 Rg8 96. Rd4 Re8 97. Kf2 Bc2 98. Bd1 Be4 99. Bf3 Bxf3
100. Kxf3 Re1 101. Rd3 Kf5 102. Rd4 Rf1+ 103. Ke3 Kg5 104. Ke2 Rc1 105. Rd3 Kg4
106. Kf2 Rc2+ 107. Kg1 Ra2 108. Kf1 Rb2 109. Kg1 Kf5 110. Kf1 Ke4 111. Rc3 Rd2
112. Kg1 Rd3 0-1

[Event "Blitz:10'"]
[Site "Microsoft"]
[Date "2012.04.04"]
[Round "?"]
[White "Komodo 3 "]
[Black "Kvicala, Miroslav"]
[Result "1-0"]
[ECO "C67"]
[BlackElo "2028"]
[PlyCount "69"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. O-O Nxe4 5. d4 Nd6 6. Bxc6 dxc6 7. dxe5 Nf5
8. Qxd8+ Kxd8 9. Rd1+ Ke8 10. Nc3 Be7 11. Bg5 Bxg5 12. Nxg5 h6 13. Nge4 b6 14.
a4 a5 15. Rac1 Be6 16. h3 Rd8 17. Rxd8+ Kxd8 18. Rd1+ Kc8 19. g4 Ne7 20. f4 Rd8
21. Rxd8+ Kxd8 22. Kf2 Nd5 23. Nxd5 cxd5 24. Ng3 Bd7 25. Nf5 Bxf5 26. gxf5 c5
27. Kf3 Ke7 28. Kg4 Kd7 29. Kh5 Ke7 30. h4 d4 31. b3 Kf8 32. f6 Kg8 33. fxg7
Kxg7 34. f5 Kh7 35. f6 1-0

[Event "15 min. game"]
[Site "Microsoft"]
[Date "2012.04.02"]
[Round "?"]
[White "Komodo 3"]
[Black "Kvicala, Miroslav"]
[Result "1-0"]
[ECO "B01"]
[BlackElo "2028"]
[PlyCount "66"]

1. e4 d5 2. exd5 Qxd5 3. Nc3 Qa5 4. d4 Nf6 5. Nf3 Bf5 6. Bc4 e6 7. Bd2 c6 8.
Qe2 Bb4 9. Ne5 Nbd7 10. Nxd7 Nxd7 11. a3 Nf6 12. O-O-O Bxc3 13. Bxc3 Qc7 14. f3
h5 15. Qe5 Qxe5 16. dxe5 Nd5 17. Bd4 Nb6 18. Bb3 O-O-O 19. Rd2 Rd7 20. Rhd1
Rhd8 21. a4 Kc7 22. a5 Nc8 23. Bc4 a6 24. h4 Bh7 25. b4 Bf5 26. Be2 Rd5 27. g4
Bg6 28. Bc4 R5d7 29. Bf1 Rd5 30. Be2 R5d7 31. Bc4 Kb8 32. Ba7+ Kc7 33. Bb8+
Kxb8 1-0

[Event "15 min. game"]
[Site "Microsoft"]
[Date "2012.03.08"]
[Round "?"]
[White "Kvicala, Miroslav"]
[Black "Stockfish 2.2.2, Samsung 9220"]
[Result "0-1"]
[ECO "E64"]
[WhiteElo "2028"]
[PlyCount "72"]

1. d4 g6 2. c4 Bg7 3. Nc3 d6 4. Nf3 Na6 5. g3 Nf6 6. Bg2 O-O 7. O-O c5 8. d5
Bf5 9. Nd2 Qd7 10. e4 Bh3 11. Qc2 Bxg2 12. Kxg2 e6 13. a3 Nc7 14. b3 b5 15. Bb2
Rfe8 16. Rfe1 bxc4 17. bxc4 Rab8 18. Rab1 Nh5 19. Ba1 Bd4 20. Rxb8 Rxb8 21. Rb1
Rxb1 22. Qxb1 e5 23. Ne2 Ne8 24. Bxd4 exd4 25. Nc1 Qg4 26. f3 Qg5 27. Qc2 Qe3
28. Qd3 Qg5 29. Ne2 Qd8 30. Kf2 Nhf6 31. Nf4 g5 32. Ne2 Nd7 33. f4 Nef6 34. h3
g4 35. h4 Qa5 36. Nc1 Qxd2+ 0-1

[Event "15 min. game"]
[Site "Microsoft"]
[Date "2012.03.03"]
[Round "?"]
[White "Stockfish 2.2.2, Samsung 9220"]
[Black "Kvicala, Miroslav"]
[Result "1-0"]
[ECO "B06"]
[BlackElo "2028"]
[PlyCount "87"]

1. e4 d6 2. d4 g6 3. Nc3 Bg7 4. Nf3 e6 5. Be2 Ne7 6. h4 h6 7. Qd3 b6 8. Bf4 Bb7
9. O-O-O Nd7 10. Qd2 a6 11. Kb1 Rb8 12. Qe3 Qc8 13. g4 Qd8 14. Rhg1 Qc8 15. e5
d5 16. Na4 c5 17. Rge1 c4 18. Nc3 b5 19. a3 Bc6 20. Qd2 a5 21. Na2 Qa6 22. c3
Qb6 23. Rh1 Qb7 24. Rdg1 Nb6 25. Ne1 Na8 26. h5 g5 27. Be3 Nc7 28. Nc2 Na6 29.
f4 gxf4 30. Bxf4 b4 31. axb4 Ba4 32. Bd1 Bxc2+ 33. Bxc2 axb4 34. Ba4+ Kf8 35.
Rh3 b3 36. Nc1 Nc7 37. g5 Nf5 38. gxh6 Bxh6 39. Bxh6+ Rxh6 40. Qg5 Ra8 41. Rf3
Rxa4 42. Rxf5 Ra1+ 43. Kxa1 Qa6+ 44. Kb1 1-0

[Event "15 min. game"]
[Site "Microsoft"]
[Date "2012.03.01"]
[Round "?"]
[White "Stockfish 2.2.2, Samsung 9220"]
[Black "Kvicala, Miroslav"]
[Result "1-0"]
[ECO "B01"]
[BlackElo "2028"]
[PlyCount "69"]

1. e4 d5 2. exd5 Qxd5 3. Nc3 Qa5 4. d4 Nf6 5. Nf3 Bf5 6. Bc4 e6 7. Bd2 c6 8.
Qe2 Bb4 9. Ne5 Nbd7 10. Nxd7 Nxd7 11. a3 Nf6 12. O-O-O Bxc3 13. Bxc3 Qc7 14. f3
b5 15. Ba2 O-O 16. g4 Bg6 17. Qe5 Qxe5 18. dxe5 Nd5 19. Bd4 a5 20. Bc5 Rfe8 21.
Rd4 Ra6 22. a4 b4 23. Bc4 Raa8 24. Re1 h5 25. h3 Red8 26. Rd2 Re8 27. Rg1 Kh8
28. Rf2 Kg8 29. f4 hxg4 30. hxg4 Be4 31. Re1 Bg6 32. f5 exf5 33. Rh1 fxg4 34.
Rfh2 f5 35. e6 1-0

overlord
Posts: 198
Joined: Sun Jun 03, 2012 4:46 pm
Location: Trinec, Czech Republic

Re: Man vs machine - how to beat chess engine

Post by overlord » Fri Jun 08, 2012 8:00 pm

Nobody is able to play against something better than normal general book. I have free style book and would never get a good result using it. Moreover i7 performance is far higher than old Q6600 or even the best smartphones...it´s performance is similar to old Pentium CPU. Another factor is content - if you put higher value...human chances decreases drastically, but I am seaking about normal setting...nothing specially tuned against human being :)

overlord
Posts: 198
Joined: Sun Jun 03, 2012 4:46 pm
Location: Trinec, Czech Republic

Re: Man vs machine - how to beat chess engine

Post by overlord » Fri Jun 08, 2012 8:03 pm

On my smartphone I have 32 bit version of Critter. I don´t know if it is the same version as for normal pieces. Maybe author could tell more.

Post Reply