Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

Main

Date: 15 Oct 2004 17:24:54
From: Don't Panic
Subject: Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

Weirdest thing I have ever seen in computer chess or just normal chess.

3r4/1k3P2/2n1qPP1/p3P3/1p5P/2B3R1/6BK/8 w - - 0 45

This position arose after moves of a game between Fritz 8 and Shredder 8. I
was watching the game and could not believe that white lost.

Black has just played 44...Bb2

Fritz evaluated the position as + 4.47 suggesting the move 45.Be1
Shredder meanwhile evaluated the position as = -0.55after the move 45.Bb2

In the event black (Shredder won), but even in infinite analysis Fritz and
Shredder are both convinced white is winning.

Put this position into your own machines and see what your eyes and your
engines think, I would think that white is winning, but can't see how... (of
course the mistake may have happened before this - becasue even Shredder a
few moves before thought that it was getting cufffed)

Here's the game:
(19) Fritz 8 - Shredder 8 [C18]
4+2 shredder8 v Fritz8 , Blitz:4'+2" NOTEBOOKK (8), 15.10.2004
[1.97;2.56]

W=8.1 ply; 460kN/s B=10.8 ply; 194kN/s 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5
5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 Qc7 8.Qxg7 Rg8 9.Qxh7 cxd4 10.Ne2 Nbc6 11.f4 Bd7
12.Qd3 dxc3 13.Qxc3 Nf5 14.Rb1 d4 15.Qd3 0-0-0 16.Rg1 f6 17.g4 Nh4 18.exf6
e5 19.h3 e4 20.Qxe4 d3 21.cxd3 Rge8 Both last book move 22.Qc4 1.97/11 16
Nf3+ (Be6) 2.56/13 18 23.Kf2 2.18/12 18 Nxg1 (Be6) 2.53/14 16 24.Kxg1
2.18/10 3 Be6 2.77/14 26 25.Qc3 (Qc5) 2.18/10 14 25...Bd5 2.47/11 10 26.g5
(Ng3) 2.34/10 7 26...Be6 (Bf7) 2.78/12 20 27.Ng3 (d4) 2.37/9 7 27...Qh7
3.14/12 33 28.Rb5 (Qb2) 2.50/10 10 28...Bd5 (a6) 3.42/12 48 29.f5 (Kh2)
3.06/11 14 29...a6 (Qh4) 3.94/11 22 30.Rb4 3.53/11 11 a5 (Kb8) 4.16/11 6
31.Rg4 (Rf4) 4.12/10 5 31...Qh8 (Bf3) 4.47/11 8 32.Bb2 (g6) 4.25/10 5
32...Qf8 4.57/12 13 33.g6 4.59/10 4 Qd6 5.07/11 14 34.f7 (Ne4) 4.84/10 8
34...Re3 4.83/11 4 35.Ne4 (Kh2) 5.22/8 1 35...Bxe4 3.33/11 19 36.dxe4 (g7)
5.75/8 2 36...Rxc3 1.32/11 3 37.Bxc3 3.87/9 2 Qc5+ 0.69/9 2 38.Kg2 (Kh1)
3.62/8 1 38...Qxa3 (Qxc3) 0.13/10 3 39.Rg3 (Kh1) 3.56/7 2 39...Qe7 0.00/9 1
40.Kh2 (Rg4) 4.18/7 2 40...b6 1.68/9 6 41.f6 (e5) 5.34/8 2 41...Qe6 0.97/10
2 42.h4 (Rd3) 5.31/7 2 42...Kb7 (Kc7) 0.64/10 6 43.Bg2 (Bh3) 4.40/7 2
43...b5 (Rf8) 0.54/10 5 44.e5 (Rd3) 5.09/8 2 44...b4 -0.02/10 2 45.Bd4 (Rd3)
4.00/7 3 45...Kc7 -0.99/10 4 46.Ba1 (Bb2) 3.53/6 1 46...Qc4 -1.99/10 5
47.Bxc6 (h5) 0.00/7 3 47...Qxh4+ (Kxc6) -4.96/9 2 48.Kg2 (Rh3) -1.35/6 1
48...Kxc6 -10.65/9 3 49.f8Q (Bd4) -3.44/6 5 49...Rd2+ (Qe4+) -#7/13 0
50.Kf3 -#6/5 0 Rd3+ -#6/11 0 51.Ke2 -#5/4 0 Qe4+ -#5/9 0 52.Re3 -#4/4 0
Qxe3+ -#4/7 0 53.Kf1 -#3/3 0 Rd1+ -#3/5 0 54.Kg2 -#2/3 0 Rg1+ -#2/3 0
55.Kh2 -#1/2 0 Qg3# -#1/1 0 0-1

Date: 15 Oct 2004 21:24:58
From: Michael W�sch
Subject: Re: Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

Hi,

Analysis by Hiarcs 9:

1. = (0.00): 45.Ld2 Kb6 46.Le3+ Kc7 47.Lf4 Kb6
2. = (-0.01): 45.Ld4 Kc7 46.Lh3 Da2+ 47.Tg2 Dd5 48.g7 Sxd4 49.f8D Dxe5+
50.Kh1 De1+ 51.Tg1
3. -+ (-1.52): 45.Lb2 Dc4 46.Kh3 a4 47.h5 a3 48.Lxa3 De6+ 49.Kh4 bxa3
4. -+ (-1.58): 45.La1 Dc4 46.Kh3 a4 47.Tg4 Td3+ 48.Tg3 De6+ 49.Kh2 Td8
50.Te3

Analysis by Deep Fritz 8:

1. +- (3.85): 45.Le1 Kc7 46.Te3 a4 47.Lg3 Kb6 48.Lf2 Kc7 49.Lh3
2. +- (2.09): 45.Ld2 Ka6 46.Lf1+ Kb7 47.Lf4 b3 48.Td3 b2 49.Txd8 Sxd8 50.f8D
3. +- (2.13): 45.Ld4 Dc4 46.Le3 Dxh4+ 47.Th3 Dc4 48.Th7 Kb8 49.g7
4. � (1.30): 45.g7 Dxf7 46.Ld2 Kc7 47.Lf4 Kb7 48.Kg1 a4

Analysis by Shredder 8:

1. = (-0.02): 45.Td3 bxc3 46.Txd8 Dxe5+ 47.Kh3 Df5+ 48.Kh2 De5+ 49.Kh3 Df5+
50.Kh2 De5+ 51.Kh3
2. = (-0.02): 45.Ld2 Dc4 46.Lh6 Dxh4+ 47.Th3 De1 48.Te3 Dh4+ 49.Th3 De1
50.Te3 Dh4+ 51.Th3 De1
3. = (-0.04): 45.Ld4 Dc4 46.Tg4 Txd4 47.Txd4 Dxd4 48.f8D Dxh4+ 49.Lh3
4. � (-0.53): 45.Lb2 Dc4 46.h5 a4 47.e6 Dxe6 48.h6 a3 49.g7 axb2 50.f8D

Analysis by Crafty 19.14:

1. = (0.00): 45.Le1 Dc4 46.Te3 De6 47.Lh3 Da2+ 48.Lg2 De6
2. = (0.00): 45.Ld4 Dc4 46.Tg4 De6 47.Tg3
3. = (0.00): 45.Ld2 Dc4 46.h5 De2 47.h6 Dxd2 48.g7 Dxh6+ 49.Th3 Df4+ 50.Tg3
Dh4+ 51.Th3 Df4+
4. � (-0.95): 45.La1 Dc4 46.Kh3 Kb6 47.Te3 a4 48.Te4 De6+ 49.Kg3 a3 50.h5

Analysis by Fritz 7:

1. = (0.00): 45.Ld4 Dc4 46.Tg4 De6 47.Tg3 Dc4 48.Tg4 De6 49.Tg3 Dc4 50.Tg4
De6 51.Tg3 Dc4 52.Tg4
2. = (0.00): 45.Ld2 Dc4 46.Lg5 b3 47.Te3 De6 48.Lh3 Dd5
3. = (0.00): 45.Le1 Kc7 46.Te3 Tb8 47.Lg3 Kb6 48.Lh3 Da2+ 49.Lg2
4. � (-1.12): 45.Td3 Txd3 46.f8D Kb6 47.f7 Dg4 48.Dc5+ Kxc5

Mike

Date: 15 Oct 2004 21:36:34
From: Guy Macon
Subject: Re: Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

Looking only at the next move and the top three choices
(I will explain why in a minute)...

Michael W�sch <[email protected] > says...

>Analysis by Crafty 19.14:
>
>1. = ( 0.00): 45.Le1
>2. = ( 0.00): 45.Ld4
>3. = ( 0.00): 45.Ld2
>
>Analysis by Fritz 7:
>
>1. = ( 0.00): 45.Ld4
>2. = ( 0.00): 45.Ld2
>3. = ( 0.00): 45.Le1

Crafty and Fritz 7 much agree on the next move.
(when they are all 0.00, order means little)

>Analysis by Deep Fritz 8:
>
>1. +- (3.85): 45.Le1
>2. +- (2.09): 45.Ld2
>3. +- (2.13): 45.Ld4

Deep Fritz 8 has the same three choices, but is far more
opinionated about the score for each.

>Analysis by Shredder 8:
>
>1. = (-0.02): 45.Td3
>2. = (-0.02): 45.Ld2
>3. = (-0.04): 45.Ld4

Shredder 8 agrees on two out of the three.

>Analysis by Hiarcs 9:
>
>1. = ( 0.00): 45.Ld2
>3. -+ (-1.52): 45.Lb2
>4. -+ (-1.58): 45.La1

Hiarcs 9 agrees on one of the three.

Let's say you were designing a multi-engine voting chess
program. What move would you choose given the above? Ld2?
Would looking further down the lines held it to choose?

Date: 15 Oct 2004 18:24:42
From: Robert Hyatt
Subject: Re: Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

In rec.games.chess.computer Don't Panic <[email protected] > wrote:

> Weirdest thing I have ever seen in computer chess or just normal chess.

> 3r4/1k3P2/2n1qPP1/p3P3/1p5P/2B3R1/6BK/8 w - - 0 45

> This position arose after moves of a game between Fritz 8 and Shredder 8. I
> was watching the game and could not believe that white lost.

> Black has just played 44...Bb2

> Fritz evaluated the position as + 4.47 suggesting the move 45.Be1
> Shredder meanwhile evaluated the position as = -0.55after the move 45.Bb2

> In the event black (Shredder won), but even in infinite analysis Fritz and
> Shredder are both convinced white is winning.

I am not sure why you have the "crafty wrong" in your header. here is output
from Crafty:

8 1.19 -0.66 1. Bd4 Kc7 2. Bc5 Nxe5 3. f8=Q Rxf8
4. Bxf8 Qxf6 (s=3)
8 2.67 -0.49 1. Bd2 Kc7 2. Bf4 Kb6 3. Rd3 Rxd3 4.
f8=Q Nxe5 5. Bxe5 Qxe5+ (s=2)
8- > 3.40 -0.49 1. Bd2 Kc7 2. Bf4 Kb6 3. Rd3 Rxd3 4.
f8=Q Nxe5 5. Bxe5 Qxe5+
9 7.00 -0.47 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
e6 Qxe6 5. f7 Qxf7 6. Bxd8
9- > 9.36 -0.47 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
e6 Qxe6 5. f7 Qxf7 6. Bxd8
10 16.90 -0.32 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
e6 Qxe6 5. f7 Qxf7 6. Bxd8 Qg8 7. Bxc6+
Kxc6 8. Bxa5
10- > 23.75 -0.32 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
e6 Qxe6 5. f7 Qxf7 6. Bxd8 Qg8 7. Bxc6+
Kxc6 8. Bxa5
11 54.19 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
11- > 1:08 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
12 2:18 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
12- > 2:49 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2

IE it doesn't seem to think that white is better. Note that - scores are
always good for black when using crafty, no matter whether it is playing
black or white. + scores are always good for white.

Crafty gives this as "dead even" (draw score) after 12 plies above.

> Put this position into your own machines and see what your eyes and your
> engines think, I would think that white is winning, but can't see how... (of
> course the mistake may have happened before this - becasue even Shredder a
> few moves before thought that it was getting cufffed)

> Here's the game:
> (19) Fritz 8 - Shredder 8 [C18]
> 4+2 shredder8 v Fritz8 , Blitz:4'+2" NOTEBOOKK (8), 15.10.2004
> [1.97;2.56]

> W=8.1 ply; 460kN/s B=10.8 ply; 194kN/s 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5
> 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 Qc7 8.Qxg7 Rg8 9.Qxh7 cxd4 10.Ne2 Nbc6 11.f4 Bd7
> 12.Qd3 dxc3 13.Qxc3 Nf5 14.Rb1 d4 15.Qd3 0-0-0 16.Rg1 f6 17.g4 Nh4 18.exf6
> e5 19.h3 e4 20.Qxe4 d3 21.cxd3 Rge8 Both last book move 22.Qc4 1.97/11 16
> Nf3+ (Be6) 2.56/13 18 23.Kf2 2.18/12 18 Nxg1 (Be6) 2.53/14 16 24.Kxg1
> 2.18/10 3 Be6 2.77/14 26 25.Qc3 (Qc5) 2.18/10 14 25...Bd5 2.47/11 10 26.g5
> (Ng3) 2.34/10 7 26...Be6 (Bf7) 2.78/12 20 27.Ng3 (d4) 2.37/9 7 27...Qh7
> 3.14/12 33 28.Rb5 (Qb2) 2.50/10 10 28...Bd5 (a6) 3.42/12 48 29.f5 (Kh2)
> 3.06/11 14 29...a6 (Qh4) 3.94/11 22 30.Rb4 3.53/11 11 a5 (Kb8) 4.16/11 6
> 31.Rg4 (Rf4) 4.12/10 5 31...Qh8 (Bf3) 4.47/11 8 32.Bb2 (g6) 4.25/10 5
> 32...Qf8 4.57/12 13 33.g6 4.59/10 4 Qd6 5.07/11 14 34.f7 (Ne4) 4.84/10 8
> 34...Re3 4.83/11 4 35.Ne4 (Kh2) 5.22/8 1 35...Bxe4 3.33/11 19 36.dxe4 (g7)
> 5.75/8 2 36...Rxc3 1.32/11 3 37.Bxc3 3.87/9 2 Qc5+ 0.69/9 2 38.Kg2 (Kh1)
> 3.62/8 1 38...Qxa3 (Qxc3) 0.13/10 3 39.Rg3 (Kh1) 3.56/7 2 39...Qe7 0.00/9 1
> 40.Kh2 (Rg4) 4.18/7 2 40...b6 1.68/9 6 41.f6 (e5) 5.34/8 2 41...Qe6 0.97/10
> 2 42.h4 (Rd3) 5.31/7 2 42...Kb7 (Kc7) 0.64/10 6 43.Bg2 (Bh3) 4.40/7 2
> 43...b5 (Rf8) 0.54/10 5 44.e5 (Rd3) 5.09/8 2 44...b4 -0.02/10 2 45.Bd4 (Rd3)
> 4.00/7 3 45...Kc7 -0.99/10 4 46.Ba1 (Bb2) 3.53/6 1 46...Qc4 -1.99/10 5
> 47.Bxc6 (h5) 0.00/7 3 47...Qxh4+ (Kxc6) -4.96/9 2 48.Kg2 (Rh3) -1.35/6 1
> 48...Kxc6 -10.65/9 3 49.f8Q (Bd4) -3.44/6 5 49...Rd2+ (Qe4+) -#7/13 0
> 50.Kf3 -#6/5 0 Rd3+ -#6/11 0 51.Ke2 -#5/4 0 Qe4+ -#5/9 0 52.Re3 -#4/4 0
> Qxe3+ -#4/7 0 53.Kf1 -#3/3 0 Rd1+ -#3/5 0 54.Kg2 -#2/3 0 Rg1+ -#2/3 0
> 55.Kh2 -#1/2 0 Qg3# -#1/1 0 0-1

--
Robert M. Hyatt, Ph.D. Computer and Information Sciences
[email protected] University of Alabama at Birmingham
(205) 934-2213 136A Campbell Hall
(205) 934-5473 FAX Birmingham, AL 35294-1170

Date: 16 Oct 2004 10:10:57
From: Don't Panic
Subject: Re: Fritz wrong Crafty wrong Shredder right? Human lost in a maze.... Interesting.........

"Robert Hyatt" <[email protected] > wrote in message
news:[email protected]...

> I am not sure why you have the "crafty wrong" in your header. here is
> output
> from Crafty:

I am no computer buff, I am a chess player, but the readout from my limited
hardware using Crafty I thought this indicated that 'it' thought white was
winning in that position - perhaps I did not leave the position long enough.
No offence intended, honest. We love Crafty. It's great. And correct.
Probably always. Vive La Crafty!

I would be interested in knowing - if in the game white was ever winning,
becasue even Shredder a few moves before gave a huge plus to white.

>
> 8 1.19 -0.66 1. Bd4 Kc7 2. Bc5 Nxe5 3. f8=Q Rxf8
> 4. Bxf8 Qxf6 (s=3)
> 8 2.67 -0.49 1. Bd2 Kc7 2. Bf4 Kb6 3. Rd3 Rxd3 4.
> f8=Q Nxe5 5. Bxe5 Qxe5+ (s=2)
> 8-> 3.40 -0.49 1. Bd2 Kc7 2. Bf4 Kb6 3. Rd3 Rxd3 4.
> f8=Q Nxe5 5. Bxe5 Qxe5+
> 9 7.00 -0.47 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
> e6 Qxe6 5. f7 Qxf7 6. Bxd8
> 9-> 9.36 -0.47 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
> e6 Qxe6 5. f7 Qxf7 6. Bxd8
> 10 16.90 -0.32 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
> e6 Qxe6 5. f7 Qxf7 6. Bxd8 Qg8 7. Bxc6+
> Kxc6 8. Bxa5
> 10-> 23.75 -0.32 1. Bd2 Qc4 2. Bg5 b3 3. g7 Qxf7 4.
> e6 Qxe6 5. f7 Qxf7 6. Bxd8 Qg8 7. Bxc6+
> Kxc6 8. Bxa5
> 11 54.19 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
> 11-> 1:08 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
> 12 2:18 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
> 12-> 2:49 0.01 1. Bd2 Ka6 2. Bf1+ Kb7 3. Bg2
>
>
>
> IE it doesn't seem to think that white is better. Note that - scores are
> always good for black when using crafty, no matter whether it is playing
> black or white. + scores are always good for white.
>
> Crafty gives this as "dead even" (draw score) after 12 plies above.
>
>> Put this position into your own machines and see what your eyes and your
>> engines think, I would think that white is winning, but can't see how...
>> (of
>> course the mistake may have happened before this - becasue even Shredder
>> a
>> few moves before thought that it was getting cufffed)
>
>> Here's the game:
>> (19) Fritz 8 - Shredder 8 [C18]
>> 4+2 shredder8 v Fritz8 , Blitz:4'+2" NOTEBOOKK (8), 15.10.2004
>> [1.97;2.56]
>
>
>> W=8.1 ply; 460kN/s B=10.8 ply; 194kN/s 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5
>> 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 Qc7 8.Qxg7 Rg8 9.Qxh7 cxd4 10.Ne2 Nbc6 11.f4
>> Bd7
>> 12.Qd3 dxc3 13.Qxc3 Nf5 14.Rb1 d4 15.Qd3 0-0-0 16.Rg1 f6 17.g4 Nh4
>> 18.exf6
>> e5 19.h3 e4 20.Qxe4 d3 21.cxd3 Rge8 Both last book move 22.Qc4 1.97/11 16
>> Nf3+ (Be6) 2.56/13 18 23.Kf2 2.18/12 18 Nxg1 (Be6) 2.53/14 16 24.Kxg1
>> 2.18/10 3 Be6 2.77/14 26 25.Qc3 (Qc5) 2.18/10 14 25...Bd5 2.47/11 10
>> 26.g5
>> (Ng3) 2.34/10 7 26...Be6 (Bf7) 2.78/12 20 27.Ng3 (d4) 2.37/9 7 27...Qh7
>> 3.14/12 33 28.Rb5 (Qb2) 2.50/10 10 28...Bd5 (a6) 3.42/12 48 29.f5 (Kh2)
>> 3.06/11 14 29...a6 (Qh4) 3.94/11 22 30.Rb4 3.53/11 11 a5 (Kb8) 4.16/11 6
>> 31.Rg4 (Rf4) 4.12/10 5 31...Qh8 (Bf3) 4.47/11 8 32.Bb2 (g6) 4.25/10 5
>> 32...Qf8 4.57/12 13 33.g6 4.59/10 4 Qd6 5.07/11 14 34.f7 (Ne4) 4.84/10 8
>> 34...Re3 4.83/11 4 35.Ne4 (Kh2) 5.22/8 1 35...Bxe4 3.33/11 19 36.dxe4
>> (g7)
>> 5.75/8 2 36...Rxc3 1.32/11 3 37.Bxc3 3.87/9 2 Qc5+ 0.69/9 2 38.Kg2 (Kh1)
>> 3.62/8 1 38...Qxa3 (Qxc3) 0.13/10 3 39.Rg3 (Kh1) 3.56/7 2 39...Qe7 0.00/9
>> 1
>> 40.Kh2 (Rg4) 4.18/7 2 40...b6 1.68/9 6 41.f6 (e5) 5.34/8 2 41...Qe6
>> 0.97/10
>> 2 42.h4 (Rd3) 5.31/7 2 42...Kb7 (Kc7) 0.64/10 6 43.Bg2 (Bh3) 4.40/7 2
>> 43...b5 (Rf8) 0.54/10 5 44.e5 (Rd3) 5.09/8 2 44...b4 -0.02/10 2 45.Bd4
>> (Rd3)
>> 4.00/7 3 45...Kc7 -0.99/10 4 46.Ba1 (Bb2) 3.53/6 1 46...Qc4 -1.99/10 5
>> 47.Bxc6 (h5) 0.00/7 3 47...Qxh4+ (Kxc6) -4.96/9 2 48.Kg2 (Rh3) -1.35/6 1
>> 48...Kxc6 -10.65/9 3 49.f8Q (Bd4) -3.44/6 5 49...Rd2+ (Qe4+) -#7/13 0
>> 50.Kf3 -#6/5 0 Rd3+ -#6/11 0 51.Ke2 -#5/4 0 Qe4+ -#5/9 0 52.Re3 -#4/4 0
>> Qxe3+ -#4/7 0 53.Kf1 -#3/3 0 Rd1+ -#3/5 0 54.Kg2 -#2/3 0 Rg1+ -#2/3 0
>> 55.Kh2 -#1/2 0 Qg3# -#1/1 0 0-1
>
>
>
> --
> Robert M. Hyatt, Ph.D. Computer and Information Sciences
> [email protected] University of Alabama at Birmingham
> (205) 934-2213 136A Campbell Hall
> (205) 934-5473 FAX Birmingham, AL 35294-1170