Computer Chess Club Archives


Search

Terms

Messages

Subject: Bxh2+ probably is the better move... chosen for the right reason?

Author: Dann Corbit

Date: 14:14:21 05/24/00

Go up one level in this thread


On May 24, 2000 at 16:06:47, Dann Corbit wrote:

>On May 24, 2000 at 14:44:13, Gian-Carlo Pascutto wrote:
>>On May 23, 2000 at 20:00:38, James Robertson wrote:
>>>>Comments on Insomniac's play?  Looks quite remarkable.  I don't believe the
>>>>account was restricting play to humans, which would explain such a jump.
>>>
>>>The account has played almost exclusively computers. I'm pretty happy with the
>>>results, and I hope that it is able to maintain or improve its rating. It has
>>>scored very well against several strong computer accounts, although it has
>>>trouble with some of the chess tiger clones.
>>
>>From the games I've got, I suspect Insomniac is very good tactically,
>>coupled with good king-safety.
>>
>>The following position is a nice example. Most amateur programs
>>can take a while to see the sac. Insomniac played it in less than 3 seconds.
>Unfortunately, while the flashy Bxh2+ has real flair and definitely nets a pawn
>in value, the pedantic, stodgy Qg4 puts you up by nearly a piece:
>
>[D] 4r1k1/1b3pp1/pp3n1p/1P2b3/3q4/2NBB3/P4PPP/1QR3K1 b - - acd 12; acn
>151375658; acs 1001; ce 288; pv Qg4 Ne4 Bxe4 Bxe4 Qxe4 bxa6 Qxb1 Rxb1 Bc7 Bd4
>Nd5 Rd1 Be5 a7 Bf4;

Crafty would pick Bxh2+ between 5 seconds and 29 seconds:

pondering disabled.
EGTB cache memory = 6M bytes.
hash table memory = 24M bytes.
pawn hash table memory = 8M bytes.
EGTB access enabled
using tbpath=e:/crafty/release/tb,F:/KBP
6 piece tablebase files found
1943kb of RAM used for TB indices and decompression tables


Crafty v17.10

White(1): st 999
search time set to 999.00.
White(1): epdpfga epd.epd epd.out
PFGA: EPD record: 1
middle-game phase
              clearing hash tables
              time surplus   0.00  time limit 16:39 (16:39)
         nss  depth   time  score   variation (1)
                6->   1.32   0.52   1. ... Qg4 2. Bf1 Ne4 3. Nxe4 Bxe4
                                    4. f3 Bxf3 5. bxa6
                7     3.34   0.73   1. ... Qg4 2. Bf1 Qh5 3. h3 Bxc3 4.
                                    Rxc3 Nd5 5. Rd3 Nxe3 6. Rxe3 Rxe3 7.
                                    fxe3
                7     5.14     ++   1. ... Bxh2+!!
                7     6.38   1.31   1. ... Bxh2+ 2. Kxh2 Ng4+ 3. Kh1 Nxe3
                                    4. fxe3 Rxe3 5. Bf5 Rxc3 6. Rxc3 Qxc3
                                    7. bxa6
                7->   6.54   1.31   1. ... Bxh2+ 2. Kxh2 Ng4+ 3. Kh1 Nxe3
                                    4. fxe3 Rxe3 5. Bf5 Rxc3 6. Rxc3 Qxc3
                                    7. bxa6
                8    17.05   1.39   1. ... Bxh2+ 2. Kxh2 Ng4+ 3. Kh3 Nxe3
                                    4. Bh7+ Kf8 5. fxe3 Rxe3+ 6. g3 Rxc3
                                    7. Rxc3 Qxc3 8. bxa6
                8    29.44     ++   1. ... Qg4!!
                8     1:45   2.91   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Ng4 6. Bxb6 Bxh2+
                                    7. Kf1
                8->   1:45   2.91   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Ng4 6. Bxb6 Bxh2+
                                    7. Kf1
                9     2:44   2.80   1. ... Qg4 2. f3 Bxf3 3. Bf1 Qh5 4.
                                    h3 Bxc3 5. Rxc3 Nd5 6. Rc1 Nxe3 7.
                                    gxf3 Nxf1 8. Rxf1 axb5
                9->   2:51   2.80   1. ... Qg4 2. f3 Bxf3 3. Bf1 Qh5 4.
                                    h3 Bxc3 5. Rxc3 Nd5 6. Rc1 Nxe3 7.
                                    gxf3 Nxf1 8. Rxf1 axb5
               10     5:43   2.90   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Ng4 6. Bxb6 Bxh2+
                                    7. Kh1 Bf4 8. f3
               10->   6:10   2.90   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Ng4 6. Bxb6 Bxh2+
                                    7. Kh1 Bf4 8. f3
               11    12:08   2.88   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Bc7 6. Bd4 Nd5
                                    7. Rd1 Be5 8. a7 Bf4
               11->  14:10   2.88   1. ... Qg4 2. Ne4 Bxe4 3. Bxe4 Qxe4
                                    4. bxa6 Qxb1 5. Rxb1 Bc7 6. Bd4 Nd5
                                    7. Rd1 Be5 8. a7 Bf4
              time=16:39  cpu=99%  mat=0  n=151375658  fh=92%  nps=151362
              ext-> checks=8628038 recaps=412484 pawns=137259 1rep=880483
thrt:43152
              predicted=0  nodes=151375658  evals=23545732
              endgame tablebase-> probes done=0  successful=0


Amy chooses Bxh2+ a 9 seconds and discards it at 25 seconds, only to find it
again at 11 minutes, 17 seconds:

E:\programme\winboard\Amy>amy_07f

    Amy version Winboard 0.7 (DC rev:0.5), Copyright (C) 2000 Thorsten Greiner
    Amy comes with ABSOLUTELY NO WARRANTY; for details type 'warranty'
    This is free software, and you are welcome to redistribute it
    under certain conditions; type 'distribution'

    No multiprocessor support.

$Id: score.c,v 1.44 2000/03/21 15:18:58 thorsten Exp $
$Id: search.c,v 1.57 2000/03/22 10:19:02 thorsten Exp $

Hashtable sizes: 16384 k, 8192 k, 8192 k
Found 4-men endgame table bases.

White(1): level 40/2400
Timecontrol is 40 moves in 2400 mins
White(1): test \epd.epd
Problem 1:
        +---+---+---+---+---+---+---+---+
      8 |   |   |   |   |*R*|   |*K*|   |
        +---+---+---+---+---+---+---+---+
      7 |   |*B*|   |   |   |*P*|*P*|   |   Hashkey: daf7730c
        +---+---+---+---+---+---+---+---+
      6 |*P*|*P*|   |   |   |*N*|   |*P*|
        +---+---+---+---+---+---+---+---+
      5 |   | P |   |   |*B*|   |   |   |   Black (28000, 23000)  PNBRQ
        +---+---+---+---+---+---+---+---+
      4 |   |   |   |*Q*|   |   |   |   |   White (28000, 23000)  PNBRQ
        +---+---+---+---+---+---+---+---+
      3 |   |   | N | B | B |   |   |   |
        +---+---+---+---+---+---+---+---+
      2 | P |   |   |   |   | P | P | P |   Index: 63
        +---+---+---+---+---+---+---+---+
      1 |   | Q | R |   |   |   | K |   |   MateThreat: 0 0
        +---+---+---+---+---+---+---+---+
          a   b   c   d   e   f   g   h
TC: 40 moves in 144000 secs
TL: 3150/12600
GamePhase: Middlegame
It    Time   Score  principal Variation
 6     1.1  +0.096  1. .. Qg4 2. Bf1 Ne4 3. f3 Nxc3 4. Rxc3
 6     1.6  +0.096  1. .. Qg4 2. Bf1 Ne4 3. f3 Nxc3 4. Rxc3
 7     2.8  +0.080  1. .. Qg4 2. Bf1 axb5 3. Qxb5 Bd4 4. Bxd4 Qxd4 5. Bc4
 7     4.0  +0.080  1. .. Qg4 2. Bf1 axb5 3. Qxb5 Bd4 4. Bxd4 Qxd4 5. Bc4
 8     6.1  +0.320  1. .. Qg4 2. Bf1 Qh5 3. h3 Bxc3 4. Rxc3 Nd5 5. Rd3
 8     8.8     +++  1. .. Bxh2+
 8    11.9  +1.088  1. .. Bxh2+ 2. Kxh2 Ng4+ 3. Kh3 Nxe3 4. Bh7+ Kh8 5. fxe3
                    Rxe3+ 6. Kh2 Qe5+ 7. Kg1 Rxc3
 8    12.9  +1.088  1. .. Bxh2+ 2. Kxh2 Ng4+ 3. Kh3 Nxe3 4. Bh7+ Kh8 5. fxe3
                    Rxe3+ 6. Kh2 Qe5+ 7. Kg1 Rxc3
 9    20.4  +1.088  1. .. Bxh2+ 2. Kxh2 Ng4+ 3. Kh3 Nxe3 4. Bh7+ Kh8 5. fxe3
                    Rxe3+ 6. Kh2 Qe5+ 7. Kg1 Rxc3 8. Rxc3
 9    25.2     +++  1. .. Qg4
 9    33.6  +1.824  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh4 4. h3 Bd4 5. Bxd4 Qxd4+
                    6. Kh1
 9    35.6  +1.824  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh4 4. h3 Bd4 5. Bxd4 Qxd4+
                    6. Kh1
10    49.3     +++  1. .. Qg4
10    1:02  +2.352  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh5 4. h3 Bxc3 5. Rxc3 Nd5 6.
                    Rd3
10    1:16  +2.352  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh5 4. h3 Bxc3 5. Rxc3 Nd5 6.
                    Rd3
11    2:16  +2.352  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh5 4. h3 Bxc3 5. Rxc3 Nd5 6.
                    Rd3 Nxe3
11    2:57  +2.352  1. .. Qg4 2. f3 Bxf3 3. Bf1 Qh5 4. h3 Bxc3 5. Rxc3 Nd5 6.
                    Rd3 Nxe3
12    6:19  +2.592  1. .. Qg4 2. f3 Bxf3 3. Qc2 Bxc3
12   11:17     +++  1. .. Bxh2+
12   14:49  +4.048  1. .. Bxh2+ 2. Kxh2 Rxe3 3. fxe3 Qe5+ 4. Kh3 Qxe3+ 5. g3
                    Qf3 6. Nd5 Qxd5 7. Rc6 axb5
12   15:35  +4.048  1. .. Bxh2+ 2. Kxh2 Rxe3 3. fxe3 Qe5+ 4. Kh3 Qxe3+ 5. g3
                    Qf3 6. Nd5 Qxd5 7. Rc6 axb5
13   15:35    1/48  1. .. Bxh2+

Once again, Dusan Dobes shows us that he is a genius:
[ white, 1 ]
4r1k1/1b3pp1/pp3n1p/1P2b3/3q4/2NBB3/P4PPP/1QR3K1 b - -

4r1k1/1b3pp1/pp3n1p/1P2b3/3q4/2NBB3/P4PPP/1QR3K1 b - -
    -> soft time limit 1714.28 s
    -> hard time limit 31999.9 s
  4 ->   0:00.75    27018   0 turns
  5     54    94    35153  Qd4-g4  Bd3-f1  Qg4-h4  Ph2-h3
  5 ->   0:02.31    86337   0 turns
  6     27   321   128755  Qd4-g4  Bd3-f1  Qg4-h4  Ph2-h3  Pa6xb5  Qb1xb5
                           Re8-e6
  6     32   362   147293  Be5xh2  !
  6     87   413   170479  Be5xh2  !!
  6    146   672   284835  Be5xh2  Kg1xh2  Nf6-g4  Kh2-h3  Ng4xe3
  6 ->   0:08.15   354026   1 turn
  7    206  1132   491688  Be5xh2  !!
  7    256  2960  1304025  Be5xh2  Kg1xh2  Re8xe3  Pf2xe3
  7 ->   0:36.73  1681106   0 turns
  8    316  5075  2287377  Be5xh2  !!
  8    387 11793  5374407  Be5xh2  Kg1-f1  Re8xe3  Pf2xe3  Qd4xe3  Nc3-e2
                           Nf6-g4  Bd3-h7  Kg8-h8  Qb1-f5  Qe3xc1  Ne2xc1
                           Ng4-e3  Kf1-e2  Ne3xf5  Bh7xf5  Pa6xb5
  8 ->   2:13.31  6081022   0 turns
  9    394 31699 14132932  Be5xh2  Kg1xh2  Re8xe3  Bd3-h7  Kg8-h8  Pf2xe3
                           Qd4-e5  Kh2-h3  Qe5xe3  Pg2-g3

Why is his eval ten times better than everyone else?  He cleary finds the best
line in far fewer plies consistently.

I am awestruck by the eval of Phalanx.



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.