Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: arasan2 position number 15

Author: Uri Blass

Date: 12:52:57 03/03/04

Go up one level in this thread


On March 03, 2004 at 15:23:14, John Merlino wrote:

>On March 03, 2004 at 13:48:07, Joachim Rang wrote:
>
>>On March 03, 2004 at 13:39:40, John Merlino wrote:
>>
>>>On March 03, 2004 at 10:38:23, Joachim Rang wrote:
>>>
>>>>On March 03, 2004 at 10:14:50, Uri Blass wrote:
>>>>
>>>>>I wonder if the solution f6 is correct solution
>>>>>
>>>>>[D]r3r1k1/6p1/1pqp2Np/2p1nP1P/N1PbPQ2/1P3PR1/P2K4/7R w - - 0 1
>>>>>
>>>>>I could not find a convincing line after f6 Nxc4+ but maybe I miss something.
>>>>>
>>>>>Movei could not solve it at 5 minutes per position on A1000 becasue after few
>>>>>minutes it changed its mind from f6 to Nc3 after failing low at depth 10.
>>>>>
>>>>>Uri
>>>>
>>>>I think you just discovered an error in this testsuite. Most engines fail low
>>>>over  and over after:
>>>>
>>>>1.f6+ Nxc4+ 2. bxc4 Qxa4 3. f7+ Kh7
>>>>
>>>>more (engine)-analysis will follow.
>>>>
>>>>regards Joachim
>>>
>>>I don't think there is a problem. After forcing those moves, Chessmaster 9000
>>>(SKR personality on a P4-2.4) still shows a good score for White. I think
>>>(perhaps inadvertantly) the point is that f6 is the BEST move and that the
>>>engine shouldn't fail TOO low on f6 Nxc4+...? Not sure....
>>>
>>>Perhaps the point is to see that, after the above moves, White must play 4.Rc1
>>>rather than 4.fxe8=Q which leads to a draw.
>>>
>>>I also believe, after a lot of analysis and playing around with lines, that f6
>>>does not necessarily lead to a forced win. I DO believe that it is the best
>>>move, however.
>>>
>>>Here is what The King says:
>>>
>>>Time	Depth	Score	Positions	Moves
>>>27:10	2/14	1.42	387147633	4.Rc1 Qxa2+ 5.Rc2 Qa5+ 6.Ke2 Qa4
>>>					7.Rc1 Rf8 8.Nxf8+ Rxf8 9.Rg6 Qa2+
>>>					10.Qd2 Qa7 11.Qd3 Qxf7 12.Rb1 Qf4
>>>					13.Rxb6
>>>
>>>jm
>>
>>
>>go deeper in the variations. For example after 4.Rc1 Qxa2+ 5.Rc2 Qa5+ 6.Ke2 Qa4
>>7.Rc1 Rf8 8.Nxf8+ Rxf8 9.Rg6
>>9... Qa7! draws.
>>
>>What score does CM9SKR shows after Nc3?
>
>My analysis stopped at 9.Rg6, because CM9 has a difficult time determining the
>best move here. Depending on the time control, it could play Rg6, Qd2 or Kf1.
>All have evals in the neighborhood of +0.5 for White, but the evals were slowly
>dropping. Which, as I said, leads me to believe that f6 does NOT necessarily
>lead to a forced win, but that it is the best move.
>
>After 1.Nc3, CM9SKR says Black wins after 1...Nxc4+ 2.bxc4 Bxc3+ 3.Kxc3:
>
>Time	Depth	Score	Positions	Moves
>0:00	1/4	-0.04	9967		1...b5 2.cxb5 Bxc3+ 3.Kxc3 Qxb5
>					4.Kb2
>0:00	1/5	0.42	45754		1...b5 2.cxb5 Bxc3+ 3.Kxc3 Qxb5
>					4.Nxe5 Qb4+ 5.Kc2 dxe5
>0:00	1/6	0.42	148074		1...b5 2.f6 Bxc3+ 3.Kxc3 b4+ 4.Kd2
>					Rxa2+ 5.Ke3 Qb7 6.fxg7
>0:02	1/7	0.78	434070		1...b5 2.f6 Qc7 3.Nxe5 Bxe5 4.Qxe5
>					dxe5 5.Rxg7+ Qxg7 6.fxg7 bxc4
>0:04	1/7	0.69	743221		1...d5 2.cxd5 Qf6 3.Na4 Kh7 4.Nxe5
>					Bxe5
>0:10	1/8	0.00	2007590		1...d5 2.Nxd5 Rxa2+ 3.Kd1 Ra1+
>					4.Kc2 Ra2+ 5.Kd1
>0:24	1/9	0.07	5005359		1...d5 2.exd5 Qf6 3.Rc1 Bxc3+ 4.Rxc3
>					Nf7 5.a4 Nd6 6.Re3 Qb2+ 7.Kd1
>1:05	1/10	0.00	13307035	1...d5 2.exd5 Qf6 3.Rc1 Nxc4+ 4.bxc4
>					Bxc3+ 5.Rxc3 Rxa2+ 6.Rc2 Qa1 7.Rxa2
>					Qxa2+ 8.Kc3 Qa1+ 9.Kd3 Qb1+ 10.Kd2
>					Qa2+
>3:42	1/11	0.00	46789493	1...d5 2.f6 Bxc3+ 3.Kxc3 d4+ 4.Kc2
>					Rxa2+ 5.Kd1 Ra1+ 6.Kc2 Ra2+
>4:24	1/11	-0.05	55885892	1...Nxc4+ 2.bxc4 Bxc3+ 3.Kxc3 Qa4
>					4.Qc1 Qb4+ 5.Kd3 Ra3+ 6.Ke2 Rc3
>					7.Qd1 Qb2+ 8.Qd2 Rc2 9.Rd1 Rxd2+
>					10.Rxd2
>10:45	1/12	-1.13	138913431	1...Nxc4+ 2.bxc4 Bxc3+ 3.Kxc3 Qa4
>					4.Qc1 Qb4+ 5.Kd3 Ra3+ 6.Ke2 Rc3
>					7.Qe1 Qxc4+ 8.Kf2 Rc2+ 9.Ke3 d5
>					10.Rg4 dxe4 11.fxe4 Rxa2
>18:15	1/13	-1.13	238745890	1...Nxc4+ 2.bxc4 Bxc3+ 3.Kxc3 Qa4
>					4.Qc1 Qb4+ 5.Kd3 Ra3+ 6.Ke2 Rc3
>					7.Qe1 Qxc4+ 8.Kf2 Rc2+ 9.Ke3 d5
>					10.Rg4 dxe4 11.fxe4 Rxa2
>
>So it looks like 1.f6 is definitely best....
>
>jm

You are right that Nc3 is bad and it seems to be a good test position.

Note that movei expected Nxc4+ but did not evaluate it correctly.

yes   9     17430     +215      3458470 f5f6 a8a7 f6g7 a7g7 g6e5 d6e5 g3g7 g8g7
                                        f4g4 g7h8 g4g6 e8e6
yes  10    169630      +95     33621526 f5f6 e5c4 b3c4 c6a4 f6f7 g8h7 h1c1 a4a2
                                        c1c2 a2a5 d2d1 a5a1 c2c1 a1a4 d1e2 a4a2
                                        f4d2 a2d2 e2d2 a8a2 d2d3 a2a3 d3d2
no   10    196160      +96     39194072 a4c3
no   10    216640     +117     43402717 a4c3 e5c4 b3c4 d4c3 d2c3 c6a4 f4c1 a4b4
                                        c3d3 a8a3 d3e2 a3c3 c1b1 b4c4 e2f2 c3c2
                                        f2e3 c4e2 e3f4 e2d2 f4g4 d2g5 g4h3 g5h5
no   10    238400     +117     47912417 a4c3 e5c4 b3c4 d4c3 d2c3 c6a4 f4c1 a4b4
                                        c3d3 a8a3 d3e2 a3c3 c1b1 b4c4 e2f2 c3c2
                                        f2e3 c4e2 e3f4 e2d2 f4g4 d2g5 g4h3 g5h5
no   11    281000      +91     56887247 a4c3 e5c4 b3c4 d4c3 d2c3 c6a4 f4c1 a4b4
                                        c3d3 a8a3 d3e2 a3c3 c1b1 b4c4 e2f2 c3c2
                                        f2e3 c4e2 e3f4 e2d2 f4g4 d2g5 g4h3 g5h5
                                        g6h4

Result:   Failure

It seems that movei missed the quiet move d5 in the following line

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "New game"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "r3r1k1/6p1/1pqp2Np/2p1nP1P/N1PbPQ2/1P3PR1/P2K4/7R w - - 0 1"]
[PlyCount "18"]

1. Nc3 Nxc4+ 2. bxc4 Bxc3+ 3. Kxc3 Qa4 4. Qc1 Qb4+ 5. Kd3 Ra3+ 6. Ke2 Rc3 7.
Qb1 Qxc4+ 8. Kf2 Rc2+ 9. Ke3 d5 *





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.