Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: endgame blunder by DeepFritz7

Author: Robert Hyatt

Date: 22:01:12 11/24/02

Go up one level in this thread


On November 24, 2002 at 13:22:46, Terry McCracken wrote:

>On November 24, 2002 at 10:23:38, Robert Hyatt wrote:
>
>>On November 23, 2002 at 19:16:23, Jay Hysenbeg wrote:
>>
>>>On November 23, 2002 at 08:17:02, Terry McCracken wrote:
>>>
>>>>On November 23, 2002 at 00:32:08, Robert Hyatt wrote:
>>>>
>>>>>On November 23, 2002 at 00:13:32, Terry McCracken wrote:
>>>>>
>>>>>>On November 22, 2002 at 23:45:01, Jay Hysenbeg wrote:
>>>>>>
>>>>>>>[D]8/4p3/4Q3/8/8/k1p3PP/1q3P1K/8 w - - 0 1
>>>>>>>wKh2,Qe6,Pf2,g3,h3/bKa3,Qb2,Pc3,e7
>>>>>>>
>>>>>>>is this a problem with my deepfritz7 or does this happen with your df7 too. mine
>>>>>>>fails to see that Qxe7 is a loosing line
>>>>>
>>>>>
>>>>>Is it certain Qe7 loses?  All I can see is a long series of checks and I don't
>>>>>see how black can find a safe haven from them to promote the pawn...
>>>>>
>>>>>I let Crafty run a while and it says 0.00...
>>>>
>>>>
>>>>Bob White can't lose, I just inserted the [D] to look at the position.
>>>>
>>>>It looked totally uninteresting.
>>>>
>>>>Terry
>>>chesstiger 15/14 and gambit tiger both see that Qxe7 is losing for white score
>>>jumps by 5 pawns
>>
>>
>>Would you post the "losing" line?  I don't see how black can ever escape the
>>checks.
>
>Bob I looked at it again and it appears I really underestimated the position!
>
>After 1.Qxe7+?? ...Ka2!! going to b1 and it really goes down hill for White
>quickly!
>
>I should have seen this...arrghh! getting rusty...."creak"....sigh...
>
>Terry
>
>P.S. 1...Kb3 is ok I think but Ka2!! and White will have to resign, Black will
>get out of check and promote his pawn.
>
>Try it yourself, you'll see White starts to lose ground for grabbing the pawn.
>
>It's a tricky endgame and Fritz 7 is slower than Tiger to spot the loss.
>HIARCS 8, Junior 7, Shredder 6 all see a lost position but not straight away.
>
>Only Tiger spots it instantly! Bravo Christophe!
>
>
>Juni


I guess I am way beyond rusty, because I +still+ don't see it.  :)

Qxe7 Ka2 (your move)
Qe6+

Now what?  Kb1?  If so, I'd probalby try Qe4+ and expect c2.   Now I play
f4 and wait for black to try to unpin the pawn.  Ka1 Qa4 Kb1 and we dance
again.

How does black make progress here?

Note that this is human analysis, not computer analysis.  I let crafty crank
on it a while yesterday or whenever and it simply said 0.00 very quickly and
didn't change.

I'm not sure the above is very accurate.  I didn't dwell on this for many
minutes, so I probably overlooked something simple, but I've generally found
that with a queen, a queen and pawn is usually not enough of a shelter to stop
the forever checks.

I was trying to create a position where

(1) black controls the b1-c2-d3-etc diagonal so that white can't pin the
pawn there;

(2) the black king would then be at b1, and somehow checks along the b-file
are impossible so that a check on the 8th rank could be met by c1=Q blocking
the check;

(3) I don't see how to do both, although from your comments apparently it is
possible or else white can be forced to a queen location where the pin along
the diagonal can't be played?




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.