Author: Martin Baumung
Date: 13:47:39 04/17/05
Go up one level in this thread
On April 17, 2005 at 14:23:10, John Merlino wrote:
>On April 17, 2005 at 06:44:26, Martin Baumung wrote:
>
>>On April 16, 2005 at 18:38:59, John Merlino wrote:
>>
>>>On April 16, 2005 at 18:16:22, F. Huber wrote:
>>>
>>>>On April 16, 2005 at 17:28:27, John Merlino wrote:
>>>>
>>>>>On April 16, 2005 at 17:15:21, F. Huber wrote:
>>>>>
>>>>>>On April 16, 2005 at 16:24:31, John Merlino wrote:
>>>>>>
>>>>>>>Just for fun, I made a few tiny adjustments to the position, and came up with
>>>>>>>this:
>>>>>>>
>>>>>>>[D]8/8/pp1p4/6p1/2ppK1Pk/3P3P/5PP1/8 w - - 0 1
>>>>>>>
>>>>>>>After playing around with several lines, I'm guessing that this is a Mate for
>>>>>>>White in no less than 20 moves. The King can't find it easily unless I push all
>>>>>>>the way past White's 8th move, and then it says Mate in 14.
>>>>>>>
>>>>>>>White has to spend a lot more time in this position ensuring that he wins the
>>>>>>>pawn race AND avoids stalemate.
>>>>>>>
>>>>>>>What does your awesome machine say? How about ChestUCI, Franz?
>>>>>>
>>>>>>"awesome"? My old Celeron/400? LOL!
>>>>>>
>>>>>>Well, I´ve tried it now (#22) with almost the same parameters as for the
>>>>>>original #13, but had no luck yet. :-(
>>>>>>
>>>>>>This seems to be a lot harder, and maybe the restriction K=1 (i.e. black king
>>>>>>can´t move at all) won´t hold any longer for _this_ position.
>>>>>
>>>>>You're correct, it will no longer hold. To avoid stalemate, White will have to
>>>>>play g3+, allowing Kxh3 and Kxg4. Eventually, there will only be 5-6 men on the
>>>>>board, so tablebases might help.
>>>>
>>>>Stalemate??? Do you want to say that in your solution mate has to capture
>>>>almost all black pawns, even the ones an a/b-line?
>>>>Then your way of solution must really be completely different than the one
>>>>for the original #13, and I´ve absolutely no idea how it might look like.
>>>>OTOH if the black king captures h3/g4-pawns, how should he be mated then -
>>>>at least in ´only´ 22 moves?
>>>>
>>>>Would be interesting to see your solution ...
>>>>
>>>>Franz.
>>>
>>>Indeed, by my playing around, it appears that all Black pawns except for the
>>>g-pawn will definitely be captured. White's d-pawn will queen, and White's h3
>>>and g4 pawns will be captured.
>>>
>>>Here's what I believe to be best play:
>>>
>>>[Event ""]
>>>[Site ""]
>>>[Date "2005.4.16"]
>>>[Round ""]
>>>[White "Merlin"]
>>>[Black "Merlin"]
>>>[TimeControl "-"]
>>>[Result "*"]
>>>[Setup "1"]
>>>[FEN "8/8/pp1p4/6p1/2ppK1Pk/3P3P/5PP1/8 w - - 0 1"]
>>>
>>>1.dxc4 a5 2.Kxd4 a4 3.Kc3 d5 4.cxd5 b5 5.d6 b4+ 6.Kxb4 a3 7.g3+ Kxh3
>>>8.Kxa3 Kxg4 {and now White has a Mate in at most 12 moves with 9.d7, resulting
>>>in a total mate sequence of at most 20 moves -- slightly better than my initial
>>>estimate}
>>>
>>>The other main option for White is:
>>>
>>>6.Kb2 a3+ 7.Kb3 a2 8.Kxa2 b3+ 9.Kb1 b2 {and now 10.f3, 10.f4 or 10.g3+ are
>>>required, and all lead to slower mates than the line above -- The King reports
>>>that 10.f3 is a Mate in 14, for a total length of at most 23 moves}
>>>
>>>jm
>>
>>
>>Hi John,
>>
>>it looks like it's a mate in 19 - maybe even less.
>>
>>After 1.dxc4 a5 ChestUCI finds a mate in 18 within a reasonable amount of time
>>using the parameters P=3 and X=8 (IIRC). It would take too long from the very
>>beginning though...
>>
>>Martin
>
>That's great, Martin. Because any reply by Black other than 1...a5 results in a
>Mate in 12 or less. So we can safely say that the position is no more than a
>Mate in 19.
>
>Many thanks! I wonder who will be the first to report an engine finding it from
>the initial position....?
>
>jm
If you have a lot of time, ChestUCI with the parameters C0/R0/K0/P4/X23 will
find the solution from the initial position. But it could take several hours ;-)
Cheers,
Martin
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.