Author: David Dahlem
Date: 07:00:48 01/01/06
Go up one level in this thread
On January 01, 2006 at 07:38:42, Sune Larsson wrote:
>On January 01, 2006 at 07:14:42, Joseph Ciarrochi wrote:
>
>>>
>>>It is known that part of the positions are unbalanced so I do not understand why
>>>do you even try to do comprison between rating and results.
>>>
>>>Note also that even in case that the positions are balanced you can expect
>>>smaller difference relative to difference in real games because in most cases
>>>the weaker engine does not get equal endgames in real games.
>>>
>>>Uri
>>
>>Yes, I realize this. I suppose rather than saying "its amazing" I should say " I
>>was amazed, or impressed by yace." I thought it was cool that a much older
>>engine could still be competitive.
>>
>>The only way to tell what constitutes a "big" or "small" difference would be to
>>run a massive endgame tournament. Then we could get some stats on whether, say,
>>a +7 score out of 40 is statistically unusual, etc.
>
>
> Yup - and thanks for the games! I wonder if there is any documentation about
> the Nunn endings. I haven't studied them much myself. Anyway - pos 1 is a draw
> with best play. The same for n:o 3 (in my suite), n:o 4 I think, and n:o 7.
> Maybe all are drawn with best play?
>
> /S
Hi Sune
I'm not sure where i got this, i think it was from the Chessbase site, but the
link is no longer there. Here are the endgames in pgn format with descriptions
of the positions ...
[Event "1/99-60 John Nunn"]
[Site "?"]
[Date "1998.??.??"]
[Round "?"]
[White "NET01: Rook Ending"]
[Black "[+0400.66f2f7]"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/1ppr1kp1/p1p4p/8/8/5P2/PPP1RKPP/8 w - - 0 1"]
[PlyCount "0"]
{A position which might have arisen from the Exchange Variation of the Spanish.
White's better pawn structure gives him some advantage, since he can create a
passed pawn on the kingside while Black, owing to his doubled pawns, cannot do
the same on the queenside. It is not clear if this advantage is sufficient to
win. In practical play I would expect wins and draws to be about equally
divided.} *
[Event "1/99-60 Budapest"]
[Site "Budapest"]
[Date "1952.??.??"]
[Round "?"]
[White "NET02: Barcza"]
[Black "Golombek"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/1pk3pp/p7/3p1p2/8/3K2P1/PP2PP1P/8 w - - 0 1"]
[PlyCount "0"]
[EventDate "1952.??.??"]
{This endgame certainly favours White. Black's pawn structure is inferior
because of the isolated d-pawn, while White's king is the first to occupy the
fourth rank. White has the obvious plan of putting his king on d4 followed by
using zugzwang to penetrate with his king to c5 or e5. This was in fact the
plan adopted by Barcza in the game, but accurate defence by Golombek enabled
him to hold the draw. Later (in 1966) Bondarevsky analysed the ending and
concluded that White could win with a different and far from obvious plan:
playing Ke3, threatening Kf4 and inducing Black to weaken his kingside pawn
structure further by playing ...g5. It is hard to say whether this is correct,
but at any rate it is clear that this position is finely poised between a draw
and a win.} *
[Event "1/99-60 John Nunn"]
[Site "?"]
[Date "1998.??.??"]
[Round "?"]
[White "NET03: Database position"]
[Black "[+4000.10b2f4]"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "2q5/8/6Q1/8/1P3k2/8/1K6/8 w - - 0 4"]
[PlyCount "0"]
{Including a database position is perhaps a little controversial, but why not?
A program which uses the 5-man databases will play the endgame more strongly
than one which does not, so this should be a part of the Nunn-test The
database shows that this position is a win, but only by 1 Qd6+!. If neither
side has access to a database, it is harder to predict the result; between
human players I would expect White to win more often than not.} *
[Event "1/99-60 Telegraph match"]
[Site "?"]
[Date "1896.??.??"]
[Round "?"]
[White "NET04: St Petersburg"]
[Black "London"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/4k3/r4pp1/pR6/P4KP1/5P2/8/8 b - - 0 1"]
[PlyCount "0"]
[EventDate "1896.??.??"]
{A classic ending. The game was halted at this point with London conceding the
game. Chigorin then published some analysis claiming that White could win. For
decades this was the accepted opinion, but theory advances and it gradually
became recognised that defending such position often requires the sacrifice of
the weak pawn (on a5 here). Fine, in his book Basic Chess Endings, suggested
that this plan should lead to a draw. His analysis contained some errors, and
was refined by Levenfish and Smyslov in their classic book Rook Endings. They
supported Fine's conclusion, although some of the detailed variations differed
from Fine's. That is the state of play today. Can computers unravel a position
which took human analysts over half a century to unravel?.} *
[Event "1/99-60 Dresden"]
[Site "Dresden"]
[Date "1936.??.??"]
[Round "?"]
[White "NET05: Keres"]
[Black "Alekhine"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/p1p1k3/1p1p2p1/3Pq2p/2P1P3/1P3QPK/P7/8 w - - 0 1"]
[PlyCount "0"]
[EventDate "1936.??.??"]
{In this game between two of the leading players of the pre-war period, Black
has the advantage because he can create a kingside passed pawn while White,
thanks to his backward e-pawn, cannot create a passed pawn. However, in a
queen ending matters are never so easy, due to the ever-present danger of
perpetual check. In the game Black did succeed in winning after a hard fight,
although Konstantinopolsky later suggested an improvement for White which
might have placed Black's win in doubt.} *
[Event "1/99-60 Uzhgorod"]
[Site "?"]
[Date "1972.??.??"]
[Round "?"]
[White "NET06: Holmov"]
[Black "Moiseev"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/p3ppk1/1p4pp/3nN3/6PP/1P3P2/P3PK2/8 w - - 0 1"]
[PlyCount "0"]
[EventDate "1972.??.??"]
{A tricky knight ending. White won the game, but the winner's very detailed
analysis (covering several pages) suggested that the correct result should be
a draw.} *
[Event "1/99-60 Cetinje"]
[Site "Cetinje"]
[Date "1992.??.??"]
[Round "?"]
[White "NET07: Ilincic"]
[Black "Cabrilo"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "7k/pp4pp/2n5/8/8/P7/1P4PP/2K1B3 b - - 0 25"]
[PlyCount "0"]
[EventDate "1992.??.??"]
{Although this position appears fairly drawish White won the game, and in his
notes indicated no real improvement for Black. However, I analysed this ending
in my book Secrets of Practical Chess, and came to the conclusion that it
should probably be a draw. What do the machines think?} *
[Event "1/99-60 Siegen"]
[Site "Siegen"]
[Date "1970.??.??"]
[Round "?"]
[White "NET08: Bellon"]
[Black "Minic"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/8/p4Bp1/1pPb2P1/1P2kp2/P7/5K2/8 w - - 0 1"]
[PlyCount "0"]
[EventDate "1970.??.??"]
{A fantastic position. An opposite-coloured bishop position with equal pawns
looks as if it should be a draw, and indeed the players agreed to a draw at
this point. However, White a serious problems because of his weak a3-pawn,
which cannot be permanently defended by his bishop. Once this pawn has been
captured, Black has the chance to make a second passed pawn by playing ...a5,
which will stretch White's defence to the utmost. Dvoretsky and Yusupov
subjected this ending to detailed analysis, and came to the astonishing
conclusion that it is a draw, but only if White starts with the incredible 1
c6!!. It is hard to imagine that any machine would find such a move, but who
knows what might happen in the future...} *
[Event "1/99-60 Semmering-Baden"]
[Site "Semmering"]
[Date "1937.??.??"]
[Round "?"]
[White "NET09: Eliskases"]
[Black "Capablanca"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "8/4k3/p1B4p/2K5/1P4bP/8/8/8 b - - 0 1"]
[PlyCount "0"]
[EventDate "1937.??.??"]
{Another classic ending. White has little trouble winning the a-pawn, but will
this be enough to win the game? White is handicapped by the fact that his
bishop is the wrong colour for the h-pawn, so even winning Black's bishop may
not be sufficient. In the game White won, but various analysts who have
examined this ending in detail have come to the conclusion that Black should
draw with accurate play. But if even the great Capablanca could not find the
draw, can a machine?} *
[Event "1/99-60 Hannover"]
[Site "Hannover"]
[Date "1926.??.??"]
[Round "?"]
[White "NET10: von Gottschall"]
[Black "Nimzowitsch"]
[Result "*"]
[Annotator "John Nunn"]
[SetUp "1"]
[FEN "3R4/1p6/2b5/2P1k2p/p3p2P/P6r/1P2KB2/8 b - - 0 1"]
[PlyCount "0"]
[EventDate "1926.??.??"]
{Once again we have opposite-coloured bishops and equal material, but here the
defender's task is complicated by the presence of rooks. Indeed, the practical
difficulties proved too much for White in the game, and he lost. However,
later analysts indicated possible improvements for White - my own view is
that the ending should be a draw, but it would certainly be unpleasant to
defend against a strong endgame player. Wieder einmal haben wir
ungleichfarbige Laeufer und gleiches Material.} *
Regards
Dave
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.