2 knights v pawn

Technical questions regarding Openings, Middlegames, Endings etc.
User avatar
Andy Burnett
Posts: 88
Joined: Wed Dec 31, 2008 12:19 pm
Location: Scotland
Contact:

2 knights v pawn

Post by Andy Burnett » Wed Jan 28, 2009 9:39 pm

Lawrence Trent had an excellent result today drawing as black against GM Ivan Sokolov (2657). I thought he might even be winning at some stage.

Anyway, to get to the point, Sokolov ended up with 2 knights against a rook's pawn and didn't even come close to showing the technique needed to win this. (I'm not saying I would have had the foggiest idea either, but then I'm not rated 2657!)

With the black pawn blockaded on h5 it is winning for the knights according to Batsford Chess Endings. There is a Troitzky line (a4-b6-c5-d4-e4-f5-g6-h4) - if the (black) pawn is further advanced than these squares it is only a draw (barring obviously favourable positions of the knights and kings).

So, is this a case of the modern time control eroding accurate endgame play, or a very strong GM not having a scooby what to do, or a bit of both?

The game should be viewable here:-
http://www.gibraltarchesscongress.com/g ... index.html

James Coleman
Posts: 377
Joined: Fri Nov 21, 2008 7:11 pm

Re: 2 knights v pawn

Post by James Coleman » Wed Jan 28, 2009 9:55 pm

Yeah I noticed that too.

Philosophical point though - is the position winning if it can't be won vs best defence ? That looks like it is the case here.

After Trent's h5 the position is mate in 80. Within that 80 move sequence there is a 50+ move sequence where Black doesn't have to move his pawn.

I found the winning procedure basically incomprehensible !

Paul McKeown
Posts: 3388
Joined: Thu Apr 12, 2007 3:01 pm
Location: Hayes (Middx)
Contact:

Re: 2 knights v pawn

Post by Paul McKeown » Sat Jan 31, 2009 12:09 am

I remember Conor O'Shaughnessy winning a game in the 4NCL in this endgame. Turned out that the endgame was a theoretical draw (according to the endgame tablebase), but in practise was impossible to defend.

Paul McKeown
Posts: 3388
Joined: Thu Apr 12, 2007 3:01 pm
Location: Hayes (Middx)
Contact:

Re: 2 knights v pawn

Post by Paul McKeown » Sat Jan 31, 2009 12:16 am

And as for not having a scooby, what is it with KBN v. K? Pretty straightforward really, I won it once in the 4NCL, and I saw recently in the Thames Valley League Tim Seymour's opponent from Maidenhead simply resigning, saying he couldn't imagine anyone over 140 not delivering mate. Yet I remember my old mate Jim Stevenson (having lost 2 queens in the same game to knight forks!!! - how do you do that???) grimly hanging on in the BCF Major Open (2000 I think; might have been 1999) to draw this endgame against his young opponent who became British U18 champion that year. The youngster bashed out about 20 moves quickly without thought, before sinking into ddddddeeeeeeeepppp thought. We (Rick McMichael, Jonny Kay and myself) all headed of to the pub, safe in the knowledge that Jim was heading for a draw.

User avatar
IM Jack Rudd
Posts: 4008
Joined: Tue Apr 17, 2007 1:13 am
Location: Bideford

Re: 2 knights v pawn

Post by IM Jack Rudd » Sat Jan 31, 2009 12:55 am

I successfully held this endgame against Bogdan Lalic once. Got my pawn to g5 before it was blocked, then made sure that if my king was forced into a corner, it was a1. (Thus allowing my pawn to queen with check on the critical move.)

Keith Arkell
Posts: 748
Joined: Tue Dec 30, 2008 2:10 am

Re: 2 knights v pawn

Post by Keith Arkell » Tue Mar 24, 2009 3:08 am

Some 25 years ago at Hastings Gary Lane was adjourned in a position which was inevitably going to reduce to N+N v P. He wanted to find his opponent and offer him a draw.I stopped him and gave him a brief explanation of why the position was winable. upon resumption Gary's opponent,who was making his moves quickly,with irritation, and without any thought,got himself checkmated very easily. In those days the concept that the position was winable was far less generally known.Gary was surprised,and his opponent was obviously totally oblivious of the fact.

Simon Spivack
Posts: 600
Joined: Wed May 14, 2008 4:06 pm

Re: 2 knights v pawn

Post by Simon Spivack » Tue Mar 24, 2009 6:21 pm

Keith Arkell wrote:Some 25 years ago at Hastings Gary Lane was adjourned in a position which was inevitably going to reduce to N+N v P. He wanted to find his opponent and offer him a draw.I stopped him and gave him a brief explanation of why the position was winable. upon resumption Gary's opponent,who was making his moves quickly,with irritation, and without any thought,got himself checkmated very easily. In those days the concept that the position was winable was far less generally known.Gary was surprised,and his opponent was obviously totally oblivious of the fact.
But Knight Endings by Averbakh and Chekhover was published in English in 1977, more than thirty years ago. There is a chapter by Chekhover on this very ending. Am I to infer that I am one of the few dumb enough to own a copy, or at the very least have it on my shelf, read. :-)

Ian Kingston
Posts: 1070
Joined: Sat Apr 14, 2007 3:16 pm
Location: Sutton Coldfield
Contact:

Re: 2 knights v pawn

Post by Ian Kingston » Tue Mar 24, 2009 8:13 pm

I was made aware of the possibility of winning this ending by the second chess book I ever read: Further Steps in Chess, by Owen Hindle (Routledge & Kegan Paul, 1968). It's a book for novices, the second in a series (Book 1 was An Introduction to Chess, by Leonard Barden; I've lost my copy of that one).

Hindle gives just one example (White: king on f7, knights on d2 and f5; Black: king on h7, pawn on d3) to show how the pawn prevents the stalemate defence. It surprises me that this was not well known in the mid-80s.

Database trivia: In 2007, Arkadiusz Bebel (2142) won this ending twice in the same tournament (Rewal Open, Poland).

John Hickman
Posts: 195
Joined: Sun Sep 07, 2008 8:35 pm

Re: 2 knights v pawn

Post by John Hickman » Tue Mar 24, 2009 9:09 pm

FM Jack Rudd wrote:I successfully held this endgame against Bogdan Lalic once. Got my pawn to g5 before it was blocked, then made sure that if my king was forced into a corner, it was a1. (Thus allowing my pawn to queen with check on the critical move.)
I assume Jack that you then knew how to win with Queen vs 2 Knights :wink:

Angus French
Posts: 1640
Joined: Thu May 15, 2008 1:37 am
Contact:

Re: 2 knights v pawn

Post by Angus French » Wed Mar 25, 2009 11:20 am

Wikipedia has an entry on the ending here: http://en.wikipedia.org/wiki/Two_knights_endgame

Keith Arkell
Posts: 748
Joined: Tue Dec 30, 2008 2:10 am

Re: 2 knights v pawn

Post by Keith Arkell » Wed Mar 25, 2009 6:19 pm

Ok yes of course these things were in one or two books,but in my experience they were not common knowledge in quite the way they are today

Simon Spivack
Posts: 600
Joined: Wed May 14, 2008 4:06 pm

Re: 2 knights v pawn

Post by Simon Spivack » Thu Mar 26, 2009 7:22 pm

Keith Arkell wrote:Ok yes of course these things were in one or two books,but in my experience they were not common knowledge in quite the way they are today
There is a danger that we may reach an agreement, assuming I can persuade Keith to take a small number of steps.

I have looked at some endgame books available in the '80s, these are in addition to the two books mentioned in previous posts. I have included comments inline in the following list:

Basic Chess Endings by Reuben Fine, 1972 reprint ; covered on page 86

Practical Chess Endings by Paul Keres, 1974 edition; not mentioned

Chess Endings: essential Knowledge by Yuri Averbakh, 1973 reprint; mentioned on page 8

360 Brilliant and Instructive Endgames by Alexis Troitsky, published 1968; I couldn't find a mention, unfortunately, though, there is no index, so I could be in error.

One book I have buried somewhere, but can't find, is Hooper's book on endings.

I probably shouldn't write this, and Keith is far better placed to judge than me, but I am none too impressed by the Keres book. I am of the view that this great player did not stretch himself in its writing.

Thus, I'd suggest that of the main endgame books available to UK players then, BCE and the Averbakh series both covered it. The former more as a sketch, the latter thoroughly. It would be useful if someone has a copy of Hooper.

A professional chess player should attempt to eliminate as many lacunae from his game as possible. However, the majority of UK chess players are amateurs who play for pleasure. I don't knock anyone for preferring not to read a book on the ending; but what I do suggest is that that was a common attitude in the '80s.

I should like to know, however, what changes have caused the increased awareness. Obvious candidates are the omnipresent Internet and the increase in the number of Quickplay Finishes. But is this a correct surmise?

Keith Arkell
Posts: 748
Joined: Tue Dec 30, 2008 2:10 am

Re: 2 knights v pawn

Post by Keith Arkell » Thu Mar 26, 2009 8:32 pm

I can honestly tell you Simon that I don't own or have any obvious access to any endgame books whatsoever,so can't comment on the Keres one.

Maybe from time to time in the past I've owned one or two.I remember annotating a game I played v Panzer,Hastings '90,for BCM [Article heading 'Nothing new under the sun'].This was Q+RP v Q,and afterwards when I got home I consulted an endgame book to see what it had to say on this ending.I found the analysis was almost identical to the way the game went.I wonder if that book was the Keres one? Anyone with a full collection of BCMs could check this out because I think I made reference to which book it was.
I think that the ''English Chess Explosion'' of the late 70s resulted in a lot of talented practical players being unleashed,who,as a generalisation,though of course not in every case, were perhaps not as clued up on the accumulated knowledge of the game as they/we could have been.
I think that my own intro to the N+N v P ending came about by accident because of an article I read somewhere.

I wonder for how many I speak when I say that any skill I may have acquired in the endgame has come about almost totally through practical play.
People probably assume wrongly that just because I have learnt how to squeeze the maximum out of an ending,that this also means that I know which ones are winning,which ones are drawing,and at what point for example the transition from winning to drawing occured.
I saw an annotation by John Nunn of one of my(17!) R+B v R endings in his excellent book on pawnless endings,and it fascinated me to see how my opponent and I drifted in and out of positions which were winning and which were drawing.During the game all that I was aware of was that I was giving my opponent difficult practical problems to solve.

Leonard Barden
Posts: 1552
Joined: Wed Dec 24, 2008 11:21 am

Re: 2 knights v pawn

Post by Leonard Barden » Fri Mar 27, 2009 12:21 pm

Yesterday Anand reached 2Ns v P in the final round of Nice Amber. Winning it would have shared first prize, and he was playing the bottom marker Wang Yue, but.....it was the blindfold game, and the world champion only drew.

The now defunct magazine Inside Chess found a picturesque title for 2 Ns v P - the "Halley's Comet endgame" because you should see it only once in your lifetime.

User avatar
John Saunders
Posts: 1246
Joined: Thu Apr 19, 2007 3:10 pm
Location: Kingston-upon-Thames
Contact:

Re: 2 knights v pawn

Post by John Saunders » Fri Mar 27, 2009 3:51 pm

Keith

Here is that article you are referring to, from the April 1991 issue of BCM...

http://www.bcmchess.co.uk/img/bcm199104arkell.jpg
Personal Twitter @johnchess / Personal Website http://www.saund.co.uk / Britbase http://www.britbase.co.uk

Post Reply