PDA

View Full Version : K+N+N+2P v K+N+N



Desmond
18-09-2006, 10:25 PM
4k1n1/6p1/7p/2NN4/4K3/8/7n/8 w - - 50 1
I had a position something like this arise in a rapid game on the weekend.

I had the white pieces, and eventually lost.

Of course, I would love to give up my knights to get the two black pawns, but could not find a way to achieve this.

I found that my knights constantly had to give up ground to avoid exchange and due to having the move.

Is this position lost? Would my drawing prospects be better if the pawns were in the middle (eg. d- and e- pawns)?

Garvinator
18-09-2006, 10:29 PM
While I was watching I was thinking that a tablebase would just love some of these positions, especially after a pair of knights got exchanged for the pawn.

Desmond
18-09-2006, 10:33 PM
While I was watching I was thinking that a tablebase would just love some of these positions, especially after a pair of knights got exchanged for the pawn.
Unfortunately, I have no tablebase.

Yes, I should clarify. In the game, I saced a N for the h-pawn. In hindsight, I probably should have tried to grab the g-pawn instead.

Garvinator
18-09-2006, 10:38 PM
Unfortunately, I have no tablebase. details, details :P

Rincewind
18-09-2006, 11:50 PM
details, details :P

Far too many pieces on the board for tablebases to be of much use.

To my limited endgame undertstanding it looks won for Black with best play to me. If White sacs a knight for a pawn Black should be able to avoid the other sac being a pawn and knight better. If a pair of knights are swapped then it is very grim for White as knight and pawn will nearly always win, even if the pawn in a rook pawn. There are some exceptions but the position is far from problematic especially if the pawn is not far advanced.

Perhaps if Black's king was further away there might be drawing chances.

Garvinator
18-09-2006, 11:53 PM
Far too many pieces on the board for tablebases to be of much use.
I did say at the end of my comment about after a pair of knights and h pawn was exchanged.

Rincewind
19-09-2006, 12:04 AM
I did say at the end of my comment about after a pair of knights and h pawn was exchanged.

Well if you are talking about a K+N+P v K then you hardly need a endgame tablebase to help there. Unless the attacker can win the pawn by force he will lose regardless unless he gets to one of very few drawing positions. AFAIK provided the attacker doesn't play his pawns up the board too quickly it should not be a danger. Even if the attacking king is far away, as long as the knight can defend the pawn from behind all is well.

One of the drawing positions (maybe the only main one) is with a rook pawn (I think the defending king in front of a lone rook pawn on the seventh rank is a draw) so perhaps Boris should have tried to take the g-pawn rather than the h-pawn (as he said) but even then I would suspect Black would have had to stuff it.

MichaelBaron
19-09-2006, 02:27 PM
I think this is a very hard position to analyze....

I suspect that with precise play, white should be able to draw...but I am not sure:hmm:

Desmond
19-09-2006, 02:43 PM
I think this is a very hard position to analyze....

I suspect that with precise play, white should be able to draw...but I am not sure:hmm:Thanks Michael.

Would you agree that the drawing chances would be better with centre pawns - more dancing space for the knights?

Garrett
19-09-2006, 05:18 PM
Perhaps oyu could try putting it into the computer and letting it play it out over a day or two at high level. It might give you an idea.

Phil Bourke
19-09-2006, 05:58 PM
The original position is drawn. It took a few hours, not solid, but just kept coming back to it, to find the drawing idea/s.
1. Kf4 Nf1 (1... g5+ is the only other move that deals with the threats of Ne3 and Kg3 2. Kf5 Ne7+ 3. Nxe7 Kxe7 4. Ne4 Nf3
5. Kg6 g4 6. Kxh6 Kf7 7. Kh5 Ne5 8. Kh4 Kg6 9. Nf2 Kf5 10. Nxg4) 2. Kf5 Kf7 (
2... Ne7+ is a good try to confuse matters 3. Nxe7 Kxe7 4. Kg6 Ng3 5. Kxg7 h5 6. Nd3 Ne2 7. Kg6 h4 8. Nf2 Nf4+
9. Kg5 Ng2 10. Nh3) 3. Ne4 g6+ 4. Ke5 Nh2 5. Kf4 h5 6. Ne3 Nh6 7. Nf2 h4 (7...
Kf6 8. Kg3 N2g4 9. Nfxg4+ Nxg4 10. Nxg4+ hxg4 11. Kxg4) 8. Kg5 Kg7 9. Kxh4 Nf3+
10. Kg3 Ne5 11. Kf4 Nhf7 12. Nf5+ Kf8 13. Ne4 Nd3+ 14. Kg4 Ke8 15. Nh4 Nde5+
16. Kg3 g5 17. Nxg5
Will be interested in replies tearing apart this analysis as I found the position to be quite instructive.

Bill Gletsos
19-09-2006, 08:15 PM
Fritz 9 evaluates it as = (0.00) from the starting position.

1.Kf4 Nf1 (1... g5+ 2. Kf5 Ne7+ (2... Nf3 3. Kg6 g4 4. Ne4 Kf8 5. Ne3 Ne5+ 6. Kh5=) 3. Nxe7 Kxe7 4. Kg6 g4 5. Ne4 Ke6 6. Kxh6 Ke5 7. Ng3 Kf4 8. Nh5+ Kf3=) 2. Kf5 Kf7 3. Ne4 g6+ 4. Ke5 Kf8 (4... Nh2 5. Kf4 h5 6. Kg3 Ng4 7. Nf4 N8f6 8. Nxh5 gxh5 9. Kh4 Kg6 10. Ng3=) (4... Kg7 5. Kf4 h5 6. Kg5 Nh2 7. Ng3 Nf3+ 8. Kf4 Nd2 9. Nxh5+ gxh5 10. Kg5=) 5. Nf4 Kg7 (5... g5 6. Nxg5=) 6. Ne6+ Kf7 7. Nd6+ Ke7 8. Nc8+ Kd7 9. Nd6 Kc6 10. Nf5 h5 11. Nf4=

Desmond
20-09-2006, 09:44 AM
Thanks Phil and Bill.

My position in the game was I think a little different. The white pieces were as displayed, but I think black's g8-N was further up the board and I think he king stood better. I can't remember exactly the position, but I think the tricks you found regarding the stray h2-N were not available to me.

Certainly your analysis is very helpful, as it shows that the ending is not always won.