White to move. Is this a win, draw, or loss for White? How should White proceed?
8/6p1/4p3/4P3/3P3P/1n2K3/8/k7 w – – 0 1
Marysko, 1943
Chess Daily News from Susan Polgar
Difficult endgame improvement
About Author
Probably draw.
I’ll give 2 lines, that I think will convincingly show that White wins:
1.d5 exd5 2.e6 Na5 [ d4+ 3.Ke2 ] 3.e7 Nc4+ 4.Kd4 Nd6 5.Kxd5 Ne8 6.Ke6 Kb2 7.Kf7 Nd6+ 8.Kxg7 Nf5+ 9.Kf8 Nxe7 10.Kxe7
1.d5 Nc5 2.Kd4 [2.dxe6? Nxe6 3.Ke4 g6!] Nb3+ 3.Kc3 Nc5 4.Kc4 Nd7 5.dxe6 Nxe5+ 6.Kd5 Ng6 7.Kc6 [ 7. Kd6? Nxh4 8.e7? Nf5+ ] Kb2 8.h5 Ne7+ 9.Kd7 Nd5 10.e7 Nf6+ 11.Kd8
Well, I see two possible plans here- either Ke4 followed by a pawn exchange at d5 to create a passed e-pawn, or an interpolation of moves in this plan- or make a dash for the g-pawn. Now, an immediate problem with 1.Ke4 is the knight check from d2. One other idea might to be do an end-around maneuver by bring the king to d3, c3, b4. I am going to do the analysis on the dash for the g-pawn first just to prove to myself it doesn’t actually work:
1. Kf4 Nd4 (why not?)
2. Kg5 Nf3 (damned knights!)
3. Kg6 Nh4
4. Kg7 Nf5!
And it is clear that the black king can reach c3 long before white’s king can reach the knight at d4. So that plan is out.
With Nd2+ to muck up 1.Ke4, I think white must play d5 first if he is to create the passed e-pawn. Let’s see where we can get to with this idea:
1. d5
Now, black can take at d5, or play Na5 or Nc5. Let’s take each move in order:
Variation A with 1. ….ed5:
1. …..ed5
And it is easy to see that white can just push as long as he is careful not to fall into insidious knight fork traps (something I am doing constantly)
2. e6 Na5 (d4 is below)
3. e7 Nc4
4. Kd4 Nd6
5. Kd5 Nb5 (one last trap)
6. Kc6! and now the pawn will queen. At move two for black, he could play d4, and white must play very carefully to avoid forks on the e-pawn and to deal with black’s d-pawn:
2. …..d4
3. Ke2! d3 (Nc1 4.Kd1+-)
4. Kd1 and the e-pawn can’t be stopped. Clearly, black cannot play 1. …ed5.
Variation B with 1. …Nc5:
1. d5 Nc5
2. Kd4 Nb3 (Na6/d6 3.de6+-)
3. Kc4 Nd2 (Na5 4.Kb5+-)
4. Kc5 Nf3 (Ne4 5.Kc6 ed5 6.Kd5+-)
5. de6 Ne5 (what else?)
6. e7 Nd7
7. Kd6 Nf6
8. Ke6 Kb2 (everything loses)
9. Kf7 Kc3 (Ne4 10.Kg7 Nd6 11.h5)
10.Kg7 Ne8
11.Kf8 Nd6
12.h5 and one of the pawns will queen. Finally,
Variation C with 1. …Na5:
1. d5 Na5
2. Ke4
This is the quickest win. I spent a lot of time on 2.ed6, but found a drawing line, at the very least, for black, believe it or not, in which black causes white to waste so much time getting the king to d7 to support the advance of e7, that the black king has time to reach d4 to claim e5 and the knight sacrifices himself for e7. Continuing:
2. …..Kb2 (everything loses here)
3. de6 Nc6
4. Kd5 Ne7
5. Kd6 Nc8 (Nf5 6.Kd7+-)
6. Kd7 Nb6
7. Kd8 Nc4
8. e7 Ne5
9. e8(Q) and there is no counterplay for black- he is toast.
So, I think white wins outright with 1.d5. The knight is helpless against the king and the resulting passed pawn, but there are numerous traps to be dodged, as always when there are knights hopping about.
1. d5 Nc5 (1. … exd5 2. e6 d4+ 3. Kf2) 2. Kd4 Nb3+ 3. Kc4 Nd2+ 4. Kd3 Nf3 5. dxe6 Nxe5+ 6. Ke4 looks like a win for White.
e.g., 6. … Ng6 7. Kd5 Kb2 8. Kd6 Kc3 9. e7 Nxe7 10. Kxe7 Kd4 11. Kf7 Ke5 12. Kxg7 Kf5 13. h5
Actually I think the correct line might be this:
1. d5 Nc5 2. Kd4 Nb3+ 3. Kc4 Nd2+ 4. Kd3 Nf3 5. dxe6 Nxe5+ 6. Ke4 Ng6 7. h2 Ne7 8. Ke5 Kb2 9. Kd6 Nf5+ 10. Kc7 (10. Kd7 Ng3) Kc3 11. Kd8 Kd4 12. e7 Nxe7 13. Kxe7 Ke5 14. Kf7
Typo in previous comment: 7. h2 should be 7. h5
1. d5 Nc5 2. Kd4 Nb3+ 3. Kc4 Nd2+ 4. Kd3 Nf3 5. dxe6 Nxe5+ 6. Ke4 Ng6 7. h5 Ne7 8. Ke5 Kb2 9. Kd6 Nf5+ 10. Kc7 Kc3 11. Kd8 Kd4 12. e7 Nxe7 13. Kxe7 Ke5 14. Kf7
1. d5! wins for white.
d5
ment1.d5 exd5 (Nc5 2.Kd4)
2.e6 Na5
(2…d4+ 3.Ke2! d3+ 4.Kd1)
3.e7 Nc4+ 4.Kd4 Nd6 5.Kxd5 Ke8 6.Ke6 Kb2 7.Kf7 and white is winning
First move seems obvious: 1. d5, then:
1. d5 exd5, but probably not
2. e6, because of d4+
3. Kd3 Nc5+ and Nxe6 or
3. Kf4 d3
4. e7 d2
5. e8Q d1Q
No idea, who wins this, but this is not, what we want. So let’s play:
2. Kd3 Nc5+
3. Kd4 Ne6+
4. Kxd5 Nd8
5. e6
The black knight has to go for the passed pawn and the better king position wins the game for white. If
2. Nc1+
3. Kd4 Ne2+
4. Kxd5 Nf4+
5. Kd6, this makes no difference for black. greets, jan
No there is a forced win, I won’t tell you what it is because I found it with an engine 🙁 Having gave up after many tries.
Looks very tempting to play
1.d5 Nc5
[1…exd5 2.e6 +-]
2.Kd4 Nb3 3.Kc4 Na5
[3…Nd2 4.Kb5 exd5 5.e6 +-]
4.Kb5 +-
I don’t know if I missed something here.
– Shant
Without a board and the help of any of our electronic friends I just can give an estimation.
It seems to me that
1. d5 brings Black in difficulties. However if
1. exd5
2. e6 Sc5
3. Kd4 Ne7
White can’t bring his King to the Knight
So better is
1. Kd3 and now after
1. … Kb2
2. Kc4 renews the d5 idea with more success and after
1. … Na5
2. d5 also is strong
2. … exd5
3. Kd4
Black can’t prevent White to support the e pawn with his King and later take the g7 and probably win.
d5, and white wins
Best from Hungary
With accurate play, this should be a win for White. In general, a knight can only stop a pawn from advancing if the pawn has no support.
In this position, White will achieve a passed pawn by playing d5, which Black’s knight cannot stop without sacrificing itself. Because White’s king is in a position to help his pawns, a pawn will advance to promotion unless the knight captures it.
Should Black trade pawn for pawn and knight for pawn, White’s king will be closer to g-file than Black’s king. At that point, it becomes an exercise in move counting.
In an endgame like this, pawns need to be pushed. Push that d-pawn and see if either it or the e-pawn can out-run the knight.
Look for Knight tricks, its the ULTRAMODERN WAY!!! They even got robots playin that they think is ultramodern! hehe.