Playing White against a computer, I reached this position (White to move):

[fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

White has a queenside pawn majority, but Black's king is closer to the action, and its knight has the idea of Nd4-c2. I played this position through to the end about a dozen times, and usually only managed to draw: Black was able to pick up my queenside pawns, and even in lines where I got one of Black's kingside pawns in compensation, the position was still drawn. A couple of times I did manage to push the queenside pawns and win, but I didn't keep track of exactly what I did and it's possible the engine made a defensive error (it was set at a fairly easy level). So what should White's plan be in this position? Does White have a win here, or is this a draw?

  • 1
    I've been running this position in Stockfish for the last few days. So far the winning method seems to be to force Black to commit his King to the defense of the Queenside. This allows White's King to run rampant on the Kingside. – Tony Ennis Jan 3 '14 at 13:11
  • I have tried my best to help you with my answer. Hopefully you can make use of it. Good luck and best regards. – AlwaysLearningNewStuff Jan 7 '14 at 2:26

EDITED ( January 7th, 2014 ):

Based on the member shivsky's comments I have edited the entire solution.

Does White have a win here, or is this a draw?

Yes it is won for White.

So what should White's plan be in this position?

At first, I wanted to provide an answer with 1.a4 and simply force the win, but I do not have computer strong enough to analyze the variations I come up with, so my conclusions can not be 100% true.

I do not think that you can achieve much by creating the passed pawn immediately, at least my engine shows some convincing defensive lines.

And again you were right-even if you trade queen side pawns for his king side pawn, he might still draw.

My engine shows some lines where 1.a4 continuation allows you to get f7 pawn for your queen side pawns-giving White an advantage, but I strongly believe you can not win since all the pawns are at one side and Black's Knight can do a good job at defending and blocking.

Being strong endgame player, I must say that I do not like 1.a4 anyway.


Your only chance is to deaden the position and to win by using your queen side pawns to distract his pieces from defending king side, when your pieces could invade there and get enough material to enforce queening. You need to start the action at the right moment.

1.Kf1 can achieve that, when you could freely maneuver into the favorable setup which will create passed pawn and win.


I have tested this line with the engine to see if there is a valid defense, and engine could find none.

I have tried to find one myself and 1. ...Nh4 was the best try I could find.

It is the only way to prevent White from activating his king ( if Black allows that then its game over ) and the idea is to harass White's King side pawns so Black could play f6 + g5 + Nf5 hoping to achieve something like this:

[fen "8/8/p3kp2/5npp/PPN4P/3K2P1/5P2/8 w - - 0 1"]

1. hxg5 fxg5

This would give him some counter-play ( he threatens to queen with h4 ) and would provide him a chance to exchange all king side pawns.Then he could try to catch the queen side pawn ( a6 pawn will 100% be exchanged for the b5-White has to play it at some point ) or to sacrifice the Knight for a pawn ensuring the draw.

This is how I would try to save the game as Black, but White can refute this plan:

[fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

1.Kf1 Nh4 2.g3 Nf3 3.h4 Nh2+ 4.Ke2 Ng4 5.f3 Nh6 6.Ne3! 

This move restricts the Knight completely refuting the above plan.Without the ability to create counter-play Black is simply lost-White creates the queen side passed pawn distracting one or both of Backs pieces and then wins by taking all of Blacks pawns on the king side.


First rule of endgames-king is extremely powerful piece-use it!

Second rule-of all the available moves that can implement your plan always play those that make the least commitment. This gives you the flexibility to adapt when position changes in its character drastically.

The 1.a4 continuation breaks both rules, as it neglects the king and forces the matters-which unnecessarily sharpens the game and provides Black with counter-play based on his strong King.

In endgames you need to slow down the pace and see the whole picture. Raw calculation here is useless if you do not have a clear plan.

To create a plan, we always start with the analysis of the position. Yours is sufficiently good for a start so let us go from there:

White has a queenside pawn majority, but Black's king is closer to the action, and its knight has the idea of Nd4-c2.

Third rule in endgames-in order to win, most of the time you will have to create at least two DISTANT weaknesses in the opponents camp ( if they are close then opponent might defend successfully ). Here we already have one-extra pawn on the queen side.

Our pawn structure is healthy-we have no pawn weaknesses on queen or king side ( doubled/isolated pawns... ).

The weakest piece in our setup is our king and that is our only weakness.

Our Knight is A MONSTER! This piece alone actually won this endgame! He is ideally posted-he can aid in creation of a passed pawn, he can attack on the queen side, he restricts the opposing Knight and King, he can harass the king side pawns if black king abandons them and he can defend our king side. WHAT A BEAUTIFUL PIECE! A HERO!

Now let us see what Black has:

Black's king IS active as you have correctly noticed, but he is badly placed-he needs to be closer to the center to balance the lack of a pawn at queen side ( e6 would be nice )-and that is something you have missed.

Black's knight is OK, he is posted well, but is decentralized which limits his actions-he can only attack on the king side at the moment.

Black has no pawn weaknesses, but is a pawn down. He also has started to probe the king side since h-pawn has advanced and can be supported with he knight and a king. He should try to create counter-play there, since he has no time to attack the queen side.

Let us see the entire picture:

Black's pieces are misplaced ( you missed this fact and this is important to notice ), they need 1 or 2 moves to centralize them selves ( they must do this in order to hold queen side ), and will spend even more moves to organize a defense/create counter attack.

This will give White enough time to activate his king.

Now we see the first phase of our plan-improve the king's position.

The second part should be about creating the second weakness-this type of endgame is well known and the usual plan is to create passed pawn on the queen side to distract king/knight/both and then to use our extra pieces to break through on the king side.

You saw what happens when Black tries to generate counter-play on the king side, now let us see what happens when he tries to centralize his pieces so he can attack on the queen side:

 [fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

 1.Kf1 Nd4 2.Ke1 Ke6 3.Kd2 Kd5 4.Kd3

At this point Black is in zugzwang, and no matter what he plays, White will simply parry his threats and eventually play a4 fulfilling the plan described above. Let us just see one sample continuation so you "get the feel":

 [fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

 1.Kf1 Nd4 2.Ke1 Ke6 3.Kd2 Kd5 4.Kd3 Nc6 5.Ne3+ Ke5 6.Kc4 f5 7.a4 f4 8.Nd5 g5 9.Nc7 Nb8 10.b5 axb5+ 11.axb5 Kd6 12.b6 Kc6 13.Nd5 Nd7 14.Kd4 Nxb6 15.Nxb6 Kxb6 16.Ke5

And White wins. The queen side pawn majority did its job-it has distracted the knight, and eventually the king from defending the king side which was exploited with White's last move.

One thing I must mention is the usage of zugzwang. At certain point Black did manage to centralize his pieces, but so did White. At that moment everything boils down to small nuances and here it is White knight that saved the day-by restricting both king and knight he enabled the creation of zugzwang after which the creation of the passed pawn was easy.

Notice also, how the activity of the Black's pieces just wasn't good enough for a draw, since they were badly misplaced and notice how that gave White the time to improve his own king.

I think that this is enough for you to understand how to win this alone, so set up the position on your computer and try to win this time.

Good luck!

Best regards and a Happy New Year!

| improve this answer | |
  • Please considering revision with a concrete and complete answer. A work-in-progress doesn't fit too well. – shivsky Jan 1 '14 at 11:41
  • 1
    @shivsky: Based on you comment, I have edited my answer ( although I believe you misunderstood my answer ). Please reconsider your down vote, and tell me what you think so I can improve it if necessary. Best regards. – AlwaysLearningNewStuff Jan 7 '14 at 2:24
  • In your second diagram, 3...Nh2+ loses a lot of time. 3...Nd4! is much stronger, having in mind ...Nf5, ...f6, ...g5 as you mention, but also preventing Ke2, covering b5 so that WHite cannot create a passer too fast, and even forking pawns with ...Nc2 if WHite is careless. 4.Ne3 seems necessary, but then 4...Ke5 and 5...Ke4 creates some counterplay. White can probably still win, but the path is far from easy. – Evargalo Nov 17 '17 at 9:30
  • 1
    @Evargalo: Thanks for pointing it out, I think I have analyzed that continuation but I really can not remember what happens in the end... I remember though that I too liked your suggestion. I will try to spare some time to look at it again and will post my findings if anything interesting appears. Best regards! – AlwaysLearningNewStuff Nov 17 '17 at 17:18

It's technical, but white wins in this position. White's plan should be to execute typical "outside passed pawn" tactics:

  1. Activate King!
  2. Play with zugzwang

The idea is that white threatens to attack on both flanks with pawn advances. Black can't stop two threats at the same time because the battlefield is too big.

Without the knights it would be easier, but with the knights it is a bit harder. With a bishop, rook or queen the plan becomes quite difficult to execute because these pieces can defend both flanks at the same time.

Here are two sample solutions:

[fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

1. Kf1 Nd4 (1...Ke6 2. Ke2 Nh6 3. Ne3 Ng4 4. Nxg4 hxg4 5. Kd3 Kd5 6. h3 gxh3 7. gxh3
Kc6 8. Kc4 Kd6 9. f4 Kc6 10. a4 Kb6 11. Kd5 Kc7 12. Kc5 Kb7 13.Kd6 1-0)
2. Ke1 g5 3. Kd2 Nf5 4. a4 Ke6 5. Kd3 Nh4 6. Ne3 f5 7. Kc4 f4 8. Nd5
Nxg2 9. a5 f3 10. b5 Kd6 11. bxa6 Nf4 12. a7 Nxd5 13. a8=Q Nc7 14. Qxf3 g4 15.
Qf6+ Kd7 1-0

A more detailed explanation on this type of position is given here: http://www.mark-weeks.com/aboutcom/aa03g19a.htm

| improve this answer | |
  • 1
    Could you provide some short,concrete variations on how White can win? What is the first move for White in your winning plan? The endgame you have linked to will not suffice, in my opinion, as there Black king is equally strong as White, and position is much better for White since he can maneuver, yet in this concrete example every tempo counts. I am really interested to see your solution and learn from it. Best regards. – AlwaysLearningNewStuff Jan 3 '14 at 7:18
  • I editted the solution. – Rafiek Jan 3 '14 at 7:43
  • I ran the main line through Stockfish at about 5 minutes per move. According to Stockfish, it's generally reasonable except Black's move 8, which is fatal. However, Black seems doomed anyway. – Tony Ennis Jan 5 '14 at 16:32
  • 1
    I also ran the 1...Ke6 variation through Stockfish. I think the entire line is bad as Black defends poorly. Black loses instantly when he allows his Knight to be traded. – Tony Ennis Jan 5 '14 at 17:28
  • Rafiek, if the main line came from brain and not an engine, I'm impressed. One thing I have noticed is that Stockfish really wants White to play a4 to prepare for creating a passed pawn. It piles pressure onto Black and reduces his options. It's a common theme in your main line, your variation, and the analysis I posted elsewhere on this thread. It's interesting. – Tony Ennis Jan 5 '14 at 17:43

These moves were selected by Stockfish running on a decent 2-core CPU at 30 minutes per move. While Stockfish is a fine engine, computers aren't great at the end game, so I expect that there could be some improvements.

[fen "8/5pp1/p4k2/5n1p/1PN5/P7/5PPP/6K1 w - - 0 1"]

1.Kf1 Ke6 2.Ke2 Kd5 3.Kd3 Ne7 4.a4 Nc8 5.Ne3+ Kc6 6.Nf5 g6 7.Nd4+ Kc7 8.Nb3 h4 9.Nc5 Kb6 10.Nd7+ Kc6 11.Ne5+ Kd5 12.Nxf7 Nd6 13.Nxd6 Ke6 14.Ne4 *

...and it's lights-out for Black.

What we see here for the first 7 half-moves are both players centralizing their Kings and preparing for a skirmish on the Queenside. Black's 4th move looks to be a clunker though it is still about bringing the Knight to the Queenside. What's happening here is that White is taking the slack out of the Queenside and limiting Black's response. This makes additional calculation easier.

White punishes 4...Nc8 by a series of Knight moves that harass the Black King, which must stay near the Queenside. A fork develops and Black loses material without compensation on the Kingside. Black can honorably resign after White's 11th move, IMO.

After 13.Nxd6 there are no good moves and Stockfish goes wonky. It hardly matters - Black is done.

The strategies in the position are: 1. Black will try to equalize on the Queenside 2. White will be willing to surrender two pawns for one pawn on the Queenside, as long as the Black King is misplaced (decoyed!) to the rim. White's King can then run amok on the Kingside.

To answer the OP's question, yes this is winnable. First, it looks to me like White can win with proper play. Second, unless you're playing at a very high level, nearly every position is winnable - you just have to capitalize on the opponent's mistakes. Chess is a game about outwitting your opponent, not making perfect moves.

Final edit - I don't see a variation where Black can force a draw. White's extra pawn forces Black to make decisions that are fatal in the long term.

| improve this answer | |

Analysis with Houdini 6.02 1 Kf1 Nd4 2 Ke1 Ke6 3 Kd2 h4 4 a4 Kd7 5 Kd3 Nc6 6 Kc3 Kc7 7 f4 Ne7 8 Ne3 Ng6 9 Nd5+ Kd6 10 Kd4 Nf8 11 Ne3 Ne6+ 12 Ke4 Kd7 13 Ke5 Nd8 14 Kd4 Kc6 15 Nf5 Ne6+ 16 16 Ke5 Nf8 17 Ke4 Nd7 18 Nxg7 Nb6 19 Ke5 Nxa4 20 Nf5 Nb2 21 Kd4 Kb5 22 Kc3 Nd1+ 23 Kb3 Nf2 24 Nd6+ Kc6 25 Nxf7 Kd5 26 Ka4 h3 27 g3 Ng4 28 Ka5 Nxh2 29 Ng5 Kc4 30 Nxh3 Ng4 31 Ng5 Nf6 32 Nf3 Nd5 33 Ne5+ Kd4 34 Nc6+ Ke4 35 Kxa6 Kf3 36 b5 Kxg3 Mate in 21 according to Nalimov tablebases , starting with f5. Two iterations of 20 and 15 minutes. Thanks.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.