This subforum needs to get more use. Endgame study is VERY important, and it's too often neglected.

 

I've been through Pandolfini's Endgame Course four times now. I only actually read all the text the first time, and I've just been drilling the positions as puzzles since then, only checking the book for explanations afterwards. I'm still having trouble telling the difference between similar positions, and knowing how to play each one correctly, especially all the similar rook and pawn vs rook positions.

 

So to learn these essential endgames really well, I've photocopied the whole book, cut it up, and put the positions on index cards with the solutions on the back. Trying to spot the solutions without the text or answers in front of me is great practice. And I shuffle them up to take away contextual clues of knowing that a position is similar to the one before it. I figure I should know them all well enough to spot the solutions instantly once I go through them another three or four times.

 

Anyway, I was just going through some of them, and this one really got my attention. Considering all the possibilities leads to some weird lines. After a couple of minutes, I figured out what I thought was the correct solution, but when I checked the answer, I noticed that he assumes black's defense will be different than what I thought of. His solution is pretty much the same as mine, just with a slightly different defense by the losing side. So my solution still works - I think.

 

In looking over ALL the possibilities, I've found that white really has to play accurately to avoid the draw. In fact, I even found one line of reasonable seeming moves that leads to a black win. And there's one "incorrect" line that I think might still work to win for white, but I need to think about it more to be sure. Anyway, I thought I'd present it to the group to see what you all think. It's instructive, if nothing else. I'll post the solution and some variations I thought about later.

 

 --Fromper

I would start out with Ke4. That forces Black's king back.

This looks like a case where white can advance his king and win the black pawn. There are some drawing tricks, but white should be able to avoid them. You can't be serious that a reasonable looking line won for black. There must have been 3 serious blunders by white in that line.

 

Here is a sample:

1. Ke4 Kf6 2. Kf4 (not falling for 2. d5 cxd5 3. cxd5 Ke7 with a draw) Ke6 3. Kg5 Kd6 4. Kf6 Kd7 5. Kf7 Kd6 6. Ke8 Ke6 7. Kd8 Kd6 8. Kc8 Ke6 9. Kb7 Kd6 10. c5+ Kd5 11. Kc6 Kxc4 12. Kxb6 and white easily shepherds his extra pawn to the end of the board.

 

 Black can try throwing in the moves ... c5 d5 at some point, but then the black king is restricted to watching white's d5 pawn and white will easily win the c5 pawn. (If the black king ever goes to d4, white wins the pawn race easily.)

Spike, 1. Ke4 doesn't force black's king back (see the sample line I gave above), but it is a good way to start. Either the black king goes back or the white king goes around the black king.

Loomis, 1. Ke4 Kf6 2. d5 cxd5? 3. Kxd5 is also an easy win for white. However, 2... Ke7! is annoying and draws as far as I can see. Both 3. dxc6 Kd6 and 3. Ke5 cxd5 4. Kxd5 Kd7 are a draw.

 

But your variation looks good. I don't see a good improvement for black anywhere.

 

More interesting would be if all pieces went one row down. Then it's trickier for white to win, because after a timely 1...c4 2. d4 Ke4 black has drawing chances with 3... Kd3 4. d5 Kxc3 5. d6 Kb2 6. d7 c3 7. d8=Q c2.

I'm not sure white can win then. Maybe something to think about?

Phobetor, thanks for catching my mistake on black's drawing chances.

 

Moving the pieces down a rank does give black that extra drawing chance. Moving them over a file takes it away again!

 

In your example position, after c4 and d4, white can march his king around to a3, meeting black's ... Kd3 with Kb4, so black has to defend by bringing his king to b5 or a5 to prevent white from playing Kb4. Black also has to have his king in position to recapture on c4. If white is allowed to play dxc4 and hold on to both pawns, it's an easy win for white.

Ok, since this thread has already gotten so much attention, I guess I should post the lines I was looking at.

 

My first instinct was to grab the opposition with 1. Ke4, which is correct. But being the patzer that I am, I wanted to run through the full line to make absolutely certain it works, and I also wanted to see what would happen if I tried other first moves. This is a seemingly simple position, which I'm pretty sure I'd play correctly in a game, even in time trouble, just by following general principles. But by beating the heck out of every possibility in my head, I'm learning and improving. Bear in mind that all my initial analysis (including the line where black wins) was done in my head without moving the pieces. I'm not perfect, but this is great practice for me. While posting the lines here, I'm moving pieces to make sure I don't blunder any further. 

 

1. d5+ draws pretty easily after 1. ... cxd5 2. c5 Ke5. If white pushes the pawn, the black king catches up to it. If he moves his king 3. Kd3 d4 and black's king goes after the white pawn, while white's king blocks the black pawn. Dead drawn.

 

1. c5 looked tempting to me at first glance. As I said, it (correctly) wasn't my first instinct, but I wondered if it might work. That's where I found a couple of lines where either side can win if the other plays improperly, but it should end up a draw.

 

Here's the line where white tries to outflank black and loses both his pawns, giving black the win: 1. c5 Kd5 2. Kd3 Ke6 3. Ke4 Kf6 4. Kf4 Ke6 5. Kg5 Kd5 6. Kf5 Kxd4. I know it seems stupid, but I wandered into that line in my analysis. Looking at it after moving the pieces, I don't think I'd blunder that badly on the 5th move in an actual game, but I did blunder that badly analyzing in my head trying to see that many moves ahead. But I also correctly analyzed the result of my blunder in my head, so I guess it works out.

 

However, if black plays into the line 1. c5 Kd5 2. Kd3 Ke6 3. Ke4 and white doesn't incorrectly try to outflank him that way, white wins:

1. c5 Kd5 2. Kd3 Ke6 3. Ke4 Kf6 4. d5 cxd5+ 5. Kxd5 Ke7 6. Kc6 Kd8 7. Kb7

or 1. c5 Kd5 2. Kd3 Ke6 3. Ke4 Kf6 4. d5 Ke7 5. d6+ Ke6 6. Kf4 Kf6 7. d7 Ke7 8. Ke5 Kxd7 9. Kf6 Kd8 10. Ke6 Kc8 11. Kd6 Kb7 12. Kd7 Kb8

or 1. c5 Kd5 2. Kd3 Ke6 3. Ke4 Kf6 4. d5 Ke7 5. d6+ Ke6 6. Kf4 Kf6 7. d7 Ke7 8. Ke5 Kxd7 9. Kf6 Kc7 10. Ke7 Kb8 11. Kd7 Kb7 12. Kd6

or 1. c5 Kd5 2. Kd3 Ke6 3. Ke4 Kf6 4. d5 Ke7 5. d6+ Ke6 6. Kf4 Kd5 7. d7

 

That's the one I mentioned in my initial post about being a line where I thought white could win by a different method than the book line. But looking at it further, I realized black's mistake is 1. ... Kd5, where he's just going to get forced back anyway, and white will gain the opposition and win. I kept looking at that line at first when analyzing this first move. Then I realized black could draw by simply grabbing the opposition:

1. c5 Ke7! 2. Ke4 Ke6 3. d5+ (3. Kf4 Kf6, maintaining the opposition and not letting white make progress) cxd5+ 4. Kd4 Ke7 5. Ke5 (5. Kxd5? Kd7) d4 6. Kxd4 Kd8 7. Kc4 Kc7 8. Kb5 Kb7

or 6. c6 d3 7. c7 d2 8. c8=Q d1=Q 

 

So that brings us to the winning move: 1. Ke4, grabbing the opposition. I had to consider every possibility, so I wondered what would happen if black advanced his pawn instead of moving his king: 1. Ke4 c5 2. dxc5 Kd7 3. Kd5 Kc7 4. c6 Kc8 5. Kd6 Kd8 6. c7+ Kc8 7. c5 Kb7 8. Kd7 and white wins.

 

Then I considered 1. ... Kd6 by black, which is Pandolfini's "book" move, but I came up with a slightly different line than him. First, if black tries pushing his pawn on the second move:

1. Ke4 Kd6 2. Kf5 c5 3. d5 Kd7 4. Ke5 Ke7 5. d6+ Kd7 6. Kd5

or 1. Ke4 Kd6 2. Kf5 c5 3. d5 Ke7 4. Ke5 Kd7 5. d6 Kc6 (5. ... Kd8 6. Kd5) 6. Ke6

 

So if black doesn't push his pawn, my line goes:

or 1. Ke4 Kd6 2. Kf5 Ke7 3. Ke5 Kd7 4. Kf6 Kd6 5. c5+ Kd7 (5. ... Kd5 6. Ke7 Kxd4 7. Kd6) 6. Kf7 Kd8 7. Ke6 c7 8. Ke7 Kc8 9. Kd6 Kb7 10. Kd7

 

Pandolfini has black play 2. ... Kd7 instead of Ke7, which pretty much leads to the same thing:

1. Ke4 Kd6 2. Kf5 Kd7 3. Ke5 Ke7 4. c5 Kd7 5. Kf6 Kd8 6. Ke6 Kc7 7. Ke7 Kc8 8. Kd6 Kb7 9. Kd7

And then there's the 1. ... Kf6 lines that were mentioned by other posters that work out pretty similar. The moral of the story is that white wins by grabbing the opposition, but other moves draw if black defends well. This should be a familiar refrain to anyone who has studied endgames for more than 10 minutes in their life. Endgames must be played accurately, which is why I'm studying them so thoroughly.

 

As for Phobetor's suggestion of shifting the pieces down a row and black playing c4 at the right time, I'm not sure I understand. You mean like this?

1. Ke3 Kf5 2. Kf3 Ke5 3. Kg4 c4 4. d4+ Ke4

I don't think that helps black that much, because you skipped a white move in your post. You had black playing Ke4 then Kd3, without saying what white does in between. If white pushes the d pawn immediately, he still wins:

1. Ke3 Kf5 2. Kf3 Ke5 3. Kg4 c4 4. d4+ Ke4 5. d5 Kd3 6. d6 Kxc3 7. d7 Kb3 8. d8=Q c3 9. Qb6+

White just needs to stop black from advancing that pawn to c2, keep checking with the queen to get the black king to c2, blocking the pawn, and advance his king towards the pawn while the black king is blocking it. Eventually, white will win that pawn and the game. 

 

Taking the d pawn doesn't seem to help black, either:

1. Ke3 Kf5 2. Kf3 Ke5 3. Kg4 c4 4. d4+ Ke4 5. d5 Kxd5 6. Kf5 Kd6 7. Ke4 Kc5 8. Ke5 Kc6 9. Kd4 Kb5 10. Kd5

 

Or am I missing something?

 

--Fromper 

Talking about elementary endings and how important they are, yes you are missing something! At your last line, try 10... Kb6 11. Kxc4 Kc6

 

I actually missed it once myself in an over the board game. I had been defending a worse position (pawn down) for hours. Then at the end my opponent messed up, gave up a pawn and apparently gave me a won position. I had a similar position as above. I thought I was winning until he moved his king back from the pawn (like 10... Kb6 above), not resigning. Then I noticed why he didn't resign: because it's obviously a draw.

 

In a "one pawn stuck against one pawn"-ending you may think the one pushing the other king back wins, but it's always a draw. 

You're right. I missed that. The outflanking maneuver for head to head pawns only wins if the attacker's pawn is on the 5th rank. If it's further back, on the 2nd, 3rd, or 4th ranks, it's a draw. I've read that in books before (just looked it up and Pandolfini says it specifically), but that detail never sunk in. Again, this is why I'm going over the positions in that book over and over - to train myself to remember all the little details, so I can tell similar positions apart and know how to play them when they come up.

 

Thanks for the feedback and interesting conversation on this position. We need to get more endgame analysis going on around here. 

 

--Fromper 

Spike_Mason wrote: I would start out with Ke4. That forces Black's king back.

---

Just take the opposition!!!