Tuesday, October 08, 2013

Important Q&P endgame improvement

White to move. Can white win this? How should white proceed?

Source: ChessToday.net


Yancey Ward said...

I don't have the time this ending will require, but I have observations.

White isn't going to queen that d-pawn- that much is sure. If white exchanges queens at e7 immediately, the resulting K+2P vs K+2P is actually losing for white because his king is out of position. So, if white can win this, I think it has to start with h4. The idea would be to advance the pawns on the kingside far enough to make the threat of Qe7 dangerous to black (change the nature of the resulting king and pawn ending). I can't really say right now what the best defense for black might be, but if he allows the pawns on the king side to lock up on h4/g5 for white and h5/g6 for black, then the exchange at e7 is going to lead to a lost king and pawn ending. Black will have to find some sort of defense using his queen to harass white, I think.

Lorfa said...

h4 actually wins after much resistance.

Tatiana Kosintseva marry me!!! <3

Anonymous said...

1. Kc5!! puts black in Zugzwang (If Q moves that leave e7 undefended say 1.... Qa8 2. Qe7#
1....Ke5?? 2. Qe7+ Qxe7 3. dxe7 promoting to a queen next move,
1...Kf6 2. d7 and the pawn can't be prevented from promoting to a queen)

Yancey Ward said...

The problem with 1.Kc5 is that black just pins the pawn with Qf8. White can try to unpin it with 2.Kc6, but then black checks from f3. After that, I don't really see how white can both evade the checks and not suffer a double attack on d6 losing the extra pawn for no gain.

Lucymarie said...

White has her King and Queen well placed, and the question arises: can the queens be profitably exchanged on e7-square yet? Well, not yet, since Black king could pick off White pawns while White was picking off Black pawns. So White must advance kingside pawns first, starting with 1.h4 before trading queens on e7.

Lucymarie said...

By the way, if White plays 1. Kc5???, as Anoymous suggests, then Black will play 1. .. Qf8 forcing the Queen trade in the premature circumstances that I previously mentioned.

mesi said...

In answer to Anonymous,
after 1. Kc5, then
1....Qf8 prevents d7.

mesilah said...

In answer to Anonymous,
after 1. Kc5, then
1. ...Qf8 prevents
2. d7.

Yancey Ward said...

Like most queen and pawn endings, this is devilishly difficult. One obvious try is to take at g7. I dismissed this idea yesterday without putting much thought into it because it eventually drops the d-pawn and leaves the white king to fend for himself against the black king and queen:

1. Qg7 Qb8!

Probably required. Taking at d6 with the king just loses to Qxh6 with check, and if the king goes to the e-file, white checks from e3 and exchanges queens for an easy win, and if the king goes to d7, white checks from c6 and exchanges queens, and if to d5, white checks from d3 followed by a check from the e-file to win. More testing for black is the check from d8: [1. ...Qd8?! 2.Kb5! Kd5! 3.Qh6 and now, can white get the king some place where he isn't checked? I don't know for sure, but I like white chances here.] Continuing from 1. ...Qb8+ above:

2. Qb7

Of course, 2.Kc5/c6/a6 allows Qxd6 with check and an almost sure draw, and 2.Ka5 isn't going to allow the white king to evade the checks and keep the d-pawn at the same time. Best attempt, I believe is to just concede the d-pawn and get the queen back to help out the king and hope black goes wrong, but.....

2. .....Qd6
3. Qc6

White can't avoid the exchange:

3. .....Qc6
4. Kc6 Kf6 (or Ke5)
5. h4

Else, black plays Kg5 and will win at h4 and white can't win at h6 without also dropping the g4 pawn. Continuing:

5. .....Ke5 with an easy draw.

In my next comment, I will try to put some flesh on the idea I offered yesterday- 1.h4 which is an attempt to create a real threat of Qe7+ followed by an exchange down to a winning king and pawn ending for white.

Yancey Ward said...

Yesterday, I envisioned a position where the king side pawns had locked themselves up on (black)-h5 and g6, and (white)-h4 and g5. If such a position arose with the d-pawn, queens, and kings on the same squares as in the starting position with white to move, then Qe7+ wins for white quite handily- the resulting K+2P vs K+2P ending is decisive for white. To try to bring this about requires white to advance 1.h4 as a start:

1. h4 g6 (alternatives later)
2. g5 hg5

Now, here, if black plays h5, white checks from e7 to get the winning king and pawn ending I described above: [2. ...h5 3.Qe7 Qe7 4.de7 Ke7 (or Ke7 or Kf7 won't save after d8Q) 5.Kc7 is won for white]. Continuing:

3. hg5

That is the easy part, the rest is very difficult and unclear. Now, there is no threat of Qe7 since the exchange at e7 wins for black- in this line there is no h4 pawn protecting g5 and black gains the time to win at g5 while protecting g6 should white exchange the queens. However, black still needs to find an appropriate defense. Let's take a "quick" look at the options:

3. .....Qa4 (still eyeing d7)
4. Qc6

Threatening d7 again:

4. .....Qb4
5. Kc7 Qf4 (pinning the pawn)
6. Qe8 Kd5 (Kf5 7.Qf7+-)
7. Qa8 Ke5 (Ke6 8.Qa2)
8. Qa5 Ke6 (Ke4 9.Qb4+-)
9. Qa2 Ke5 (Kf5 10.Qf7+-)

Now black must find a perpetual or some sort of double attack to win the d7 pawn. He has four checking moves- three uncovered checks on Kf5, Ke4, and Kd4, and Qc1:

10. ....Kf5
11.Kc6 Qc1 (Qe4/f3 12.Qd5!+-)
12.Kd6 Qf4 (Qd1 13.Qd5;Qg5 13.Qa5)
13.Ke7 Qg5 (Qb4/e4 14.Kf7)
14.Ke8! Qe3 (what else now?)
15.Kf7 Qd3 (Qb6 16.Qd5+-)

Now, all the way back at move 3, black could try Qh8 as a way to get at the white king from b2:

3. .....Qh8
4. Qe7 Kd5 (Kf5 5.Qf6+-)
5. Kc7 Qc3
6. Kd8 Qh8 (Qa5 7.Ke8+-)
7. Kd7

And now black is in a kind zugzwang. If black puts the king on the c-file, white plays Qf6: [7. ...Kc4 8.Qa8 Qe6 should win]. Also [7. ...Qh3 8.Ke8 Qh8 9.Kf7 Qh7 10.Kf6 Qh8 11.Kg6+-]; or [7. ...Qg8 8.Kc7 Qa8 9.Qf7 Ke5 10.Qf6 should win]. I can't really be 100% sure this won after 7.Kd7, but I don't see a ready defense for black, and after the g6 pawn is dropped, I can be sure those lines are losing because I verified them on the Nalimov Tablebase.

Most of the other defenses after white's third move just look like variations on a theme, and look to eventually lose. In my next comment, I will finish up by looking at black's first move alternatives.

Yancey Ward said...

In my previous comment, I addressed the following line:

1. h4 g6
2. g5 hg5
3. hg5, and I think I showed that white has excellent winning chances (the position may even be decisive since I couldn't find a really clear defense for black). Now the question is, can black improve on that first move? Let's look at the options:

1. h4 g5
2. h5

Here, I don't think the exchange at e7 is a threat to black with this pawn structure. The resulting ending looks drawn to me since the white king can't allow the black king to reach e5 and f4 to attack the backward g-pawn, and the kings just look at each other from the c and e files as a result. So, I think the critical moves for black here are Qh8 and Qa4:

2. .....Qh8
3. Qe7 Kd5
4. Kc7 Qc3
5. Kd8

Now, you tell me- is this better or worse for black than the analogous lines in the previous comment with just two king side pawns at g5 and g6? I don't see how it is better for black.


2. .....Qa4
3. Qc6 Qd4
4. Kc7 Qa7
5. Kc8 and how is this better for black?

Back to move 1:

1. h4 h5
2. gh5 Qf8
3. Qc4 Ke5
4. Qc3 Ke6
5. Qh3 Ke5
6. Qe3 Kd5
7. Qd3 Ke5
8. Kc7 looks winning to my eye. Of course, I am leaving out clearly losing lines in the moves above, to save some time this afternoon.

Finally, at move 1:

1. h4 Qa4
2. Qc6 Qd4
3. Kc7 Kf7 (Qa7 4.Kc8)
4. Kc8 looks winning to me, too.

I could spend another day on this ending, it is so complex, but I feel strongly that white wins after 1.h4.