Tuesday, February 18, 2014

Rook + 5th rank pawn vs Rook

Yesterday we looked at the challenge I issued on Friday (Rook + 5th rank pawn vs Rook) and the methods which both sides use to force a win or a draw.

White to move, can he win?

To repeat again, these are the goals which both sides hope to achieve:

  • Black wants to get his king in front of the enemy pawn, thus making it possible for him to create a Philidor Position and drawing the game
  • White wants to prevent Black's king from getting in front of his pawn, by using his rook to cut the enemy king away from the battle. This makes it easier for him to create a Lucena Position, which will win for White.

(Once again if you still have no idea what a Lucena/Philidor Position is, or need revision on basics, I'll place this here for convenience: http://nushsblackknights.blogspot.sg/2013/12/revising-basic-rook-endgames-part-1.html)

Though it may seem tedious to calculate all the variations, there is fortunately for us a simple guideline when it comes to combating a pawn on the 5th rank: The stronger side wins only if the weaker side's king is cut off by one or more files on the long side of the board, or two or more files on the short side of the board. This guideline is mostly true for central, bishop and knight pawns on the 5th rank (save a few exceptions we will look at later); as usual, rook pawns form exceptions and we'll discuss them in the last part of this article.

If you're confused over the term "long side of the board", the following diagram will help to clarify:

Yellow: Long side. Green: Short side

The long side of the board (the yellow squares in the diagram above, from the a to d-files) is the side of the board with the greater number of squares between the pawn and the side of the board. Conversely, the short side of the board (green squares, f to h-files) is the side with the smaller number of squares. So if the pawn were on f5 rather than e5, the long side of the board will be the a to e-files, while the short side will be the g and h-files.

In such positions where the weaker side's king has been cut off by the enemy pawn, he should strive to place the king on the short side of the board; the reason being that his rook wants to rain checks upon the enemy king from the longer side of the board (remember that rooks are strongest when far away from the opponent's king!) and of course the king should not get into the way.

Part 1: The long side of the board

Let's return to the position we saw earlier and see how this works out:

White to move, can he win?

And now it should be quite clear that White can win, because Black's king is cut off on the long side of the board. Thus Black cannot use the rook-checking trick to force a draw, because his king gets in the way and the short side of the board is-- well, as its name suggests-- too short for the rook to check safely.

The outcome is still the same whichever side it is to move, so for clarity's sake we'll give White the move here:

1. Kf5

Exploiting the short side of the board, and preparing the advance the pawn

1... Rf8+

If Black tries to check on the short side of the board with 1... Rh8, then after 2. e6 Rh5+ 3. Kg6 Black can no longer check safely, and soon falls to 3... Rh2 4. e7 Rg2+ 5. Kf7 Rf2+ 6. Ke8 with a Lucena Position in White's favour.

The attempt to check on the long side of the board also fails because Black's king and White's pawn get in the way: 1... Ra8 2. e6 Ra5+ 3. Kf6

2. Ke6 Re8+
3. Kf6 Rf8+
4. Ke7

Chasing away Black's rook and allowing the pawn to advance again

4... Rh8

Once again we see that after 5... Rh7+ 6. Kf6 Rh6+ 7. Kf7 Rh7+ 8. Kg6 Black's rook can no longer keep up the rain of checks

5. e6 Rh2
6. Kf7 Rf2+
7. Ke8 Kc7

The variation 7... Rh2 8. e7 Rh8+ 9. Kf7 Rh7+ 10. Kf6 Rh6+ 11. Kg7 also loses for Black. Hopefully by now you are familiar with how the Black rook's checking ability is crippled when on the short side of the board.

8. e7

Reaching a Lucena Position, with a win for White.


So that looked simple enough. But now let's see how even the smallest change to the initial position can make things very different!

Can White still win here?

Black's king still appears to be cut off by the enemy rook, but in fact he is not because Black has a special move at his disposal: Rd8, where after trading rooks Black's king forces its way to the front of the pawn!

In such a position the outcome of the game is decided by not whether a Lucena/Philidor Position can be reached (because the rooks have already been traded!) but rather which side has the opposition. I leave the remainder of the analysis as a very simple challenge to y'all: Can White win if he has the move, and will the outcome be different if Black had the move instead?

Part 2: The short side of the board

So how does the weaker side draw when his king cut off by only one file on the short side of the board? Let's take a look:

Draw regardless of who to move!

So let's say it's White to move, and he tries to perform the same trick of using the king to help advance the pawn. Since the close proximity of the Black king prevents White from using the f-file, the game may continue along the lines of:

1. Kd5 Ra8

And now Black threatens a barrage of checks from the long side of the board. This time, unlike our first example, White cannot run away from the checks!

Attempting to force the king away with 1. Rg1+ leads to 1... Kf7, allowing the Black king to get in front of the pawn and paving the way for a draw. Notice that if the Black king were on g5 rather than g6, then 1. Rg1 will have very different results altogether... which we will see later on. Lets concentrate on the current position first.

2. e6

Once again, attempting to block the checks with 2. Rc1 fails to 2... Kf7

2... Ra5+

There is no way White can run away from the checks now because Black's rook is sufficiently far away. So for example the game can continue something like this:

3. Kd6 Ra6+
4. Kd7 Ra7+
5. Kc8 Re7
6. Re1 Kf6

Piling up upon the e6 pawn and winning it, thus simplifying to a rook vs rook draw.


So as I mentioned earlier, why will things turn out differently if the king were on g5 rather than g6? Let's take a look:

White to move wins

If it were Black to move he will just play 1... Kg6 turning the position into the drawn game we saw earlier. But if we give White the move:

1. Rg1+!

And now Black's king cannot step in front of the pawn because the White pawn covers the f5 square.

2... Kh5
2. Kf5

With Black's king cut off by two files, White can safely step onto the f-file where it is immune to checks from the long side of the board. The rest of the game is easy:

2... Rf8+
3. Ke6 Re8+
4. Kd6 Rd8+
5. Ke7 Rd2
6. e6 Re2
7. Kf7 Rf2+
8. Ke8 Re2
9. e7 Rd2

Creating a winning Lucena-like Position for White. Note that it isn't exactly a Lucena Position because White's king is on h5 and prevents 10. Rg4, but after 10. Rf1 Kg5 11. Kf7 Black can no longer check safely and the pawn is free to promote.

Part 3: Rook pawns

As I have stated earlier rook pawns always form special exceptions to general endgame rules, and so we'll see how different they are in this position.

No win for White over here!

A quick look at the position should be enough to tell that this is a dead draw: With the pawn huddled in a corner White can never hope to step forward to support the pawn. Both 1. Rc1+ and 1. Kb4 (attempting to step out to help the pawn) fail to 1... Kb7 where Black's king gets in front of the pawn with a draw. And with rook pawns, there is no other side (no short side this time!) for the White king to step onto.

If you fancy another challenge, you can work out how many files does the Black king need to be cut off by in order for the game to be a win for White. (Perhaps this link to one of our earlier topics may help: http://nushsblackknights.blogspot.sg/2013/12/a-lucena-with-rook-pawn.html)


Was that too many examples for y'all to absorb? If so, then let us summarize things to make our understanding easier:

In this rook + pawn vs rook position where the pawn is on the 5th rank, the stronger side can win if:

  • The weaker side's king is cut off by at least 1 file along the long side of the board. This prevents the weaker side's rook from forcing a draw through repeated checks, as the short side of the board is not long enough to do this safely and the king gets in the way on the long side of the board
  • The weaker side's king is cut off by at least two files along the short side of the board. If the weaker side's king is too close, then the stronger king is forced to step onto the long side of the board and expose himself to enemy rook checks.
  • If the weaker side can get his king in front of the stronger side's pawn, then his chances of drawing becomes much higher.

This can be summed up into the simple guide which holds true for most cases: A win is more likely with the weaker king on the long side, while a draw is more likely on the short side (or with the weaker king in front of the pawn). Of course, we must note down the important exceptions to this rule, as we have investigated in the examples earlier.

Steady on, my brothers. The journey will get even tougher as we progress deeper into the mysteries of rook endgames.

"Silman's Complete Endgame Course" by Jeremy Silman

No comments:

Post a Comment