Sunday, October 26, 2014

A short break from exams

Now if you're getting tired from all those mugging, take a short break and try out this composed problem:

Ahues, Herbert Siegfried Oskar
Die Schwalbe, 1999
1st HM
White to move and mate in 2

Have fun, and all the best for your remaining papers (:

Monday, October 20, 2014

A curious case of bishop vs knight

To let y'all prep for your upcoming exams, I will go on a hiatus for a couple of weeks after this post. I might still put up a puzzle or two to give you a break from studying, but other than that there'll be no articles during your exam period.

In our last challenge, the following position was given:

Deep Fritz 12 - Checkerboard 5
5 minute blitz
White to move, can he win?

There are several things that work in White's favour:

  • White is up by 2 pawns.
  • His kingside pawn majority gives him good chances to create a passed pawn on the f or g file.
  • There are pawns on both sides of the board. This favours the bishop over the knight; the bishop is able to exploit its long range abilities and switch attacks from one side of the board to another quickly. On the other hand, Black's short-range knight can only guard one side of the board and is slow in responding to threats on the other side.

Based on his pawn majority and superior bishop, one would naturally think that White should be able to create a passed pawn and win the game easily. However, there are also elements in the position that are more favourable for Black, and make life difficult for White:

  • All of Black's pawns are on light squares. This means that White is unable to utilize his dark-squared bishop to attack Black's pawns and tie down his king, leaving the Black king free to create potential blockades. Had the bishop been on a light square, White could decide the game very quickly by attacking Black's pawns and forcing them onto vulnerable squares.
  • Black has a very strong blockade on the queenside; the knight on c5 is untouchable. Moreover, Black's knight also denies White's king of the d2 and e3 squares, preventing him from making a quick incursion into Black's territory via e3.

That guy's tougher than you think

Although White can actually win this game, we realize that the task is not as simple as it looks. White must find a way to get past Black's blockade and create an unstoppable passed pawn.

But how? White has no way to force Black into zugzwang; Black can simply shuffle his king along the squares in the centre, discouraging White from advancing his pawns. An attempt to create a passed pawn is 1. gxh5 gxh5 2. Bg3, but after 2... Kf5! Black has blockades on all of White's pawns and the bishop cannot chase them away.

So White must find a way to force his king into Black's camp and attack his pawns. Since the e3 square-- the most direct route-- has already been fenced off by Black's knight, White must either penetrate via either the queenside or kingside. The first variation we look at is one where White attempts to enter via the queenside:

1. Kc2 Kd7

White wants to penetrate via b4 and a4, so the Black king must move over and stop him.

2. Kb3 Kc6 3. Ka4 Kb6 4. Kb4 Kc6 5. Bc1?! Kb6 (D)

Position after 5... Kb6

White has drawn Black's king away from the centre, so his pawns are free to strike on the kingside... 6. gxh5 gxh5 7. Bg5 Kc6 8. f4 Nd6!

Or can he? Black still has one more trick up his sleeve

9. Be7 Nf5! 10. Bf6 Kd7 11. c4 Ke6 (D)

Position after 11... Ke6

Here we see the power of the knight as a blockader. By playing 5. Bc1?!, White relinquished control over the d6 square and allowed Black to perform the maneuver Nd6-Nf5, blockading White's passed pawn and at the same time attacking two other pawns! With White's bishop under threat, he has no choice but to lose a pawn, e.g. 12. Bg5 Nxd4 13. c5 Kd7 and it would be difficult for White to make progress.

Now you might be wondering: Since 5. Bc1 allows the Nd6-Nf5 maneuver, how about playing 5. Bh2/5. Be5 instead (White must move his bishop anyway if he wants to exploit his passed f-pawn)? The answer is that it still won't work; now instead of d6, Black can use the e3 square: Ne3-Nf5 to achieve his objective.


So we've seen that exploiting the queenside doesn't work since it allows a strong knight blockade to be set up. The next variation-- and the main line-- we will be looking at is one where White crosses over the to kingside instead.

1. Be5! (D)

Position after 1. Be5

Black wants to keep his king in the centre, so White restricts the number of squares he can move to.

1... Ke7

1... Nxe5+? loses even faster; after 2. dxe5 Kxe5 3. gxh5 gxh5 4. Ke3 the resultant king-pawn endgame is an easy win for White

2. Ke2 Ke6

Of course 2... Nb2 leads to 3. Ke3 penetrating Black's position.

3. Kf2 Ke7
4. Kg3 Nd2

4... Ne3, trying to create the blockade on f5, fails now because after 5. Kf4 White has control over the blockading square. Thus we see the importance of the White king's presence to protect his pawns!

5. Kf4

White has achieved his objective.

5... Ke6
6. gxh5 gxh5
7. Bc7 (D)

Position after 7. Bc7

The bishop will eventually go back to chase away the knight, and White will either safely advance his f-pawn or win Black's h-pawn.


So going the queenside way failed, but the kingside worked. Why? The main reason is that Black's knight was a very powerful blockader with the threat of repositioning quickly to f5, thus White needed the presence of his king on the kingside to deny Black this opportunity.

Here are some of the things you can take away from this endgame:

  • Knights are very good blockaders. You have already seen in the game how Black's knight effectively immobilized the entire queenside, while controlling a large number of squares around it and attacking other pawns simultaneously.
  • One good trick to defending against enemy bishops in the endgame is to put your pawns on opposite colour to the opposing bishop, thus making it impossible for the bishop to attack them. In this game Black's pawns were well-placed, making White's winning task more difficult.
  • A well positioned bishop or knight in the centre can deny the enemy many crucial squares. Which was why 1. Be5 was played in the game.
  • When advancing pawns, use the king! Even though White was able to get a passed pawn in this game, he could only advance it safely when his king came over to help.

Oh, and just in case you're curious to what happened in the actual game... it ended in a draw. In mutual time trouble neither me nor Fritz were able to fully analyze all the variations, thus I escaped with a lucky draw from the silicon monster (:

With that, I shall go on my hiatus while y'all prepare for your exams. All the best, guys!

Wednesday, October 15, 2014

Challenge yourselves: A curious case of bishop vs knight

The following position arose after I was playing a particularly bad game with Fritz:

Deep Fritz 12 - Checkerboard 5
5 min Blitz
White to move

As you can see in this position, White is clearly better since he has two pawns up. However, in endgames featuring a struggle between the bishop and knight, there are always complications that make the resulting variations very unclear. In this case, Black has plenty of tricks up his sleeve that he can use to outsmart White and turn the game into a draw.

So the challenge is clear: If you were White, how would you go about winning this position? Or will it end up as a draw?

I will go through the analysis in the next article. In the meantime, have fun!

P.S. Some of you might already have seen this position in the Chess Teachers and Learners Group. If so, refrain from commenting and give others a chance!

Monday, October 13, 2014

Rook-Pawn endgames: Three vs Two (Part 2)

In the second part of our discussion I will stray off slightly from the main focus (pawns on one side of the board) and take a look at what may possibly go awry in a situation where there are outside pawns.

NN - Staunton
London 1842
White to move

Over here White is a pawn down, and there are pawns on both sides of the board. Now White faces a tough choice: Should he trade rooks?

I'll let y'all think about it for a while; assess the position and determine whether White should trade rooks, before scrolling down for the rest of the analysis.


And the answer is yes: White should trade rooks. While this might sound counterintuitive compared to what i have been harping on for the past couple of articles (that the weaker side should strive to keep the rooks on the board!), this is an exception. The presence of the kingside pawns means that a rook trade enables White to obtain an outside passed pawn, which he can use to distract the enemy king and force a draw:

1. Rxf6 gxf6
2. Kg4

As with all endgames, it is always important to centralize the king.

2... Kg7
3. Kf5 c5
4. h5 c4
5. h6+

White's outside passed pawn distracts Black and pulls his king to the side of the board.

5... Kxh6
6. Kxf6 b5
7. Ke5 (D)

Black's king is at the sidelines

White's king is more centralized; by the time Black's king crosses over the the Queenside, White would have been able to win at least one pawn and hold the draw.


That was simple enough, but as math and science peeps we'd prefer making life more difficult for everyone. So rather than just saying "Hooray, draw!" and ending here, many of us would ask the question: What if White were to reject the rook trade? Can he still defend his position successfully? To satisfy our curiosities, we shall take a look at what happens should White play something else; let's say 1. Rd2:

Position after 1. Rd2: Can White draw?

Once again, you can spend some time analyzing the position before scrolling down.


 My answer to the earlier question is: Yes, White should be able to draw, because I have not yet found a line where Black can win convincingly. In any case, with rooks on the board, things suddenly get complicated... and so much more fun to analyze.


After 1... Rf7 2. Rd8+ Kh7 3. Rd6 c5 4. Kg4 Rf2 5. Rd7 we have a position similar to the main line analysis.

2. Kg4 a4
3. Rd7!

Following the same principle of keeping the rook active by tying down the enemy king to the pawns. And now I will diverge into two lines: The main line  3... Rf7 and the variation 3... Rf2. Let us go through the Rf2 variation first:

3... Rf2

Black also keeps his rook active by attacking White's pawns, and this soon leads to a massive trade of pawns:

4. Rd6 Rxb2 5. Rxc6 Rb3 6. Rc8+ Kh7 7. Rb8 7... Rxa3 8. Rxb6 Rc3

After 8... Ra1 9. Ra6 a3 10. Kg3! a2 11. Kg2! g6 12. Ra5 (D) we arrive at the following position:

A familiar position

Does this position ring a bell? Yes, once again we see the power of Tarrasch's rule: White's rook is behind the enemy passed pawn, while Black's rook is tied down to defending the pawn and at the same time obstructs its promotion. So long as White's king stays on h2 and g2 (which is why 10. Kg3 and 11. Kg2 were played; to ensure Black doesn't try something nasty like ... Rg3+ which clears the way for his pawn!) and shuffles his rook on the a-file, the game is a dead draw. This means that even if Black wins the pawn on h4, he still cannot win since the g-pawn cannot chase White's king away from g2 or h2.

To refresh your memories with more details, here are the links to the relevant positions:

Returning to the position after 8... Rc3:

8... Rc3 9. Ra6

Same as always: Put your rook behind the pawn!

9... a3 10. Kf5 Rh3 11. Kg4 Rd3 (D)

Position after 11... Rd3

It should not be too hard to tell that Black cannot make progress in this position. Both White's king and rook work together to prevent Black's king from touching the h-pawn, while Black's rook is once again tied down to defending the a-pawn. After something like 12. Kf5 Rd4 13. Rxa3 Rxh4 14. Rg3 the a and h-pawns have been exchanged, and White's king will be in front of Black's remaining pawn... which means an impending Philidor Position.


So that was one variation cleared. Now let's return to the position after 3. Rd7 and proceed with our main line: 3... Rf7

3... Rf7 (D)

Position after 3... Rf7

4. Rd8+

This is the time when White should not trade rooks, since he can no longer get an outside passed pawn: 4. Rxf7? Kxf7 5. Kg5 b5 6. h5 c5 7. Kf5 b4 and the series of pawn trades on the kingside will give Black the outside passed pawn instead.

4... Kh7
5. Rd6!

Once again we see the power of keeping your rook active.

5... Rc7

This is a passive move, but we have already seen that the more active variation 5... Rf2 leads to a draw.

6. Kf5

Preparing to march over to the kingside to assist the White rook. 

7... b5

6... Rc8 7. Ke5 g6 8. Ke6 Kh6 9. Kd7 and White still wins the pawn

7. Ke5 Kg8 (D)

Position after 7... Kg8

White might be a pawn down, but can hold out thanks to his active rook and more centralized king.

8. Kd4 c5+
9. Kd5 c4
10. Rb6 c3
11. bxc3 Rxc3
12. Rxb5 Rxa3
13. Rb4

Or 13. Ra5 Rd3+ 14. Kc5 (not 14. Kc4?! Rh3! with some problems for White) preparing to win the a-pawn.

13... Rd3+
14. Ke4 Rh3
15. Kf5 a3
16. Ra4 Kh7
17. Ra6 Rc3
18. Kg4 (D)

Position after 18. Kg4

And we have reached a similar position as the one in our first variation. Since Black cannot make any progress, the game can be agreed drawn.


Thus, we have seen that despite the complications of having extra pawns at the side, a Three vs Two situation in rook endgames can still be a draw as long as the defender remembers the correct drawing technique. As always, active play is key!

P.S. In the actual game, White refused the rook trade, and then proceeded to lose the game after making a couple of mistakes.

Two vs One:
Three vs Two (Part 1):

"Silman's Complete Endgame Course" by Jeremy Silman

Saturday, October 4, 2014

Rook-Pawn endgames: Three vs Two (Part 1)

As we have seen earlier with Two vs One scenarios, rook endgames with the pawns on one side of the board is usually a draw as long as the weaker side can keep his rook active, trade pawns, and simplify the position to a Philidor/Inverse Philidor. When we extrapolate to a Three vs Two, we can see that this should be a draw too so long as the defender plays accurately according to the guidelines in our earlier post (

However, if the weaker side plays too passively then sometimes he might end up in a situation where he ends up getting "squeezed" by his opponent's pawns:

White to move wins

In this position White's pawns eat up most of the space and squeeze Black's king into a fatally cramped position; such a troubling position is often caused by passive play from the defender.

1. Kh5

White wants to trap Black's king on the edge of the board, where he will slowly be crushed by the weight of White's pawns and the threat of back-rank mate. 1. Kh5 prevents ... Kg6 after the White pawn advances to f6.

1... Rd8

The only defence against back-rank mate; 1... Rb7? 2. f6+ Kf8 3. Ra8+ leads to mate on the next move.

2. f6+ Kg8

2... Kf8 3. Kh6 (threatening the h-pawn) Kg8 leads to a similar position as the main line.

3. Kh6 Rb8

There is no better move; taking the rook off the back rank allows mate after Ra8.

4. Ra7

The threat is 5. e6 fxe6 6. Rxh7, where White's pawns should easily crush the Black king to death (just don't stalemate!).

4... Re8?

An attempt to stop White's threat. Unfortunately, it fails after:

5. e6!

5... Rxe6 6. Ra8 leads to mate, while 5... fxe6 6. f7 wins the rook.


The following is another example of how things can go bad for the weaker side if he gets squeezed:

Black to move wins

Here, it is obvious that Black enjoys a lot of space in the centre. How should he convert this advantage into a win? The plan lies in two parts:

  1. Get the king and rook out of the way, and reduce White's checking distance so as to avoid any annoying back-rank checks. Once the king has been evacuated from the c5 square, Black can proceed to push his pawn to c5 and create the "squeeze" we have seen in our previous example.
  2. Place the rook on the 7th rank and exploit the pin of White's c-pawn to his king by playing d4-d3.

1... Re3

It is important to occupy the e-file: In some lines where Black plays ... Kc3, he can defend from a check on the 6th rank by returning the Black rook to e3.

2. Rc8

After 2. Rb7 Re2 3. Kb1 Kc3 White is pushed into a fatal corner, while after 2. Rf8 Re2 3. Rf3 Kb5 4. Kb3 c4+ Black has also achieved the "squeeze" position as we have seen in our first example.

2... Re1!

This move forces White to play 3. Rc7/Rc6, reducing checking distance and spoiling any potential back-rank checks when the Black king moves off the c5 square.

3. Rc7

We have already seen that 3. Rg8 Re2 4. Rg3 Kb5 is bad for White, while 3. a3 (trying to exchange pawns, but Black already has too much space in the centre) 3... bxa3+ 4. Kxa3 Rb1 traps the White king, with the upcoming threat of ... Rb5 followed by ... Kc3

3... Re8!

Taking control of the last rank: Now White's rook cannot use the last rank to make any back-rank checks once the Black king has left c5!

4. Rc6

Once again 4. Rg7 fails to 4... Re2 5. Rg3 Kb5

4... Kb5
5. Rc7 Kb6

Preventing Rb7+ after ... c4

6. Rg7 c4

The "squeeze": Black's pawns have reached their ideal positions.

7. Rg6+

An attempt at long-distance checks, but now Black has a safe hiding square on a4. The other variations are:

7. Rd7 d3 8. cxd3 Re2+ 9. Kb1 c3 where Black's pawn is too far advanced, and winning is only a matter of technique and time-management.

7. a3 Re2 (7...bxa3+ also wins but not as quickly: 8. Kxa3 Re2 9. Rg6+ Kc5 10. Rg5+ Kd6 11. Rg4 Kd5 and Black should win the resulting two vs one endgame since his opponent's king is too far away) 8. axb4 d3! (exploiting the pin along the 7th rank) 9. Rg6+ Kb5 10. Ka3 Rxc2 11. Rg5+ Kc6 12. b5+ Kd6 and Black's pawn will queen first.

7... Kb5
8. Rg5+ Ka4 (D)

Looks familiar?

We have reached a similar position as that of the first example, which is an easy win. Now I will leave the rest of the game to you as a personal challenge: Find the fastest way to finish off White!


So it should be clear to us: There is no way the defender can allow such a terrible position to happen, which is why he has to keep his pieces and pawns as active as possible in such endgames.

Black to move

Here, Black must not allow White to push his pawn to f4, else it will create the dreaded "squeeze" position. With this in mind, the candidate move is not difficult to spot:

1... g5!

Locking White's pawn majority. Now, any pawn advance by White helps simplify the position into an easily drawn 2 vs 1 position.

2. Rd5

Hoping for a miracle after 2... Rxd5?? 3. exd5 Kf7 4. f4 where the outside passed pawn proves decisive for White. But Black is not so compliant as to trade rooks:

2... Rb3

Black simply follows the same guidelines as we have seen in our earlier article: Keep the rook as active as possible by tying down the enemy king to its pawns.

3. e5

Repeated checks with 3. Rd7+ Kg6 don't make any progress, while after 3. Kf2 Rb2+ 4. Ke3 Rb3+ White's king cannot hope to get closer to the rook without hanging his pawn.

3... fxe5
4. Rxe5 Kf6
5. Ra5 Kg6 (D)

A dead draw

With the Black king in front of the enemy pawns, any additional pawn trade will simplify the position into a draw Philidor/Inverse Philidor.


In the tussle among rooks and pawns, the goal of the stronger side is always to try and promote his extra pawn. To do this, there are two different tricks:

  • Entice the weaker side into trading rooks so as to create an outside passed pawn from the recapture, which we know by now is a valuable winning asset.
  • Advance the pawns to gain as much space as possible, so as to "squeeze" the opponent. When that happens, the combination of the pawns and the threat of back-rank mate should slowly crush the enemy king to death.

Not a pretty sight either (Image from Wiki)

As for the weaker side, one must make sure never to fall into the traps as stated above! To do this, we just needs the same set of guidelines (and I state "guidelines", not "rules"!) that I have provided in our earlier article:

  • Avoid trading rooks
  • Put the king in front of the enemy pawns
  • Keep the rook as active as possible (i.e. tying down the enemy king and pawns)
  • Seek pawn exchanges that help simplify the position
  • Additionally, use appropriately-timed pawn pushes to deny the stronger side any chance to grab space in the centre with his own pawns.

With that in mind, the defender should be able to avoid the terrible "squeeze" positions that we have seen earlier, and slowly simplify the position into a drawn Philidor/Inverse Philidor. Keep in mind: To defend well, active play is needed!

With that I stop here, and will proceed with a more complicated example in Part 2.

Two vs One:

"Silman's Complete Endgame Course" by Jeremy Silman