Category Archives: Endgames

Queen Against Pawn

Last time we looked at a pawn ending played between two young players (about 1500-1600 strength) at Richmond Junior Chess Club. After various misadventures, during which Black miscalculated badly in a position where he had a simple win, this position was reached, with White to play.

Before we continue looking at the game, some basic endgame knowledge. Everyone needs to know the ending with queen against pawn on the 7th rank supported by the king. If the pawn’s on a centre file or knight’s file the queen wins. You force the king onto the queening square and advance your king. Against a bishop’s pawn or a rook’s pawn, though, it’s a draw unless your king’s close enough to take a hand in a checkmate. With a bishop’s pawn, the defender can move his king into the corner so that taking the pawn will result in a stalemate. Likewise, with a rook’s pawn, the king in the corner will be stalemated.

Another piece of basic knowledge is that you can stop a pawn on the 7th rank easily if you can put your queen on the promotion square. All you have to do then is approach the pawn with your king.

Bearing that in mind, let’s see what happened in the game, with White to play his 60th move.

Black has the potentially drawing c-pawn, and two others as well, but his king is on d3 rather than d2. White has several ways to bring home the full point. A nice winning move is 60. Qh3+, when Kd2 walks into 61. Qe3+ Kd1 62. Qe1#, while moving back to, say, c4 allows Qe3, controlling the queening square. White can then follow up with Qc1 and just take all the black pawns. A similar idea is 60. Qh6, again followed by Qc1. But instead the game continued:

60. Qd8+ Kc3 61. Qxf6+

In some lines White might want to keep the f-pawn on the board to prevent the stalemate defence, but after this White’s still winning.

61… Kd3 62. Qf3+ Kd2

Allowing an immediate mate, but otherwise the king will be cut off on the fourth rank.

63. Qe2+(?)

Missing the mate in 2: 63. Qe3+ Kd1 64. Qe1#. White’s still winning at the moment, though.

63…Kc1 64. Kxg2?

This is the move that throws away the win. It’s not so easy at this level, but the winning idea was 64. Qb5 (avoiding the stalemate defence) Kd1 65. Qb3 Kc1 66. Kxg2 Kd2 67. Qb2 Kd1 68. Kf2 c1Q 69. Qe2#.

64… Kb1 65. Qd3 Kc1?

Now White’s winning again. Instead, Ka1 was drawing.

66. Kf2?

It looks natural to move the king in but now Black has the chance to revert to the stalemate defence. Again, the win was to be achieved by occupying the b-file. For example: 66. Qb3 Kd2 67. Qb2 Kd1 68. Kf2 Kd2 69. Qd4+ Kc1 70. Qb4 Kd1 71. Qe1#.

66…Kb2 67. Qd2 Kb1 68. Qb4+ Kc1?

The final mistake. Black still had a draw by moving to the a-file.

69. Ke3

White had to be careful: Ke2 and Ke1 were both stalemate. There was another mate in two, though: 69. Kf1 Kd1 70. Qe1#.

69…Kd1 70. Qd2#

Once more, then, a lot to learn from this game. These endings with pawn on the 7th rank against queen are so important and essential for understanding many pawn endings. As I tell all my students, you can’t understand other endings until you understand pawn endings, you can’t understand middle games until you understand endings, and you can’t understand openings until you understand middle games.

For the record, here’s the complete game.

Richard James


Amatuer Versus Master: Game Six

My opponent in this game is from Russia (I think Siberia) and is the second highest rated player in this section. At the time that I am writing this, Norchenko has one win and four draws, including the one with me, and is in second place in this section. I am in third place with five draws and one loss. So far, there are only four wins and thus four losses in this section. The remaining 13 concluded games are all draws.  I do believe that the ultimate winner of this section will be whoever gets a plus score. The top two places in this section advance to the next round.

When this game started I decided to play the White side of the Sicilian Defense because I wanted to try the Smith-Morra Gambit on him. I almost never play the White side of the Sicilian Defense in a rated game, but I did this time. I messed up the move order and decided not to play the gambit because the move order that I played favored Black. After I made this decision updates to my database showed that I could have played the Smith-Morra Gambit and been OK.

Black’s fourth move surprised me a little, as did many of his moves afterwards. I had never seen this line or variation in any other game that I have played before or after this one. Fortunately, most of what he played was in my database. When he varied from my database I was able to figure out good enough moves to hold the draw.

From move number 19 on we were out of my database. On move number 26 I played what the chess engines considered to be a second-best move. The “better” line would still have been even and thus I would still end up with a draw. I played what I thought was the more impressive or cuter line.

I believe that this is the highest rated player that I have drawn on ICCF.

Mike Serovey


Sometimes it is Better to be Lucky Than Good

This game was my last game in this section to finish. My opponent is from England. My opponent kept declining my draw offers because he thought that he had a better pawn structure.

I was, once again, mislead by the chess engines into playing an inferior line and could have lost the endgame if my opponent found the winning idea on move number 41. Instead, he moved his King in the wrong direction and then agreed to a draw.

I ended up with an even score in this section which netted me third place. Although I have won several Walter Muir sections, and these are played on the ICCF server, this third place finish is my best result so far in an international section. The Walter Muir sections are for players in the USA only and I am not allowed to use chess engines in those events.

On move number 6 White captures on c6. This gets me out of what I wanted to play, but I usually do OK with it as Black.

Although White grabs some space in the Center with his pawns on e5 and f4, he leaves his King a bit naked. I was never able to take advantage of that, though.

On move number 15 both players still have their kings in the Center and neither one can castle. I never did get to castle my King.

On move number 27 I pinned White’s Bishop to his King. After some fancy moves we traded off some minor pieces and rooks, but I never got an advantage out of it. On move number 28 I got convinced by Houdini 3 that the line that I played was better than the one that I wanted to play. I now think that the other line that I rejected was better.

On move  number 36 I was up a doubled pawn. I also had two passed pawns. Even so, I was unable to win.

Mike Serovey


Amateur Versus Master – Game Five

This game is from the final round of the 2011 Golden Knights Correspondence Chess Championship. Because John and I agreed to email our moves to each other instead of using snail mail, this game finished well ahead of the other ones in this section. The rest of my games in this section are still in the openings or are transitioning to the middle games.

John is the lowest rated opponent that I have in this section. I lost playing the Black side of the Benko Gambit. This loss, combined with a few other ones, has convinced me to stop playing the Benko Gambit in correspondence chess. I used to win whenever my opponent fully accepted my gambit. Lately, I have been losing whenever White shoves that passed pawn down my a file!

I am the only non master in this section. Therefore, I have no delusions of grandeur about winning this section. I am simply trying to get an even score and this loss will not help me any.

On move number 6 I decided to change up my usual move order because I was hoping to confuse my opponent and thus gain a psychological advantage. This almost worked. John did get confused a little, but I lost anyway.

Whenever White allows Black to capture the Bishop that is on f1 White gives up the right to castle. This is where Black gets his compensation for the sacrificed pawn. I am no longer able to keep my advantage in this variation.

Black completes his basic development on move number 11 and then White begins his assault by moving that passed a pawn down my throat. I still need to find Black’s best reply to that.

By move  number 14 Black is  bringing his rooks and knights over to the Queenside to launch his counter attack. White is going to break open the Center.

On move  number 15 White anchors a Knight on b5 and this Knight creates problems for me for quite a while afterwards.

Looking back at move number 16, I now doubt that trading my fianchettoed Bishop on c3 was the best move for Black. Allowing White to get a pawn on c4 created many problems for me. From move number 21 on Black is losing.

Mike Serovey


What To Do About Gambits

Speedy development is often worth the investment of a pawn in the opening. Examples include the Smith-Morra Gambit of the Sicilian 1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 when White has a very promising initiative for the pawn that often brings dividends. Devotees of this line can become highly attuned to its nuances. If that is the case you have to ask yourself as Black whether taking them on in their most familiar territory is the most intelligent thing to do. You might decide it is better to avoid it than to try to refute it. Even if you like spending many an hour with opening books, there is no substitute for hours of practice playing the line over and over again – which White will of course be doing. Perhaps Smith-Morra Gamiters’ would find the Caro-Kann or the French Defence, or something else, really annoying. If so, play that against them! It is wise to get your opponent out of their familiar territory.

This is the sort of thing that can be considered if you know your opponent and you’re playing them in an over-the-board game. Of course, if someone plays a gambit against you in a correspondence game and you are allowed to use software for help, then that is a different matter. For example, silicon monsters nowadays are less impressed with the Smith-Morra than we humans are. Below is one of my own correspondence games against a line of the Smith-Morra that I would have found difficult to play against over-the-board. But, with assistance from HIARCS, I found it easier to deal with. It takes a long time, but eventually White’s initiative dissipates and then it is all about whether Black can convert the ending. The knight and pawns ending was particularly pleasant to play for Black. If you would like to play some correspondence chess online, try FICGS – The Free Internet Correspondence Games Server.

Angus James 


A Scrappy Example of Psychology and Luck in an Ending

In the sixth (and final) round of the Pittsburgh Chess Club Championship, I played one of those unfortunately scrappy games I have been playing recently. From an easily winning position, I carelessly threw away the win to reach an ending that (to me) was obviously a draw. However, I kept playing for a win, hoping for a swindle, aided by the fact that my opponent had very little time on the clock and appeared to have spent a lot of energy earlier, and now appeared to be still nervous (indicating that he was not certain, unlike me, that the ending was a dead draw). Our subsequent play was sloppy, to say the least, but I got the win (aided by my incessant blitzing that left him in fact losing on time in the final position), and ended up just catching the leader to tie for first place in the tournament, to become one of two 2014 Pittsburgh Chess Club co-champions. I am happy that I achieved this, but know full well that I got there with a lot of luck in all the rounds that I aim to replace in the future with new skill (for example, every ending that I didn’t do right, I have studied after the fact).

Flaws aside, I think it’s useful to see how, amidst imperfect play, having a possible swindling winning idea is useful, because with luck it might actually work out. We are human beings, not computers, so there will always be some luck involved in human chess. I’d like to think there is a little bit of skill in pursuing a swindling idea, latching onto interesting aspects of a position and trying to make use of them.

In the sport of chess you have to do what you can even after misplaying an earlier part of the game. The swindle involved making moves that were risky or had obvious (to me) defenses, but part of the art of swindling involves trying to guess that your opponent might not see what you see and setting possible traps.

The ending

We reached a position with equal material: Two Rooks and one Bishop and four Pawns on each side. As White, I had a single b-Pawn and three King side Pawns. Black had an a-Pawn and b-Pawn but a fragmented King side with an f-Pawn and h-Pawn. So I concentrated on hoping to make something of Black’s weak King side before Black’s Queen side majority became a factor.

So one observation I immediately made was that perhaps I could make progress by getting my f-Pawn to f6 to make Black’s King inactive, and also to semi-trap it and bring my Rooks over to the half-open g-file, or even to try to win the h6 Pawn. Or try to round up the f7 Pawn. Meanwhile, having the move, I had an opportunity to block Black’s a-Pawn on a7 and artificially isolate the b5 Pawn. So I played Ra6, an active-looking move attacking Black’s h6 Pawn. I did this even though I knew Black could play the simple and effective …Bb6, because I had to try something. I gambled that my opponent would not want to move the centralized Bishop on d4 “backwards” as defense, but would want to keep it there to attack my undefended b2 Pawn. Yes, psychology at work.

I gambled further by not taking the offered h6 Pawn in return for my b2 Pawn, because simplification, even though objectively this was clearly the “best” move, because the “best” move doesn’t mean much if it only reduces swindling opportunities in a dead draw, and again because of psychology: my opponent had not played …Bb6 in the previous move, and probably would not play it again, and therefore would be playing the passive …Kg7 instead. And that happened. And I continued with advancing my f-Pawn to f5.

There was some risk in playing these suboptimal moves: a good defense would have left me fighting on the worse end of a draw. But I needed to win this game, and had plenty of time on the clock, so I was willing to fall back on defending a draw if anything did not work out.

Then I offered to trade our opposite-colored Bishops, by unprotecting my Bishop while moving my Rook up to “threaten” to come to the g-file. There was no real threat, but as I hoped, my opponent eagerly swapped Bishops, thinking (correctly) that this would neutralize the “attack”. However, objectively, the trade only benefited me. I got rid of a strong Bishop and lost my weak one.

A couple more gambles, and I made progress, losing my b-Pawn in return for his a-Pawn but now having one Rook on the g-file and one Rook on the 7th rank. Optically it looks a little scary, but that’s an illusion. Nevertheless, when an opponent is short on time, creating illusions can be useful.

After more passive moves by Black, I achieved my final dream position: Pawn on f6, Rook on a7, Rook on g7, about to win the f7 Pawn. It’s amazing how this fantasy position I had imagined early on actually came about. Still a draw, of course. But Black remained passive, and after a trade of Rooks, we actually ended in a Rook and Pawn ending that was winning for me. Unfortunately, at move 42, with the win in sight, I hastily made a passive move myself (Rg4) that threw away my win. I realized a few move later that the game was a truly dead draw. But I kept playing. A few seconds before his flag fell, my opponent made the only losing move, trading Rooks into a lost King and Pawn ending (two Pawns to one). Tragic, but in this tournament, where none of us were Masters, and endgame knowledge is weak, this stuff happens.

The moral of the story:

  • Endgame knowledge is very important. I’m not going to lie: I’m currently remedying my defects in the endgame (better late than never). I’m tired of displaying my games in which my weakness is obvious.
  • Even if you know an ending should be a draw, press on because you might get lucky.

The annotated game

Franklin Chen


Rook And King Vs. King

Learning elementary checkmates is a first step towards learning endgames. Today I’m going to discuss the checkmate with rook and king against the king. Students must have a prior knowledge of checkmate with a queen and king and also stalemate. In case you need some guidance here is the nice article written by Richard James.

In order to do checkmate with rook and king, I have divided the process into 3 parts.

1) Reduce the box. For example your rook is on a1 and opponent king is on d5 then the opponent king has freedom to move in (from b2-b8-h8-h2-b2) box and our aim is to reduce that box to force the opponent king to move to the edge of the board. As we all know that a queen is alone able to force the opponent king to move to edge of the board, but the rook can’t do this so you need your king to be there to support the rook.

Image 1

2) Once the opponent king is at edge of the board, keep your rook away.

3) Try to force the position where both kings are opposed to each other with a distance of a square.

Here is an example to illustrate the process.

Ashvin Chauhan


Endgame Studies for Training Purposes

Studies are ideal for training your calculation skills, because they have a unique solution and “there are no positional assessments in studies” (Botvinnik), so there is no argument if your solution is correct or not. But you need to choose the right studies. Many modern studies are unsuitable for the average player, because they are too complex, and rely on computer-generated key positions, such as reciprocal zugzwangs, etc. Many others are just too difficult for the average player to have any real hope of solving. Sadly, books such as John Nunn’s “Endgame Challenge” and Dvoretsky’s “Studies for Practical Players” are just too hard for the average player.

Instead, look out for studies by composers from the earlier part of the 20th century, especially Kubbel, the Platov Brothers, Fritz, and Troitzky. They usually have relatively few pieces on the board, and single-variation solutions that are only 6-7 moves long. These are ideal for calculation practice, and within the average player’s solving capacity. Two excellent books of such material are Jeno Ban’s “The Tactics of Endgames” and Beasley & Whitworth’s “Endgame Magic”.

Here is a typical example of the sort of study to look for:

Steve Giddins


“Satisficing” as a Strategy for Rapid Games

I played a rapid game bright and early this AM. I was up by 6:30 AM and playing before 7AM. My opponent and I were pretty much equally rated.

I’ve been watching Nigel’s ChessBase video on the QGD Exchange Variation all week. It’s an excellent video for chess improvers, since Nigel focuses on plans and ideas rather than memorizing lengthy variations. The last few clips discuss how the lessons one learns with focused attention on a particular opening can often be transferred to other openings.

A case in point was my game this morning. I chose to play 1.e4 and my opponent decided to play a Caro-Kann Defense. One of the examples Nigel demonstrates on his DVD is a Caro-Kann Exchange Variation that transforms to something like a QGD Exchange Variation with colors reversed. I opted to try for that. We didn’t get there, but we did get to a game where I could use some of the strategies from the QGD Exchange Variation.

Rather than adopting the minority attack, I decided to use my IQP to lever open the center. I consider the position in the diagram below and decided that 12.d5 was a strong move. I had three pieces defending my IQP on d5. Black had only two pieces defending d5 and had an undefended bishop on d6.

Black made a dubious move with 13…Rb8. Black should have played 13…Ne4. My evaluation of that position was roughly equal. We could have had a good game with reasonable chances for both sides. My reply was strong, trading knights on f6 and then bringing my rook from f1 to d1, where I could hassle the black bishops.

Black made a futile piece sacrifice, 16…Bh2. Black was down a piece with no compensation, except the h2 pawn. I didn’t make a serious blunder in this game, so the fruitless bishop sacrifice was enough to cost Black the game.

My weak play in this game was in the endgame. I waited much too long to get my connected, unopposed queen pawns moving. Delaying their movement didn’t change the outcome of the game. Two pieces up after 29.NxR, I was reasonably confident of the win throughout the endgame. Mobilizing my queen pawns earlier would have shortened the endgame. It was better endgame technique. Inefficient play is to be avoided, especially for us older players during a weekend tournament. Extra moves require extra time and mental energy, which increases opportunities for fatigue in later games and that can lead to costly mistakes.

I didn’t consciously relax my game during the endgame. I didn’t look for the quickest win, either. With rapid games, there is usually not enough excess time to calculate the most efficient solution. Instead, I often “satisfice.” I look for a satisfactory solution that is sufficient to win. That’s what I did in this game. I “satisficed” my way to a winning endgame.

Glenn Mitchell


King and Pawn

Here’s an extract from a game between two of Richmond Junior Chess Club’s less experienced members. Black, the older and stronger of the two boys, has a rook, knight and pawn against his opponent’s lone king, as well as the advantage of the move. It should be easy to win, shouldn’t it?

Mr Silicon Knowall announces mate in 4 here, starting with the rather improbable 46.. Rd4 47. Ke8 Nd8, but your first thought might be just to push f5 as the white king is behind the pawn. Black did indeed push his pawn, but only one square. (It’s interesting how often less experienced players forget that pawns can still move two squares on their first move when they reach the ending.) Still winning easily, of course, but no need to lose the knight unless your pawn’s going straight through. It looks like Black failed to ask himself the vital question “If I do that, what will he do next?” before making his move.

Anyway, White took the knight and Black continued his plan of advancing the pawn: 47. Kxe6 f5 48. Ke5 Re4+ 49. Kd5. But now came 49.. f4, repeating the same mistake from three moves ago and this time losing his important rook and ending up in a drawn KP v K ending. Most children soon learn to check that they’re moving a piece to a safe square, but what they find a lot harder is to see when they’re leaving another piece unsafe, for example, as here, by moving a defender. A minimal position such as this provides a graphic illustration of the problem. Black understandably wants to promote his pawn, but again fails to ask himself “If I do that, what will he do next?”.

So, 50. Kxe4 Kg5, reaching the next diagram.

A critical position with White to move which every serious player needs to know and understand. Most beginners’ manuals start off their coverage of KP v K with the rule of the square, but this sort of position is far more important. You’ll almost always find in practice that the kings start off close to the pawn, as here. Let’s see whether our two novices knew what to do.

The next two moves were natural and fine: 51. Kf3 Kf5 52. Ke2 (here any move to the second rank draws, but one rank down and only Kf1 would draw, which is why I teach children always to retreat to the same file as the enemy pawn) 52.. Ke4 Now White has to play Kf2 to draw, but if you don’t know this and you lack the skills to work it out it’s natural, I suppose, to play 53. Kf1 instead, which is what happened.

Now Black has two winning moves. Everyone should know that the position after 53.. Kf3 is won with either player to move. It might be harder to remember that the position after 53.. Ke3 is also won regardless of the move. But Black, again naturally if you don’t know the position, played 53.. f3 instead. It may be counter-intuitive but you have to remember to get your king in front of the pawn. Now White has one way to draw: 54. Kf2! Kf4 55. Kf1! Not knowing this, he played 54. Kg1 instead and after 54.. Ke3 55. Kh1 f2 Black managed to promote his pawn and eventually win the game. (White’s 55th move is strange but here it doesn’t make any difference. Inexperienced players often move their king away from the pawn in these endings and if you ask them why they’ll tell you that if they keep near the pawn they might get checked.)

Two important lessons for all novices (and all who teach novices) from this. Always ask yourself the question “If I do that, what will my opponent do next?” to avoid losing pieces by moving or blocking defenders. And make sure you know the basic KP v K endings, backwards, forwards and inside out, blindfold and with your hands tied behind your back.

Richard James