Category Archives: Annotated Games

Accept The Sacrifice If The Alternative Is To Lose Anyway

When I was in my final year of high school, I played in the last tournament of my life before I returned to chess two decades later: I played in the 1987 Michigan High School Team Championship. I ended up winning the first board prize with a perfect score of 5 points, but I always felt funny about how I achieved that, because in one of my games I played a sacrifice that I felt guilty about for two decades. Also, that was the only tournament in my life that I ended up losing my score sheets for, so I do not even have the full score of that game. But I do remember vividly the moves leading up to the critical position, and my mindset.

Seeing the possibility of a Greek gift sacrifice

On move 10 out of the opening, I suddenly spent a huge amount of time deciding whether to play the “Greek gift” sacrifice against my opponent’s King, sacrificing my Bishop on h7 with check.

But in my attempt to calculate a win, I could not find a forced win. I saw defensive resources, so I was reluctant to play an unsound sacrifice. But the idea of playing the sacrifice really appealed to me. You have to understand that I had never played the Greek gift sacrifice before, only read about it in books, and also I knew this might be the last chess tournament of my life, as I was going off to college in the fall, and I had actually “retired” from chess in my sophomore year of high school, and came out to play in the Michigan High School State Championship only because I had started up a chess club in my high school in the fall in hope of boosting my college application (I brought four teammates who had never played in a tournament before). I outrated my opponent by over 500 USCF rating points, so there was no need for me to play recklessly to win, so my motivation was just to finish my chess-playing days in style.

I did see that I would get compensation for the sacrifice, and therefore should not lose if I played the sacrifice, but that was all I could see. Even after I went home to analyze the game, because I did not have access to good computational power in the 1980s, I did not believe I had the full truth of the position until the 2000s, on my return to chess, when chess engines by then had become very strong.

Sacrifice declined!

I was simultaneously ashamed and relieved when my opponent thought only briefly and declined the sacrifice, and therefore easily lost, being a Pawn down without compensation, and having a weakened King side also.

My opponent must have concluded that my deep thought meant I had figured everything out, but in fact, my deep thought came from not having figured it out! Granted, I was much higher-rated than my opponent, but higher-rated players can make terrible moves too, and sometimes even deliberately as a swindle, so you should think for yourself for a bit, and not always assume your higher-rated opponent has everything figured out. Granted, psychologically it was clearly a shocker to him that I thought mysteriously for such a long time moves before the sacrifice.

In club play, I often see fear of accepting sacrifices, and painful losses resulting from declining. The loss is usually painful because a sacrifice significantly disrupts a position, so if your position is disrupted anyway, and there is no visible immediate mate, maybe you might as well grab some material for your trouble; if the attack goes wrong, then you may have a good chance of consolidating and winning as a successful defender. Part of chess is choosing to defend.

So I’m saying, accept the sacrifice if you honestly do not see anything wrong with doing so. You might be making a mistake, but at least make the mistake and lose rather than choosing the path of sure loss, losing material against a much higher-rated player.

How sound was the sacrifice?

The fact that White is missing the dark-squared Bishop and only has a Queen and two Knights really restricts White from having a win in this position. The only possible things White can do are try to push h4, maybe castle Queen side, and use the two Rooks somehow. Meanwhile, Black can defend the King and develop. Note that if White tries to win back an exchange, the result is an unfavorable balance of material in which White gets a Rook and a Pawn or two for two minor pieces, so it is no use for White to regain material.

I’ve inserted some variations into my annotations below.

Irony: there could have been an alternative Greek gift sacrifice!

The irony is that if I had played Nc3 instead of Bd2, and “normal” development had continued, with Black “castling into it”, then the Greek gift sacrifice would have been obviously sound and winning. The huge difference is that with White’s dark-squared Bishop still on the board, and guarding the Knight on g5, White does not have to support the Knight with the Queen, but can calmly play h4, followed by Qg4, with a deadly barrage of discovered checks to follow: a check with the Queen or with the Bishop on c1 if the King goes to h6.

Note that it is important to play h4 first, to avoid Black’s tempo-gaining …f5 against the Queen on g4, because with the Pawn on h4 first, then h5+ can be played at any time, and optimally when Black’s King on g6 cannot escape to f5. Check it out with a computer engine if you want to verify that it’s a quick win for White.

Why did I play Bd2 anyway? I had some vague idea that getting rid of Black’s “good” Bishop for my “bad” one was advantageous. Also, note that I recaptured “wrong” with Nbxd2; I just recently wrote an article about why Qxd2 is usually best. But in 1987, my positional understanding was not so good.

Some resources on the Greek gift sacrifice

A well-written overview by GM Daniel Naroditsky.

A previous Chess Improver article by Ashvin Chauhan.

A 2012 game of mine in which I played a correct Greek gift sacrifice.

The game (up to the point of the sacrifice)

Franklin Chen


Don’t Be In A Hurry To Regain Sacrificed Material

I have been discussing gambits with a student of mine, to explore the concept of positional sacrifice, a sacrifice of material with long-term positional objectives rather than with a clear immediate tactical winning shot. A position came up that illustrates a dilemma that often comes up when playing a positional sacrifice: when do you “cash in” your advantage? Very often, the logical result of a positional sacrifice is boxing in your opponent in such a way that tactics become possible, such as regaining your lost material. It is very tempting, after having played a Pawn down for much of the game, to see a way to finally win the Pawn back and do so, and in fact, in many situations it is entirely appropriate to win back your sacrificed Pawn “with interest”, retaining a positional advantage. But often it is actually counterproductive to “cash in” too early, if that results in giving up much of the non-material advantage you have so carefully accumulated.

Here’s an example, in which White is down an h-Pawn but finally has a chance to win Black’s h-Pawn in return. Unfortunately, doing so is terrible and results in a dead equal position in which White has nothing to look forward to and even has to be careful. Instead, White can ignore the fact of being a Pawn down and continue to build up with deadly pressure. Let’s look at why.

Assessing the position

First, let’s look at the position with White to move.

White has more space, with a Pawn on e5, a powerful dark-squared Bishop on f6 restricting Black’s activity, a grip on c5, active Rooks doubled on the h-file, and a Queen attacking Black’s Rook on h7.

Black has a pitiful Rook on h7 that is completely immobile and under attack, a Queen on g8 that is tied to defending that Rook, a light-squared Bishop that has no possible moves, a Rook on d8 that can barely move, and a dark-square Bishop on c7 that is currently attacking nothing (there is no hope of getting at White’s e5 Pawn).

So for a mere Pawn, White has a tremendous-looking position. The question is, how to cash in eventually?

What taking back the h-Pawn accomplishes

Taking the Black Pawn on h6 throws away all of White’s advantage:

  • Black gets to trade off his worst piece on the board, the trapped Rook on h7.
  • Black gets to swoop down and activate the Queen by checking on White’s unprotected back rank.
  • Tactically, the threat to win White’s f2 Pawn forces White to retreat horribly with Nd1, self-pinning the Knight (which wants to go to e4) and making it do nothing other than defend the f2 Pawn.
  • Black can start putting pressure on White’s d4 Pawn, and think about swinging the Rook to the open g-file and coming down.
  • White no longer has any threats. The Bishop on f6 was once a powerful piece keeping Black’s Rook on h7 trapped against its own h-Pawn, but both of those are now gone. White’s Queen is now just guarding the d4 Pawn.

I’ve given a sample bad continuation by White to illustrate how quickly Black can actually end up winning, if White tries to continue an “attack”.

Continuing the pressure instead

The alternative to winning the Pawn back is to observe that Black has serious problems with the Rook trapped on h7. White can try to win the Rook eventually, by driving Black’s defensive Queen away from protecting it. Rh3 and Ne4 and Rg3 not only improve White’s pieces in general but also serve to restrict Black’s activity and aim to win Black’s Rook.

A defensive counter-sacrifice!

It turns out that the only real try for counterplay by Black to avoid losing the Rook is to open up the position and try to activate the light-squared Bishop. Sacrificing the c-Pawn with …c5 is the best chance for Black, to get the Bishop to …c6, where it exerts power over the light-squared diagonal to h1, and also can at least, if needed, trade itself for White’s powerful Knight on e4.

So it is ironic that White’s best plan is to ignore the Pawn on h6 and instead force Black to give up the c-Pawn instead as a defensive sacrifice. After White gains the c-Pawn, of course, White has a huge advantage still, but at least Black is still surviving and has some practical chances.

Franklin Chen


The Difference Between a Knight Developed at c3 and at d2

There is a well-known trap in the Bogo-Indian Defense that raises an interesting question whenever I show it to someone. The trap is as follows and involves a question of how White should recapture after a trade of Bishops by Black:

The question after this trap is always, “Well, why would White ever want to play Qxd2 anyway, exposing the Queen to an attack by …Ne4? Isn’t it obviously better to recapture with Nbxd2, simultaneously developing the Queen Knight?” This is an excellent question. It is best answered by examining some long-term issues in the middlegame arising from this opening.

Comparing Knight developed at c3 and Knight at d2

Knight at c3

First, we look at what can happen if Black mistakenly allows White to recapture the Bishop with Qxd2 instead of Nbxd2, by not taking White’s Bishop early enough for the “trap”.

Black’s plan in this variation of the Bogo-Indian is to play …d6 and …e5, attacking White’s Pawn on d4 and encouraging White to close the center with d5. After the center is closed, all attention must be directed toward Pawn breaks by either side.

White is acknowledged by theory to have some advantage in this opening, having more space and a lead in development, and can think about attacking either on the Queen side (with plans such as a3, b4, c5) or on the King side (with plans such as e4, Ne1, Nd3, f4). But Black has a solid position, and can aim for counterplay with …a5 with …Na6 or …Nbd7 aiming for …Nc5, and/or …c6, to prevent White from gaining too much ground on the Queen side, and perhaps preparing slowly for …f5 to further attack White’s e4 Pawn chain base.

Knight at d2

By contrast, let’s see what happens when Black correctly forces White to recapture the Bishop with Nbxd2.

below is a sample continuation, in which at move 13, probably White’s best move is the paradoxical undeveloping move Nb1! The Knight at d2 is not doing much, being blocked by White’s own c4 and e4 Pawns. More important, it is not controlling the important a4 square (that Black can possibly aim to occupy with …a4), and it is not controlling the b5 square that could also be important (in a later attack against Black’s c7 and d6).

But this retreat wastes two moves (the original Nbxd2 and the Nb1) before getting to c3. However, in the Qxd2 situation, White wasted a move with the Queen, which is not so well-placed on d2: White’s Queen is actually better placed on d1, where it controls a4, than on d2. But White’s Rooks are not connected, so White will eventually want to develop the Queen anyway, perhaps to c2. So overall, White has lost one move, net, and, and this does make some difference in White advantage, even in a closed position, because the extra White move in the Nc3 variation makes it that much harder for Black to catch up in development and begin counterplay.


The summary of the situation is that paradoxically, since White wants the Knight on c3 anyway eventually, “saving” time by recapturing with development by playing Nbxd2 actually ends up wasting a move because the Knight will have to spend two more moves to get back to c3. Knights are funny pieces because any time a Knight has a choice to go to one of two different squares, if it chooses to go to one of them, it will always require two more moves to get to the alternate square. This is something to think about when planning Knight maneuvers: it is efficient, when possible, to plan to get to a desired square with the smallest number of moves possible (given the tactical constraints).

The other point to remember is that “wasting” moves to get a Knight to a good square may be justified. “Backwards” Knight moves are very important in chess, because a Knight on a good square can be so powerful that it is worth spending the time to get the Knight there. Look at how White thematically “undevelops” the Knight on f3, where it is doing nothing, to e1 and then to d3, to control the c5 square and b4 square (in case of a Pawn advance to b4 in the future) and also regain pressure on Black’s e5 Pawn and help support an f4 advance.

Study of typical middlegame positions in the Bogo-Indian can pay off with better understanding of the roles of both of White’s Knights and both of Black’s Knights (Black’s King Knight was not discussed here, but it has plans too).

Franklin Chen


Smyslov: Master of the ‘Coiled Spring’ Approach

I have always been an admirer of the late Russian Grandmaster, Vassily Smyslov. One of the things that drew me to his games, was his ability to take on cramped positions without becoming passive. He would then, very often unravel his position, rather like a spring which is wound and full of tension before being released. There would then be an explosion in which Smyslov would take space, and begin to relocate his pieces to more advanced positions, very often to carefully prepared squares.

Smyslov’s play, I must say, suits me very well, his style fits very well with mine and the openings I play. For players who play openings or piece setups where development is initially contained to the first 3-ranks and the opponent very often establishes in the centre, studies of his games really can not hurt at all.

Infact, this approach is one of the best ways to become familiar with an opening.

In the following game, I would like to draw your attention to how Smyslov plays quietly and subtely, but remains active (not an easy thing to do!). He controls the situation, playing accurately and responding to his opponent. Steadily, his position improves, and he is able to pounce when his opponent shows an obvious lack of technique and understanding of the type of position.

John Lee Shaw


Amateur Versus Master: Game Eleven

This chess game is one of my recently completed games form the 2011 Golden Knights Final. Bonsack is the second master that I have drawn in this section and the second highest rated in this section at 2344. Unless I am mistaken, I have drawn a few 2300 rated players in correspondence chess, but I have yet to beat one. So far, I have one loss, two draws and no wins in this section.

I am moving out of my current apartment this weekend and my opponent knew that I was taking a month off from chess for this move. I think that he felt sorry for me and that may be why he offered a draw in a position that favored him. Whatever the real reason for the draw, I’m glad.

This game transposed into a Benoni Defense. At the points in this game where I say “slightly better is” or “possibly better is”, it is because the chess engines do not agree on the moves and I am unsure myself.

At move number 11 I decided to keep the position closed for a while so that I could restrict the range of White’s bishops. In previous chess games against masters I have gotten burned when my opponent’s bishop pair came to life. I avoided that in this chess game.

Once White’s Knight was firmly established on b5, I could never dislodge it without giving him a passed pawn on the Queenside. That did not favor me, so I eventually decide to go for play on the Kingside by opening it up. That is when White started moving his King over to the Queenside. Not much happened after that.

Mike Serovey


Fianchettoing Your King’s Bishop May Weaken Your c4-Pawn

I saw a curious position recently in which strangely, White was in a position to lose a c-Pawn placed on c4. I then remembered that it is actually not uncommon for this Pawn on c4 to be undefended when White has fianchettoed the Bishop to g2, because unlike classical development of the Bishop, where the Bishop is on e2 or d3 and therefore protects the Pawn on c4, the Bishop on g2 does nothing to protect the light squares from f1 to a6. Check out this position:

A standard theme for Black counterplay

Many middlegame plans by Black in these kinds of opening development setups in fact target White’s c4 Pawn and the light squares on the Queen side in general, while White tries to make something long-term out of increased central control of e5 and d5 (over classical development of the Bishop) and of course the long diagonal from h1 to a8. These positions can be very subtle for both sides to play. In this article I’m not discussing any of these subtleties, but simply pointing out a common theme for Black.

A variation of the King’s Indian Defense:

A variation of the English Opening:

And of course, the concept behind a popular approach to the Queen’s Indian Defense:

Franklin Chen


Remember Games and Patterns

You might have heard that Carlsen can remember numbers of positions and recall them over the board in a limited amount of time. In the book GM-RAM, by Rashid Ziatdinov, the author emphasises remembering key positions and games and claims that “if you know just one of important classical games, you will be able to become a 1400 level player, to be world champion you will need to know 1,000 such games”. This may be too much but we can’t deny fact that remembering these games cold will definitely help you towards chess improvement.

I tried different ways to remember games, for example playing them over the board many times, guessing them move by move, using Chess Position Trainer etc. But they didn’t work that well for me.

Then I tried one more thing and succeeded. This method uses lots of time but definitely works; after a month without playing them through a second time I am able to remember the games and their critical positions.

The way to do this is to take a book of your favourite player where he has annotated his games. Now we are going to annotate his games in our words rather than going through author’s annotations first. You can use different software but a pen and paper works best for me.

The most important thing is that your focus must be on one direction but with inherent flexibility (if your opponent blunders you must be able to punish him). This tends to be missing from the play of amateur play as they fight in different directions. Write down your ideas for each move (for both White and Black) and don’t worry if you repeat the same thing over a series of moves. Once you finish it (normally I take 4 to 6 hours) go to the experts annotations and compare. You will find that now it is very easy to understand the author’s points and your mistakes, this wouldn’t have happened if you went directly to the author’s annotations .

It is also wise to go for a second opinion also, if someone has explained the same game. Players who have the time and work like hell will definitely get benefit from this!

If you find this is very hard and time consuming, first watch this video:

Ashvin Chauhan


Clash of the Titans

OK, both my opponent and I are experts, not yet masters. Still, this chess game was hard fought by our chess engines! We both were posting our analysis on the server. I could see what he was analyzing with Stockfish 5.0 SE and he could see what I was analyzing with Houdini and Deep Fritz. Truthfully, I doubt that either one of us would have found half of the moves that we played had this been an OTB chess game. Again, ICCF rules allow us to use chess engines.

This chess game is one of two draws that I have in this section. I also drew the player that Miloslav defeated, so Miloslav is temporarily in first place, I am in second place and Don Pedro is in third place. If I can finish my remaining games with at least a draw in each one I may remain in second place in this section.

Against unknown opponents I will often play the Modern Defense. It did not take long for my opponent to get me out of my database of games and into unique analysis. About half way through this game I realized that someone was anonymously following my analysis on From that point on, my opponent was playing whatever moves Stockfish recommended. There were a couple of times in the thick of it that my chess engines thought that something else was better for White. The notes that I made during this game (see below) explain the rest.

Mike Serovey


What Not To Do If You Have The Isolated Pawn

A typical introduction to positional principles in chess covers the advantages and disadvantages of having an isolated Pawn, a Pawn that has no Pawns on the files adjacent to it and therefore cannot be protected by another Pawn. (In particular, the most common isolated Pawn is the isolated d-Pawn.) Since it is easiest to understand why an isolated Pawn might be a long-term static disadvantage, many players reflexively go out of their way to avoid ending up with one. The situation is not helped by the use of illustrative games in which one side has an isolated Pawn and suffers quite a bit before losing the Pawn and the game.

But as a student of mine pointed out while studying such games, the situation is not actually that simple. Yes, it can be frustrating defending a position in which you have an isolated Pawn without any of the benefits (not discussed here) of having one, but that does not mean the position is actually lost. Whether your opponent can actually make any progress is another matter. It is instructive to know how to play for a draw in an unpleasant defensive position. Much chess instruction focuses on how to win, but ignores questions of how to avoid losing.

Here’s a classic isolated-Pawn game that ended poorly for the defender.

Korchnoi-Karpov, World Championship in Merano, 1981

Sliding from an advantage to equality

In the opening, Korchnoi as White accepted an isolated d-Pawn position. Karpov responded with a “Knight on the rim” move 11…Nh5 to trade off dark-squared Bishops. This wasn’t actually very good. It potentially gave White precious time to create a thematic good position: White could have played Re1, Ne5 with pressure against Black’s f7 and e6 Pawns, then begun a thematic attack on Black’s King side (especially with the h6 advance weakening the King side already), either through a Qd3/Bc2 lineup and/or a Rook lift with Re3/Rg3, something like that. (Full discussion of how to attack if you have an isolated d-Pawn is outside the scope of this article.)

White dawdled with 13 Bb3 and then 15 Qe2, which did nothing to create threats against Black’s position. And then White played 16 Ne4? which resulted in a simplification that left White fighting for a draw.

Simplifying trades are what you do if you are playing for a draw with an isolated Pawn, to reduce the other side’s attacking possibilities.

Refusing to accept that the goal should be to defend a draw

On move 19, White had the opportunity to trade Rooks and practically guarantee a draw. The fewer the pieces, especially powerful long-ranging major pieces, Rooks and Queens, the fewer opportunities for the opposing side to win the Pawn and still have a middlegame initiative to win the game. So White should have simplified here. The task of drawing would still have been slightly tricky, but doable, requiring keeping track of Black’s Queen, Rook, and Knight activity.

On move 22, White made another mistake and played the backward-moving 22 Qe1? It was best to simply wait around and do nothing, after having everything well-defended: White’s Queen was centralized at e4, protecting the d4 Pawn and exerting pressure on the d5 square.

Often, in a defensive position, the best thing to do is to wait for the draw to happen. Trying a funny plan when there is nothing really going for you can backfire badly. White has no winning chances in this position.

Unnecessary passivity

On move 23, White played 23 Rcd3? which just turned a fine Rook (on the open c-file) into a purely defensive piece. OK, the idea may have been to dissolve the isolated Pawn by playing d5, but this was easily parried by 23…Rd6.

Final simplification

On move 27, White traded the Bishop for the Knight on d5. Objectively this is OK, actually, if the goal is to draw. But the followup shows that was apparently not the goal. So the real problem is a mismatch between an idea and what is consistent with that idea.

Own pieces in disarray

28 Rb3? was a terrible move that took a defensive Rook and removed it from its defensive function, and also exposed the White Queen to a pin of the d-Pawn, in case Black ever got in …e5. 29 Qc3? compounded the problem by leaving the Rook on d1 completely undefended.

33 Qa3? removed the Queen from the action in the center and King side after White had already weakened his King’s position with the necessary 30 f4 earlier.

Now all that was needed for Black to win was to tactically take advantage of White’s uncoordinated pieces and unprotected King, and Karpov did that precisely.

An important note about how to draw

Sometimes the easiest way to draw is to just give up the weak isolated Pawn without a fight, in exchange for activity and simplification. Instead of risking King unsafety with 30 f4, White could have decided to just lose the d-Pawn but keep King safety intact and Rooks and Queen active, say with 30 Qf3. I will confess that I have been held to a draw a couple of times in games in which I expended effort to win an isolated Pawn but at the cost of massive simplification and could not win the ending.


To answer my student’s question about this game: yes, there were multiple turning points in the game at which White could have held still and played for a draw. Especially in the case of an isolated Pawn and much piece simplification, there is often no way to win the Pawn or force some other concession somewhere else, if one just puts pieces on good defensive squares and then just waits. The only way for the other side to win is to break through by distracting the defensive pieces and taking advantage of pins and the single possible Pawn break (…e5 here) to create threats elsewhere on the board (such as on an exposed King). Note that the game was not lost because of losing the isolated Pawn: the game was lost by trying too hard to hang on to it.

The complete annotated game

Franklin Chen


For the Love of Doubled Pawns

Having doubled Pawns in a chess game is rightfully taught as being potentially disadvantageous, and a vital step toward improvement in chess understanding comes from learning how to play against them as a static positional weakness in one’s opponent’s position that has ramifications in the middlegame as well as in the endgame. However, to dive into deeper into the subtlety of chess, it is important to also know of the potential dynamic strengths of owning the doubled Pawns.

For example:

  • If you get doubled Pawns from recapturing toward the center, that may increase central control, which could be very important in the middlegame.
  • If you have doubled Pawns, you have an at least half-open file that you can potentially use for attack by putting Rooks on the file.
  • The front Pawn of the doubled Pawn pair can be used for attack, while the back Pawn of the pair can be used to defend the squares that were otherwise abandoned when the front Pawn moved forward.

Examples of the power of doubled Pawns

Bent Larsen was a great chess player who was famous for playing in unusual styles. One thing that he seemed to do often was invite having doubled Pawns and then making use of them effectively. Here are two games showing off how to make doubled Pawns effective. Notice that the doubled Pawns enabled setting up a “wall” behind which gradual positional maneuvering and improvement of pieces was possible while waiting for the opponent to go astray. Playing with doubled Pawns often takes patient regrouping.

Franklin Chen