Thursday, August 14, 2008

Calculating Sacrifices

My opponent and I reached this unusual middlegame from the Classical Sicilian. Black's King is left all alone. And this usually signifies an attack. Black had just played Bb7 (see 1st diagram). So it is White to move.

Do you know what I was thinking?

1. Bxf7! was the sacrifice to open up the area around the Black King.

The question is: Do you think the sacrifice 1. Bxf7 works?

Calculate the lines after 1.... Kxf7.

Once you're ready, I'll go into the following thought processes that goes into my dysfunctional brain.

So please bear with me. :)

First of all, why 1. Bxf7? White has a positional advantage in that it has more active pieces and the pieces control a lot of squares on the board. True, I can play the slow way of squeezing my opponent to death with moves likes 1. a4, even h4 to break up the Black pawns and cramped Black further in.

Why sacrifice unnecessarily?

I believe that for a patzer player like myself, it is important to learn to see tactics and combinations where they occur. Rather than being afraid of playing bold moves, one should actively seek out good moves because that is the way a player can improve.

Anyway, back to the subject.

Does 1. Bxf7 work? Let's work it out.

What can Black reply with?

Looking at this, I calculated a few replies by Black. But I'll just explain one for the sake of brevity.

1..... Kxf7 is the natural response

2. Ng5+ (the natural followup, bringing another piece into the attack) Kg8 ( 2.... Kg7 or Kf8 runs into the horrible twin fork between King and Queen with Ne6+)
3. Ne6 (stifling the e7-pawn, effectively cutting off communication between Black Queen and King - see 2nd picture) Qd7 (forced)
4. Qd4 (Black is now forced to give up his Queen to prevent mate on g7)


I have actually calculated other lines as well including the replies
1 .... Rf8, 1.... Bxe4 but that discussion will take far too long for this post.

Instead, I'll leave you with a final puzzle after

1.... Rf8
2. Ng5 Bc8
3. Bd5 Nc6
4. Bxc6 (see 3rd diagram)

Now what happens if Black plays 4.... Qxc6?

Can you spot the mating pattern by White? Answers can be found by highlighting between the brackets. :)

[5. Rxe7+ leads to

..... 5. Kg8 leads to 6. Qc3 and mate on next move with 7. Qg7#
..... 5. Kh6 6. Rxh7#
..... 5. Kh8 6. Qd3+ Kg8 7. Qg7#
..... 5. Kf6 6. Rae1 and Black is going to be mated in the next few moves unable to stop mating threats with say Qd4+/Qf4+/Nh7+.]


  1. I am constantly amazed by good sacrifices! I applaud your calculated effort to look for these great moves, instead of just positionally "sound" moves.

    My initial thought was 1.Qd4+. Then I saw your move plan was 1.Bxf7 and my quick assessment was something like
    2.Ng5+ (wasn't sure where the King would move)

    As you can see I have trouble considering the opponent's possible plans, and generally only focus on my own plans. Not surprisingly, this usually leads to getting kicked off the board right in the middle of my plan. LOL

  2. Do you remember how long it took to do the calculations before you made the move? Seconds? Minutes?

    I'm wondering how much time people allot themselves to calculate a tactic like this.

    What was the time control on this game?

  3. Hello RT Solo,

    It was a game on FICS (think it was 20 with Fischer time controls). I had earlier blitzed out the moves in record time.

    I spent 15-30 seconds surveying my opponent's possible threats. I didn't see any but did notice Black was cramped in. I also noticed that the Black king is isolated and if I can plonk my Knight on e6, Black is in deep trouble because there is severe miscommunication between kingside and the queenside pieces.

    Like you, I saw the Queen check and thought of following it up but I couldn't find a way to make a way of attacking the Black King work. I thought that even if Black took my bishop, my position is still solid with a slight plus.

    It was at this point that I thought if I could somehow break Black's kingside pawns, I might get a win so I started looking for tactics. My mind leapt at 1. Bxf7 and the Knight followup check with gain in tempo and ability to jump/blockade on e6 attacking the Black Queen convinced me to delve further into this tree of variation and start examining possible replies.

    I spent approximately the next 5 (prob. more?) minutes or more taking my time to calculate possible replies.

    I looked at 1..... Kxf7 first and you see the line I just looked at. I then looked at 1.... Rf8 which was longish and while there was an exchange of pieces, the position looked good for White.

    The last reply I saw was 1.... Bxe4 but that just leaves Black an exchange down after 2. Bxe8 and this was quickly dismissed as Black's possible reply.

    For some unknown reason, the alarm bells in my head started to ring,"Tactics!" when the initial position was reached - which was why I paused for quite a while to take my time to calculate.

    I think it's good to take your time to calculate whenever you see a tactical shot and to take as long as possible.

    cheers and thank u

    August 16, 2008 2:17 AM

  4. Very nice tanc! In positions like this, I sometimes just act on gut feel (in short time control games anyways). I'll look at the position, see the lone king and my pieces aiming at him, and just play the sac. As Tal said - sac first, calculate later :) Doesn't always work out, and I probably wouldn't do it in a tournament game but I find myself doing it quite a bit online.

  5. Chessaholic: Thanks for the compliments! For some strange reason, I find that I'm more willing to look for sac'ing pieces than an actual tournament game.

    However, I find that one must be brave (not foolhardy) when performing a sac like this in an actual OTB game.

    I remember watching a live game between 2 prominent GMs. Black took amazing 40 minutes of clock time to calculate a particular sacrifice followed not by an overwhelming attack but by a deadly quiet move! I was gobsmacked. I'm amazed by players who are not afraid to play sacrifices when the demands call for it.