"Advanced Tactics"

Mar 30, 2008, 12:00 AM |
9 | Tactics

Yesterday's puzzle (March, 29th, titled "Advanced Tactics") was very interesting, and I posted quite a long text about it. As it is somewhere in the middle of several pages of posts, I suspect it won't be read much. Therefore I will repost it here as an article. I will give the position below, the move list contains the lines I have calculated, but I recommend you try to go through the lines in your head while reading instead of using the forward button under the diagram. This way you will probably get most out of reading this article.

So here is the position:

















And here is the post: 

Now this is a really nice and instructive puzzle. The position is quite game-like (probably it is from an actual game?), so it may be interesting for some of you to follow the systematic search an advanced player might conduct. I kindly ask you to apologize my lack in modesty to give my own thought process while I tried to solve the problem, but naturally this is easiest to write about for me Wink

I start with a quick survey of the position:


-material is even.

-there are no obvious threats to either king. 

-both sides have weaknesses: White's are the squares f3 and c4, his back rank and the misplaced knight on b2.

-Black's only real weakness is the bishop on c6, which may be left en prise if some of the other pieces move, as it is attacked by White's heavy pieces and along the long diagonal (if Black's e-pawn moves).

From the above we can conclude that chances are there is a combination working for Black, especially as it's his move. So we start looking for forcing continuations, namely checks, captures or threats.

The candidate moves (moves to be calculated) I identified were 1...Nf3+, 1...Qa1+ and 1...e3 (this move opens an attack on the bishop on g2, preventing White from taking on e3, and therefore has forcing character, too). Only after calculating these moves and not finding a convincing solution will we start looking at quiter moves with the rook or bishop.

We start with the checks: a) 1...Nf3+ 2.Bxf3 exf3 3.Qxc6 Qa1+ (3...Re1+ 4.Kh2 Qa1 5.Qxf3 is less precise) 4.Rd1 Re1+ (4...Qxb2 5.Qxe8+ is obviously hopeless)
5.Kh2 (weaker seems 5.Rxe1 Qxe1+ 6.Kh2 Qxf2+ 7.Kh3 Qf1+ 8.Kg4 f2, when the passed pawn is a dangerous force) 5...Rxd1 6.Nxd1 Qxd1 7.Qe8+ Kh7 8.Qxb5 Qe2 9.Qc5 g6 (9...Qf1 fails to 10.Qf5+ and 11.Qxf3) 10.Qe3 doesn't look winning for Black at all. This line may seem long, but it is pretty forced with little branches along the way, so we should not shy away from calculating it. However, from a practical point of view it may be sufficient to see that White is by no means forced to take the bishop c6 on move 3 but can play some cheapscate move like 3.Rd1, when forcing play seems pretty ended. This being a puzzle, one can probably discard 1...Nf3+ on these grounds alone, but in a practical game we would need to evaluate the position following 3.Rd1: After all, White has parted with is fianchetto bishop and gotten himself a coffin nail on f3.

b) So we move on to 1...Qa1+: 2.Rd1 Qxb2 obviously wins, as nothing in Black's camp is attacked, so 2.Nd1 is forced. Other players may find 2...e3 immediately here, but I had the idea to play this move with the queen still on a8, as that opens fire on the g2 bishop and prevents White from taking on e3. Consequently I only checked 2...Nc4 3.Qxc6 Nxd6 4.Qxd6, which doesn't look winning for Black at all. Had I correctly identified the candidates on move two, I would have started looking at 2...e3 now, thus saving a lot of effort and time.

Instead, I started working on c) 1...e3:

2.Bxc6 (mental note: If this turns out to be in Black's favour, we need to check if 2.Rxc6 may be an improvement.)2...Qa1+ 3.Rd1 (3.Nd1 e2, and Black wins) 3...Qxd1+ 4.Nxd1 e2 5.Bxe8 exd1D+ This position is difficult to assess, but it seems that Black has only lost himself a pawn and may be able to create some threats against White's king due to his better piece coordination, but it is unlikely he will have more than a perpetual, especially as the white queen still covers f2.

Up to now, none of the variations was satisfactory for Black, so I started looking if I had left out any candidate moves on move 1:

1...Nc4 2.Nxc4, when there is no decent follow-up for Black is discarded rather easily.

As line c) would have been absolutely winning if the rook hadn't been on e8, I tried moving it away from there with tempo: d) 1...Rb8 2.Qe3 (blockading the pawn) 2...Qa1+ 3.Nd1 Nc4 wins.

Unfortunately 2.Qd4 e3 3.Bxc6 Qa1+ 4.Qd1 Qxd1 5.Rxd1 isn't a forced win for Black.

Ok, so we can dispose of 1...Rb8 as well.

In a game this would probably have been the end of it, unless I would have still had plenty of time on the clock, so I could check over the lines I already calculated. In a puzzle, of course this is exactly what you do.

The lines a) and c) are pretty straightforward, so we arrive quite quickly at line b)

1...Qa1+ 2.Nd1, and now 2...e3 jumps to mind: obviously White can't take on c6: 3.Bxc6 e2 (winning) is a position we have already arrived at calculating line c). So what if White takes the pawn e3?

3.Qxe3 Nf3+ 4.Bxf3 Rxe3 5.fxe3 Bxf3 wins

3.Nxe3 is illegal :o)

3.fxe3 Bxg2 (getting rid of the only weakness at last) 4.Kxg2 Nc4 5.Qc6 attacking e8, damn! ...Ah yes! 4...Qa8+ 5. K anywhere Nc4 seems winning. Hooray!

We see that just bashing out plausible moves with a good tactical eye and a bit of luck might or might not have led us to solving the puzzle in a few seconds, inevitably followed by some "easy" comment. Doing it my way I spent about half an hour calculating all those above lines, but this is as it works in a game, where it can be quite rewarding to invest that much time in a single move, by the way. The only problem is to get the feel when exactly to invest this time, as obviously you can't do that on every move. There have been several attempts to answer this question, most of which included the word "intuition". While they are certainly right to some degree, it appears to me that you can develop this intuition along with another important quality: experience. Playing many, many games (I definitely don't refer to blitz games here, though!) is a much more effective pastime than learning opening variations by heart, for example, and will eventually get you there.

As I haven't checked the lines with a computer (or even on a chessboard, for that matter), they may contain heavy mistakes. If you find one or several of them, please comment!

Thanks for reading this post, I hope it is of some help to those striving to develop their tactical skills. Smile


More from Torkil
Meeting 1.e4: Sicilians and Anti-Sicilians

Meeting 1.e4: Sicilians and Anti-Sicilians

Meeting 1.e4: Semi-Open Games

Meeting 1.e4: Semi-Open Games