Ghosts And Fallability

While playing games be it, OTB or over the internet, I noticed a few 'ghosts' cropping up in my gameplay.

One of these flaws was that I have a tendency to not play the best moves after getting in a few nice moves. In other words, after making a series of good moves, I followed it up with an impulsive move which usually leads to the start of my downfall.

I will highlight this in this game here (see 1st diagram).

White has a slight edge here having a passed pawn and more active pieces. With correct play, White should be able to convert this.

But I was disturbed by the doubling of the Rooks on the h-file. I definitely do not want Black to play Bxf3 leading to a doubling of my pawn structure and even worse, allowing Black to crash through the Kingside which has become porous (with 1... Bxf3 2. gxf3 Rxh3) - my Queen had gone on a temporary excursion to wipe Black's a pawn off the board - hence the awkward position.

I looked first at 1. Re3. Not bad, this protects the Knight and indirectly the h3 pawn as well. It also provides some cover for my King should it decide to go to the e-file.

Then.... I looked at 1. Rxe4, sacrificing an exchange with Rook for Bishop and pawn after 1. Rxe4 dxe4 2. Qxe4. The more I looked at it, the better it seemed. First, I had broken up Black's pawn control of the center. The Knight is excellently posted on f3, protecting all the necessary squares of h4, g5 and e5. The Queen on e4 stops all tricks of Black pushing the g-pawn.

So the exchange sacrifice went ahead.

1. Rxe4 dxe4
2. Qxe4 Re8 (see 2nd diagram)

Hmmmm.... what now? Black realises that there is no way through on the h-file but decides to plant his Rook on e8 (?).

I half expected Black to contest the d-file with 2... Rd8 instead after which 3. Re1 (wisely avoiding the trade but having to give up the d-file). At some point Black cannot wait to trade off the Queens so that Black's Rook can be dominant.

I needed temporary control of the d-file but I needed to do something better. So I hitched upon a move that encompasses 2 different ideas.

So I played 3. Qg4 (see 3rd diagram)

Attacks the undefended Rook but more importantly, it allows my Rook to invade the 7th rank at d7 and attack the c-pawn and the Black Queen - a double attack and the c-pawn will fall.

3.... Rh6
4. h4! Rb6
5. b3 Ra8 (see 4th diagram)

In this position, I played

6. hxg5.

A mistake. See? Mistakes are now starting to creep into my play. I was concerned with

6. Nxg5 Rg6 (a Rook pin)

It is so unnatural to allow a pin unto yourself and I thought it was an uncomfortable pin that might cause me problems later. However, I failed to see that

7. Rd7 any Black Queen move
8. h5! and e6! are very strong replies and there was nothing to be afraid of.

6.... Rxa2
7. Rd7 Qe8
8. g3 (forced) Re6
9. Rxc7 Qd8? (see 5th diagram)

I had a chance to clarify things with 10. Qxe6! Qxc7 11. Qh6+ and achieve a winning endgame with such a pawn majority.

And now my ghosts are starting to haunt me.

10. Rxc5?? (see last diagram) A blunder of humongous proportions!

What I saw was that Black didn't seem to have anything after 10.... Qd3+ but there was something critical I missed.

Can you see how Black can win in this position?

My opponent didn't see it either and lost in a few moves later.

Answer can be found by highlighting between the brackets.
[ 10. .... Qd3+ 11. Kg2 Qe3!! and the White Rook on c5 is lost 12. Qd4 does not work because of Rxf2+ and the Knight is lost and the Queens are coming off in the next move.]