Round 4 vs AMARS: Board 1

It was the shortest of tournament games I ever lost, so a new personal record. In a hurry I prepared a couple opening moves. This made sense only because the record shows Bill is quite predictable. To start out with we quickly reached the desired position, the opening being the Semi Slav:

With black to move I played Bg4 a sideline suggested by Bronstein. The main line is Bf4.
This side line is probably weak for black but Bronstein would not have suggested it were it not for the fact that

1. it provides black with tactical opportunties
2. its refutation is not immediately obvious.

The idea is black entices white to the following natural maneuver

1. Ne5 with an attack on g4 and c4
2. white pushes f3 to push back the g4 bishop and support his own e4 push

Meanwhile black playes the following rather unusual moves

1. retreat his f6 knight to d7
2. then push e5 

which opens the black queens diagonal and allows for a check on h4.
All this worked out as hoped for and soon the following position was reached.

Here the best continuation for white is Ne4 or Be3 neither of which is immediately obvious. However the natural

1. dxe looses a piece to Qh4+ followed by Qxc4. 
2. e4 is weak in light of Qh4+ and after g3 the queen drops back to f6 with a double attack.

My oppnent did play e4, and I knew this was not correct, so here was my chance to materialize but alas my mind was too fuzzy and I played Qf6 without Qh4+ first. I must have been pretty damn tired and didn't even consider his most obvious reply Be3, which backs d4 and renders the h4 check pointless. So after Qf6 and Be3 the computer sees white as being better by about 2 pawns. 

What is the lesson learned?  If not familiar with the opening, and I am a novice to the Slav, don't play tricky lines, you might fall vicitim to the trap you prepared as I did.


If you want to see a much more detailed analysis, here is Bill Brock's analysis of what could have happened:
Brock analysis