Analysis of the Elitserien games - Part 3/3

It is time to show my best game from the beginning of Elitserien.
Of course that was my victory against young Swedish GM Nils Grandelius.
If you are interested in his opinions about chess, chess politics this blog and so on than you can read his interview on this blog :
Grandelius interview - Part one
Grandelius interview - Part two
The game will not be commentated on usual way, but this time from the point of view of chess psychology and it will be some details about opening preparation.

Bejtovic Jasmin E2387 - Grandelius Nils E2593

How to prepare for the game against somebody who knows you so well, maybe in chess sense better than I know myself?
Is it good just to play my usual opening and don´t think about psychology?
Maybe is good to play something new, which nobody knows well?
It depends on what you want to achieve. My problem before the game was to make a clear goal for this game. That´s why I thought 3-4 minutes on my first move, and as much on my second and third move.
Nils did the same, which told me that he has not clear plan what to do.
We both got what we actually wanted to play (I guess). I got a position which I thought that I can play on decent level, and he got a position which was by no means worse for him.
I am talking just about introduction moves.
The real question which I would like to explain to the readers came in the next position:

Before I explain things behind Black´s seventh move I would like to tell something about opening system.
I played The English Opening and Nils choose so called Fischer setup.
The most famous game in this system is Petrosian-Fischer when glorious chess legend Fischer showed ingenious manoeuvre with changing the place of his minor pieces which controlled black´s pased pawn ( with his bishop on b7 and knight on c6 Fischer played Na5-Bc6-Nb7).
The most recent example of the clash in this variation is Radjabov-Ivantchuk game where Radjabov didn´t castled, but instead played with an early h4 (trying to exploit the fact that black´s knight was not on natural square f6 but on e7).
One of the serious tries against this system is Marin´s recommendation in his book on The English Opening, but for me it is somehow not practical.
However, he explains position very well and I can do nothing but recommend you to read it.
I want to point out one very important thing for practical players.
When preparing for the game many of them (some of them which I know very well) are looking at the statistics of some moves or positions.
In this position clearly best statistic has a move:
7...Nf5
...which was played by Grandelius (albeit, he probably didn´t know this during the game).
When I say, clearly the best I mean that Black won majority of the games.
This that means that one should play Nf5 as it is the best and showed in practise?
Of course not!
If we look at the statistic just one move further than we can see that my next move:
8. Ne4
...did very well in practise.
It can tell us that 7...Nf5 did well in practise only when White failed to find the best reply, but my opinion is that even this is not true!
A statistic has been made by results of the games and not by results of the openings (which means that a late blunder or just better play in middlegame or an endgame can influence this), and we should use our common sense when preparing for the game.
The reason why 7...Nf5 did well is just because an alternative
7...0-0
 leads to a drawish position after 
8. d4, cxd4
9. Nxd4, Nxd4
10. exd4, d5
11. cxd5, Nxd5
12. Nxd5, exd5 =
...and that means that mostly better players played 7...Nf5 in order to avoid a draw against the lower rated players. It doesn´t mean that they were better after Nf5, just that they succeed in outplaying their opponent later on.
However, White got some initiative and in the post analysis both Nils and I thought that black could get an unclear play in the following position:

We both thought that Black should play:
10...Nfxd4
11. exd4, Bxd4
...and that White should have a compensation for the sacrificed pawn but probably not more.
We got the position from the diagram in the introduction.
When I was checking the game with the computer engine, Houdini showed that White is close to winning.
12. Be3!
...and it is almost all over!
If Black plays:
12...e5
only then:
13. Bg5
with a strong attack.
What else he can do?
12...Bxe3
13. Nxd6+, Ke7
14. fxe3, Qxd6
15. Rxf7+
and wins.

This is the moment when I took an initiative and didn´t lose it till the end of the game (altought I made some inaccurate moves).
13. c5, dxc5
14. dxc5, Qxb2
15. Rb1, Qxa2
16. g4
This is the critical moment for the fate of the game.
Nils played:
16...Nh6 ?!
and he never get back in the game again.
As we found after the game, and what engines confirmed:
16...Nd4 
...was the best try.
White is better but I did found nothing conclusive.
I have to mention one more thing in my play which can be criticised.
When I was very near to win the game I didn´t play in the most practical way.
Even if computers likes my solution (with taking a pawn on b7 and playing for the mate attack) I did lost control over the situation and got very bad time combined with good resistant of my opponent.
It could be fatal in the time trouble. This time I won, but I know that this kind of unpractical decisions would cost me to much in let´s say 5 similar games against such opposition.

This is the position I am talking about.
If I just played:
24. Rfd1
or
24. Bxe6, Qxe6
25. Rbd1
Nils would (as he told after the game) just resign.
I played:
24. Bxb7, Bxb7
25. Rxb7, Qc6
26. Bxh6+, Kxh6
27. Qxf7
which looks very good, but White needs to be very accurate to win in the time trouble.
For those who wants to go trought the game there is a chess viewer.