Today we have a rest day at Corus. A nice moment to brainstorm about openings. Don't expect much cohesion.
Four years ago I decided to commit the youth-sin of playing gambits. The reason for that was to gain tactical experience. Usually a grandmaster is granted a year in his chess-youth to commit this sin. Since my development is slower than a grandmaster, my chess youth lasted longer.
It took me four years to learn what can be learned from gambits only.
But now it is time to move on.
I never developed a special affinity with gambits. I played gambits because it was a logical thing to do at that stage of my development. Yesterday I played 6 fast games against a 1950 player.
I won 2 of them by tactical crushing him and I lost 4 due to well positional play from his side.
It convinced me that my tactical skill is at a 2000 level already.
But tactical means are blunt by nature. When opponents keep a keen eye at their pieces, they won't lose them to you without notice.
So the battle has to be won by subtler means first before to can finish your opponent off by tactics. It is just a logical step.
To learn a more subtle approach I need openings that are coherent with that idea. So I intend to replace my openings.
There are a few approaches from which I can choose.
If I play the moves that are played by most of the grandmasters, that can't be bad.
Just follow the mainstream.
Rybka is a pretty good engine. At my computer it can think about 22 ply ahead within a reasonable time. This approach means that novelties from opponents score always less.
A computer procudes second and third choice alternatives which often score about the same.
If I let statistics decide between them, I probably get a good mainstream line.
Openings like the 1.d4, the Caro Kan, the French etc. are very alien to me. Which probably means that I will learn a lot when I try to play them.
Think for myself approach.
Since I think that piece activity for yourself and lack of activity for your opponent should be the goal of any opening, it might be possible to find the moves according this principle myself.
What to choose?
All approaches include a decent amount of work. With traditional opening study it is very easy to get lost in the forests of variants though. The last approach (think for myself) leads to the most understanding, but it entails a certain amount of risk. I cannot know beforehand if it will give a useful endresult. It's value lies in the process and not in the result. The risk is that when it proves to be a dead end, I will have to take one of the other approaches. On the other hand I have my old repertoire as a safety net.
I'm inclined to go for thinking by myself (gee, aren't you surprised?). If I approach the opening the same way as I approach the middlegame (piece activity!) I can't believe the efforts will be in vain. Even if there is no usable end result my middlegame play will improve too due to the efforts. In fact I already did my first attempt.
My first attempt.
I looked at the begin position of the white pieces for a few hours, asking myself what would be the best future for each piece. Especially the development of the queen is problematic. I'm used to play booklines supplemented with my own fantasy lines so I always took the endresult of the opening as a random given fact.
This was the first result I came up with:
It is a Colle/London system like structure.
After I found this, I listened to Pete Tamburro's video how to defend against the Colle/London system. I found his lines very convincing. A crucial move is for instance g6, which reduces the power of the bishop and queen instantly. (Actually that screams for h4-h5, but where will the white king be save?) As black I always hated the Colle since I played e6, which leaves a nice place for a knight at e5. But Tamburo's suggestion is to attack in the center with Nbd7, c5 and e5. If the center breaks open, the white pieces are subject to attack. So in one day I learned that to play for a position without looking after the moves of the opponent is a dead end.
Of course I already knew that, but it never made a deep impression. Now it did.
Chess Training AI is Closer than We Think
3 hours ago