Non human 960 openings II

To play a 960 game against Stockfish means to open a window to another chess world. Human players have begun to play with classical position, study scientifically openings and win familiarity with some structures of pawns and pieces. As a result, our thinking is based in patterns. Sure that in 960 there are a lot more of those patterns, and, all of them unknown. So most of us try to play structures that have some similarity with what we already know. I will not enter in epistemology. If we are lost, we look for our home. Stockfish does not get lost, no home to look for. It plays without prejudices and preconceptions. It makes that module finds new plans, ideas and new positions 'where no one has gone before'.

Position 816

bbrknqrn/pppppppp/8/8/8/8/PPPPPPPP/BBRKNQRN w KQkq - 0 1

A similar position to 351 presented in first part of current blog, Non human 960 openings I
both bishops in corner and h pawn in here is weak. 816 is not the mirror of 351, but both plans are the same. We will examine a symmetric defense.

1.c4 c5 , if 1...Ng6 is possible to play like in the first part of blog 2.c5. Now, after 2.Bh7 Bh2 3.Bg8 Qg8 (it is not antichess) the game is balanced.
I will consider now 2.f4 and 2.g4. After 2.Ng3 g6 our chamber GM even gives an evaluation of -0,15

First one 2.f4. Absolutely logical, white is blocking black diagonal. Seeing that if 2...f5 3.g4 then black anticipates with 2...g5 which makes a tactical game since the second move

Position after 2...g5

I will consider 3.Bh7 or 3.g3

3.Bh7 Rg7 4.Bb1 (4.Be4 f5) 4....gf4 5.b4 And if you wanted a calm game, forget it. Here, my brain blows up with 5...b5. Not the best option, but possible. A move that I am not able to make it nor to understand it neither. Additionally, after 6.cb5 c4 !? removes the last track of racionality. A more natural move inside the chaos is 5...Rg7 6.g4, it is a position to play and the one who first melts they brain will lose.

Position after 6.g4

You can see the bar around +0,2 , gives us a balanced game, but a little unstable. A real minefield.

It is also possible to try...
3.g3 b6 4.b3 , preparing capture in h7 , 4...Qh6 and here what really shocked me was that after 5.fg5 Rg5 (5...Qg5 6.Bh7) Knight? what knight? 6.Bh8 is not the best move. 6...f6 and for example 7.Nf3 Rg8 8.Bf6 Nf6 9.Nf2 Bf3 10.ef3 Qh2 with no problems for black. Sillicon friend says that 6,Nf3 Rg8 or 6.Nf2 f6 7.Be4 is a better option. Personally I like 6...f6, looking for a good path for Nh8, Ng6 conflicts with pawn g3 and encloses rook in an unnatural position. Well, all those variations are unnatural.

Instead of 2.f4, it is possible the second one 2.g4, although it is original it is not a sharp variation and defense is simple. 2...g5 3.Bh7 Rg7 4.Be4 Nf6, (4...Bh2 is also a good choice that makes an amusing and highly dangerous symmetry 5.Rg2 Be5) followed by Bh2 gives a balanced position to black.
Nevertheless, white has an idea that tries to break position. After 5.Bf3 Bh2, the module proposes almost all possible ways to give the rook, 6.Ng3 6.b3 and the wildest 6.b4, and also the most civilized 6.Rg2 with very similar evaluations, I will analyze, following the purpose of the blog, 6.b4.
'What's it going to be then, eh?'

Position after 6.b4

As a human player, I must tell you it is difficult to play, difficult to understand and difficult to believe.
A credible variation:
6...Bg1 7.Qg1 cb4 8.Bd4, here our artificial friend gives a small advantage for black in all moves, except with the one that I would do conducting black pieces. I bet that I am not the only one that would move 8...b6 defending a7, 9.Ba8 Ra8 10.Qg2 Rb8 11.c5. White queen gets activity. For curiosity, notice that after 11...bc5 12 Bc5 a5? Evaluation is Mat in 10 moves! Starting with 13 Bb6 Kd8 14.Qb7, actually a5 is not just a bad move, it is a provocation. I must say as well that this such bold variation is better for black each time you go forward in analysis.
It is preferred 6.Ng3 Bg1 7.Qg1 Rh7 8.b3, simply developing pieces with an exchange down. With what I got surprised was that module evaluated some advantage merely because pieces coordinate better. 8...Qh6, trying to simplify. 9.d4 and all variations give compensation for white. Rudolf Spielmann, author of 'The Art of Sacrifice' would have enjoyed this variation. Even changing queens. 9...Qh2 10.Qf1 Qh3 11.Qh3 Rh3 12.dc5 Rc5 13.Bd4 Ra5 14.a4. This last position is not a firework of sacrifices, but maybe is harder for both sides a precise play. Black has material advantage and with no weakness on sight, however each one of his pieces are playing by itself while white is coordinating his pieces.

Position after 14.a4

What humans do about all this stuff ?
Sequence 1.c4 Ng6 is played in 10% of 24899 games played with 816 starting position, 1.c4 c5 2.f4 in 13 games, has a rate of 0,048%; followed by 2...g5 in 2 games. Finally 2.g4 in 0 games, nobody played it.

There is a third interesting move for white. In other similar positions seems to be stronger than in the current 816.

To be continued