using Stockfish to find opening/middlegame ideas

This post is about using engines in a slightly different way.

A classic way is to ask the engine to analyze your game. and Stockfish will point out the best move or the errors. Chess.com does this as well in the Game Report function.

I use Chessbase and Stockfish12, I do not have the hardware to run LC0,LeelaChessZero unfortunately.

A concrete example now. My statistics against the English 1 c4 are not that great. My opponents play fast, get pressure on the queenside and I score lower than against 1.e4.

Here's the position I pre-selected.

Position after 1. c4 e5 2. Nc3 Nc6 3. g3 g6 4. Bg2 Bg7 5. d3 d6 6. Nf3 f5 7.O-O Nf6 8. Rb1 O-O 9. b4 h6 10. b5 Ne7 11. Qb3 Be6

Now this is the trick.

I set up a game between Fritz 17 with White and Stockfish12 with Black. 5 minutes + 2 seconds.

And I let the engines play from that position.  I know that Fritz17 is weaker than Stockfish12 even though Fritz17 can do a decent job analyzing my games.

Now I don't let them run 1 hour to analyze at Depth 50 for the best move here, I just want to know if Black can generate attacking chances, which I am failing to do in my own games.

Either this is better for White and I should play a different system ( like 1c4 e5 2 Nc3 Nf6 3 g3 d5 or 3..Bb4) or I should learn from engines.

_____________________________________________________________________________________________________

Of course I could play Stockfish10 vs Stockfish12 but chances are the engines versions will neutralize play for a long time.

I also can play Stockfish 12 with White against Fritz 17 and see how White can go through.

Finally I can play Stockfish 12 on both sides for 10 games, and see a lot of draws but look at optimal play which tends to generate more draws.

_______________________________________________________________________________

the point is that Fritz 17 at 5 minutes + 2 seconds is vastly superior to the opponents I am facing ( I do not play GMs or IMs every day).

And here's how the game went.

Game reached the following position

now here Fritz17 play 19gxh4. That's surprising.

maybe a human player may choose 19.b6 and here Stockfish's answer: 19..ab6 20 ab6 h3! 21 Bh1 c5! 22Nc3 e4! and Black has all the chances to break through. If you have more time, you could set a 2nd game from 19.b6 and learn from Black's play.

lets follow the engine game

Black has made great progress on the kingside.

Here's the brilliant finish with a sacrifice by Stockfish12 with Black. a fierce attack !

I think we can learn a great deal from an engine-engine match, as opposed to let the computer think for 5 minutes on each move and follow the 'best' line. practical play rarely sees best line on each side. 

Of course I recommend you use this technique on the opening position you care the most about. and feel free to share what you found out !

I also think that the game had natural moves except perhaps for 19gxh4.

__________________________________________________________________________________

Now does this mean this system is bad for White ? not at all !

Here's the reverse game, Stockfish12 with White from the same position (tabyia)