I'd just spend a good while with a modern SF and friends! It just looks to be the sort of strategically and tactically messy line that NNE engines play much better than anything else before.

Also very good at showing just how hard chess often is. It gives about +0.5, yes, but it still gives white only about +0.5 in a huge mass of lines 5-10 moves deeper in. 

And those positions aren't really any simpler. Trying to tame it down to a deep main line or two isn't going to work.

11..o-o is rather more straight line ish, especially the lines with that temporary pawn sacrifice on e4.

So there's huge scope to have interesting games. Which is good if that's what you're after, less good if white is going to be preparing very specific, concrete, ideas to play against you!

I like running engine tournaments with a mix of neural net and non neural net engines in. If, eg, SF16 can beat SF10 as black then it's often an indication of a good line to try in practical games.

3...Nf6 lines were once a great favorite of mine. Theory has shifted towards 3...c5. In the last five years 3...c5 has been played more often and scored better.

I still look at the 3...Nf6 lines from time to time. In the main line (1. e4 e6 2. d4 d5 3. Nd2 Nf6 4. e5 Nfd7 5. Bd3 c5 6. c3 Nc6 7. Ne2 cd4 8. cd4 f6 9. ef6 Nf6 10. Nf3 Bd6 11. O-O), Black has a choice between between 11...Qb6, 11...Qc7, and 11...0-0. I've mostly looked at 11...0-0, which seems to stand up to engine scrutiny.

11...Qc7 is more aggressive and has accumulated a lot of published research. My general impression was that Black is falling short. However, I was surprised to see that it has done pretty well in 2024. Looking further, it seems to do well if both players are close in rating, with a difference of less that 100 points. It's a little puzzling as engines like White's chances. Maybe engines just don't see far enough and should be ignored.

White's continuation 12.Bd2 O-O 13.Rc1 scores well for White in a small number of games. Perhaps Black hangs in there with something like 13... Bd7 14. Re1 Rae8 15. Qb3 Qb8 16. Ng3 Ng4.

I also looked at 12. Bg5 O-O 13. Rc1 Bd7 14. Re1 h6 15. Bh4 Be8 16. Bg3 Bh5 17. Ne5 Be5 18. de5 Ne4 19. Be4 de4 20. Qb3 Bg4 21. h3 Bf5.

These lines and other messy lines generally favor White. However, if it looks like Black is hanging in there, he probably is.

Recently I drew as Black a game that began with



Because of general embarrassment I won't give the rest of the moves.
My question is what the experts think of 12.Bd2 and 13.Rc1 ? It's neither in Williams nor in Moskalenko.