The Levitsky Attack is an Underrated Opening

The Levitsky Attack (1.d4 d5 2.Bg5) is an underrated opening.  Just watch Black get grinder down to the very end!

Qb3 by white is an interesting choice, to be honest I wouldn't even play Bd3 most times

If Black plays 2...Bf5, then 3.c4, but with ...c6 and e3 thrown in, it's different.

I mean obviously playing Qb6 is critical against Bg5, so I'm not sure why he didn't do that. But maybe that's why he's 1800.

Still, it wouldn't really be my first choice against a lower rated player, but maybe I feel that way now because I'm beginning to enjoy going into complicated lines.

Against 3...Qb6, I gambit the pawn.

1.d4 d5 2.Bg5 c6 3.e3 Qb6 4.Bd3 Qxb2 5.Nd2 Qb6 6.Ngf3 with compensation for the pawn. White links to castle, play Qe2, and e4.

that makes a lot of sense. I'm just very wary of gambiting any sort of material.

I always think of the diabolical mastermind Paul Robinson in Neighbours when I think of the Levitsky since we once learned that's what he plays (actually Trompowsky, but similar thing).

Anyway, a problem with these sorts of deviations is that in the computer age not only can people find good lines, they can analyze what if such and such, they can find ones that they like best and staying away from the mainlines might be a good thing to confuse the opponent. There's much less of a mystique about them.

As usual the computer is having a field day with most of my suggestions, but 23. ....f5, countering the central push would have gone some way to cutting out the attack. Then you can start considering Nxe5 and facing off the rooks.

Currently it is established that the safest and soundest defence to the Trompowski (1.d4 Nf6 2.Bg5) is 2...d5, as white's "positional threat" of Bxf6 is a bluff.

And of course, the same approach can be applied against the Levitsky, although 2...c5 and 2...f6!? (3.Bh4 Nh6!) are decent candidates.

Most people that are up to date with the theory don't play 3.Bh4?!. The hot button these days is 3.Bd2!