Something else deeply amusing in its way - in the old line from Karajakin (et al as I remember?) - Yusopov, with:
8.. Nbc6 9 Qh5 Ng6 10 Nf3 Qc7 11 Be3 Nce7 12 h4 Bd7 (12 .. f6 there is reasonably playable too it seems, as is 12.. Nf5) then 13 Rb1 isn't all that strong.

Black has a horrible looking resource - 13 .. c4 14 Bxg6 fxg6 15 Qg4 and then 15 .. h5!? (yuck!). That seems to give black time to get organised to have a good chance.

Cf 16 Qh3 b5 17 g4 a5 which is enough to provoke it into 18 o-o as white. Then so on.

It somewhat prefers going 13 Qg4 cd 14 cd f5 15 Qh3 Bb5 16 Bxb5 Qa5+ 17 Bd2 Qxb5 18 h5 Nh8 19 h6 g6 which does look a bit annoying for black if presumably quite solid.

It's actually fascinating looking at these (sort of) old lines with SF16 and friends - none of them are absolutely brilliant but there's a fair few that seem to be quite holdable.

That line from the update, but also things like:
8 Bd3 Nbc6 9 Qh5 Ng6 10 Nf3 Qc7 11 Be3 c4 12 Bxg6 fxg6 13 Qg4 then a couple of fascinating things.

Firstly it points out that black can (amazingly) get away with 13 .. Bd7 14 h4 Rf5 15 h5 Rxh5 16 Rxh5 gh 17 Qxh5 Rf8 18 Kd2 Be8 19 Qh3 without any exchange sacrifices.

It thinks Rxf3 might be OK mind, but prefers just going h6, ditching e6.

Or 19 .. Nd8 20 Rh1 h6!? and then if 21 Qg3 h5!?  Looks utterly suicidal, is risky!, but isn't.

The other thing is that after 13.. Qc7 14 Ng5 Qe8 15 h4 Ne7 16 Qe2 Bd7 17 g4 everyone has been doing it a bit wrong with black.

NOT 17 .. b5. Instead 17 ..a5 18 Kd2 Rb8.

With the point that if white goes for the desired 19 h5 gxh5 20 Nxh7 Kxh7 21 Rxh5+ Kg8 22 Rah1 black still has 22.. Ba4 and is then apparently OK in the serious mess  eg 23 Rh8+ Kf7 24 Rh8h7 Bxc2 25 Qf3+ Bf5 which are some serious heroics from the bad bishop!

I can't see an especially strong waiting move for white either, so it might be moderately effective. Of course 9 Bg5 iso Qh5 is still very annoying! Much narrower for black too.
(9 Qh5 is actually SF16's third choice, the other being Nf3 f5 Qh3)

Nernstian59 wrote on 01/09/24 at 00:23:23:


A *differently* hair raising one, please

There's a lot of different lines, if you ask the modern strong computers - much improved in these sorts of positions - they can iirc (as far as I can tell) hold many of them.

Often a  bit of a struggle. Fairly terrifying to prepare as white though as there's a wide choice for black and the positions stay very, very complex for a very long time.

The January 2024 update has several poisoned pawn games with 7...0-0. I'm no expert in these lines. I followed one line with 1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. e5 c5 5. a3 Bc3 6. bc3 Ne7 7. Qg4 O-O 8. Bd3 f5!? 9. ef6 Rf6 10. Bg5 Rf7 11. Qh5 g6 12. Qd1 Nbc6 13. Nf3 Qa5 14. Bd2 Qc7!? 15. dc5! e5 16. Ng5 Rf8 17. c4! e4! where it seemed that Black was in the game. There are several other Winawer games that I think could be mined by Winawer (especially 7...0-0) proponents.

I can't see how black can save himself after Nd8.

I don't think anyone has ever claimed all of the 7 Qg4 o-o stuff to be busted, or for that matter even all of 8 Bd3 Nbc6. The specific subvariation of that stuff that Watson had does seem dead but plenty of other options.

Khalifman in the opening according to Anand books doesn't give more than +/= (which he has to claim vs everything of course ) against several of blacks main tries.

This should be winning for white instead of Ng6.

The java format is easy - there's a PGN button in the editor window which inserts a pair of pgn  /pgn tags (with square brackets round them), and you just paste  the pgn in there.