I'd be OK with 20.Bg2 Qf5 eg 21.Qc2 Qf2+ 22.Kh1 Qe2 with a messy position but indeed 17.Ne1 avoids this counterplay. When I feel like (and that's unpredictable) I'll take a look at 13...a5.

My interest is not merely abstract; I would like the Dutch to be playable (ie enabling me to play for a win, even if I would be objectively a bit worse) but run into too many lines that leave Black struggling for a draw. It's like the Kieler Gambit.

Stefan Buecker wrote on 03/04/18 at 08:01:34:

Like Brabo I think 19.Ne5 a better try. So I wondered about 16...Rad8 (I don't really understand 16...Rf6) and 17.cxd5 Qxd5 18.Bg2 Qd6 19.Ne5?! is met with the surprising b5! After 19.Ne1 Black wins a tempo with Rxf1+. Best seems 19.Qc4.

brabo wrote on 03/07/18 at 10:58:51:
Here I sugget 18...Rxf1+ (iso 18...Bf6) 19.Bxf1 Rf8 20.Nd3 Qf5 21.Nf4 g5 22.Bh3 Qf6 23.Nf4 Kg7 with a complicated position (but I don't dispute "There remains a slight disadvantage for Black in these variations").
Unfortunately I have nothing to add to the interesting 13...a5 lines at the moment.

brabo wrote on 03/07/18 at 10:58:51:

Don't you think analyzing and trying to understand complicated lines always will let you play better, even if you'll never play those particular lines?

Stefan Buecker wrote on 03/04/18 at 08:01:34:

Hi Stefan, thanks for your excellent analysis. You clearly put quite some effort in it as it definitely improves on what I already analyzed which already took quite some of my time.
Now I hadn't intended to go in correspondence-analysis-modus as this was just some fun (yes I do analyze openings while no intention to ever play them which again contradicts that I think studying openings will let me play better). Still your work deserves a reply so I went the last couple of days through a myriad of very complicated lines.

13...Bb7 14.Rc1 Qd7 15.Qa4 a5 (This is an improvement but white still has some advantage)
16.a3 Rf7 17.cxd5 exd5 18.Ne1 Bf6 19.Nd3 Nd8 20.Qxd7 Rxd7 21.Nf4 c6 22.h4
16.Bh3 Rf6 17.cxd5 Qxd5 18.Bg2 Qd6 19.Ne5 (I like this more than your 19.Ne1) Rxf1+ 20.Rxf1 Nxe5 21.Bxb7 Rf8 22.Rxf8 Bxf8 23.Qe8

13...a5 (This is indeed more accurate. I already had noticed this move in several lines but as stated earlier didn't dig deeper until now)
14.Qa4 Bb7 15.Rac1 Qe8 and now 16.Bh3 and 16.cxd5 exd5 17.a3 Kh8 18.Qc2 are very difficult to evaluate properly but definitely are no easy equality for black
14.Rc1 Ba6 15.cxd5 Bxf1 16.Qxf1 Qxd5 is also very interesting. The engines contradict each other.
14.Rf2 Bb7 and now I looked at 15.Be1, 15.b3 and 15.Qb3. Black has to play very accurately against all three lines but possibly can equalize on the long run.

So overall conclusion yes I do have to admit that I am not sure anymore if this special anti-classical Dutch-line guarantees an advantage. For sure black will be put under pressure but a perfect defense will maybe hold.

Stigma wrote on 03/04/18 at 19:13:46:

My experience of using standard games from e.g. the megadatabase to study an opening is that in 95% of the cases the played moves won't improve upon the analysis of an engine. Besides I am only looking at games of +2300 players. Nevertheless I still think it does make sense to also look to those "weak" games not from theoretical but from practical point of view. Some of the refutations are far from evident and spending some time with an engine to detect them can become later useful for practice.
Concerning the game Sbarra-Castaldo I had in my notes 14.Qd3 c6 and now the novelty 15.Bf2 with a clear advantage for white.

MNb wrote on 03/04/18 at 16:23:19:

It's not silly at all. This line seems no worse than 9.Nd2, it has a great score in practice, and it allows White to skip some sharp theory. The drawn game you mention is one of the very few times White failed to win with 9.Ne1 d5 10.Be3 on my database, and even there White had a nice advantage when he took the draw.

MNb wrote on 03/04/18 at 16:23:19:

Exactly. I'm happy that 9.Ne1 d5 10.Be3 hasn't been discussed more seriously and become better-known; I could benefit from that from both sides of the board. Maybe Williams and other 7...Ne4 specialists have known about it and taken the same attitude. But it's still surprising nobody is discussing it when it's been the choice of strong GMs like Nogueiras and Sumets.

brabo wrote on 02/28/18 at 19:22:02:


That's not strictly true: If the Gustafsson-Carlsen is given correctly on your blog, the move order there was 7...Qe8 8.b3 a5. But after 7...a5 8.b3 Williams, Reinderman and other strong players have preferred not 8...Qe8, but 8...Ne4, and only after 9.Bb2 Nxc3 10.Bxc3 does Black play 10...Qe8. In The Killer Dutch Williams even departs from the standard 10...Qe8 (because he concedes Avrukh's 11.Ne1 is slightly better for White), suggesting 10...Nd7!? instead. I guess you're aware of all of this.

In the 7...Ne4 8.Nxe4 dxe4 main line it's fun to analyze the sharp lines with 9.Nd2 d5 10.f3. But why can't White calmly delay f3 with 9.Ne1 d5 10.Be3, followed by either f3 or by Rc1 and a later f3? This setup has been known since at least 2004 and has scored very well for White in OTB games, but both Simon Williams and this forum keep ignoring it as fas as I'm aware. I could have brought this up years ago, but decided to wait and see if anyone else mentioned it.

No, it's not a sacrifice as Black cannot prevent White from winning two pawns back: 18.Ne5 Rxf1+ 19.Rxf1 Qc8 20.Qxa7 (not necesserily best) and later pawn c4 will fall as well.
16.cxd5 exd5 17.Ne1 Rxf1+ 18.Bxf1 Rf8 19.Nd3 Qf5 allows Black counterplay.
So 16.Ne1 looks best. Black's problems on the queenside don't run away and Bg5 17.Rxf8+ Rxf8 18.Nd3 Rf7 19.cxd5 exd5 20.Nf4 gives Black the kind of passive position I'd want to avoid with the Dutch.

Quote:

You can by robbing Black from all counterplay.

Quote:

In corr. chess I don't play the Dutch anymore either.