When reading Davies book I've just come to the DERL, where he recommends black a knight sac for 2 pawns which seems slightly dodgy to me, as black might need to take a perpetual early on in some side lines.

Actually, the DERL (5.0-0 Be7 6.Bxc6) looks quite interesting to me and I don't see why it's so much less popular than 4.Bxc6, as compared to the Exchange variation white has gained 0-0 while black's Be7 needs to move again to cover e5. And then d4 seems to have some more bite than in the exchange.

Anyway, I consulted Keres and he mentions that black can transpose to the Modern Steinitz with 6...bxc6 7.Re1 d6 - which looks like the more fighting choice to me. So knowing something about the Modern Steinitz can be quite useful even if you usually prefer the other main lines...

re Ruy Lopez, Modern Steinitz Defence

Last year I dared to suggest to a well-known British GM and chess author that there was a real gap in the market here - to my knowledge there has been no book in modern times treating the Modern Steinitzr from Black's point of view. Since it was a defence with a great past (Steinitz, Rubinstein, Capa, Alekhine, Keres, Smyslov, Bronstein...) It might just have a great future. All it would need would be for a great player of today to start scoring well with it for it to swing back into fashion.

I am afraid I was given an extremely sharp put-down; the idea was rejected out of hand on the grounds that he had studied the Modern Steinitz in depth some years ago and there were just too many problematic lines for Black, above all no really satisfactory answer to 5 0-0.

Re the Norwegian Variation: ChessPub gives Anand--Agdestein [10 Nh4] with the suggested improvement 16 Nf5!, and Pavlovic--Agdestein [10 Qe2] with the suggested improvement 12 ...de!, as both a bit better for White. Any improvements/updates here?

After 5 Bc6 bc 6 d4, Keres's 6 ...Bg4 looks to me an interesting alternative to the usual (and OK?) 6 ...f6. (And better than Keres's other line?: 6 ...ed 7 Nd4!.)

Re the Siesta Variation, there may be no need to give long lines, but the discussion could benefit from a bit more concreteness? What does OFWATA say is best here? I'm assuming 9 Bc2 Bc2 10 Qc2 Nf6 11 d4 e4 12 Ng5 d5 13 f3 h6 14 Nh3 0-0 15 Nd2 ef 16 Nf3 and now is 16 ...Rf7 17 Qg6 Bd6 best? -- as you can tell I don't really know this stuff! (Even if it is, doesn't it have the practical drawback that 18 Bh6 forces a draw?; and what of 18 Bf4 or 18 Nf2?)

Can Black sidestep all this with 10 ...g5!?, or is that too wild? It reminds me of a game where iirc Portisch massacred Korchnoi with an even earlier such thrust: 5 0-0 Bd7 6 Re1 g5!?.

Anyone interested in discussing my cheeky attempt (ibid.) to second-guess Mamedyarov?

Same here, suppose we're both old enough to remember the olden days when Sportverlag and Batsford were the only chess book publishers...

I think Keres used a temporary pawn sac in one of his games in the Siesta...

Willempie wrote on 02/22/06 at 22:53:41:


Yep! S bis F - though I just noticed that I also have his Vierspringerspiel bis Spanisch, though I suppose they're pretty similar in content.

No need to post any long lines. I was just curious as the Siesta resembles the KG a bit. Is that slight white edge a case of black 'probably not having quite enough for a pawn' or is it a solid positional plus?

This thread made me take some time off and revisit one of Keres' Sportverlag books on the Spanish. So far I don't have too many answers to contribute, but a few questions...

One line that looked attractive, but I don't see mentioned here is the Siesta Variation, invented by Capablanca it seems, i.e. 5.c3 f5 6.exf5 Bxf5 - is this still considered playable these days?

Another line suggested by Keres that intrigued me is: 5.Bxc6+ bxc6 6.d4 Bg4 7.dxe5 dxe5 8.Qxd8+ (8.Nbd2 f6 9.Qe2 Ne7 10.h3 Be6 11.Nc4 Qb8 12.0-0 Ng6 'and black stands very well Jimenez-Keres, Moscow 1963') 8...Rxd8 9.Nbd2 f6 10.Nc4 Be6 11.Ne3 Bc5 12-Bd2 Ne7 'with excellent black play in Klundt-Keres, Bamberg 1968.' Has white come up with anything here? Or is Bxc6 plus d4 just wrong??

And finally the line 5.0-0 b5 6.Bb3 Na5!? - could it really be this easy?! Well, probably not - but there seems to be a whole lot of strong Norwegians that play it! - Is there any particular theoretical antidote?

[quote author=Michael Ayton link=1138105012/15#20 date=1140449827]As has been mentioned in another thread, Mamedyarov’s adoption of the Modern Steinitz at Corus has highlighted some interesting new ideas, and there’s a very interesting summing-up of these in Olivier’s latest update. My first conclusion from this is that the Modern Fianchetto Variation (5 0-0 Bd7 6 c3 Nf6 7 d4 g6 8 Re1 b5 9 Bb3 Bg7) is under something of a cloud because of Kuzmin’s 10 Bg5! h6 11 Bh4 0-0 12 a4!. (Might 12 …Qe7 be an idea here, I wonder?)

The ‘compensation’ is that Olivier’s notes imply the Bronstein Variation (5 c3 Bd7 6 d4 g6 7 0-0 Bg7) is an interesting fighting choice, since after 8 Re1, 8 …Nge7!? is in business: 9 d5 Nb8 10 Bd7 Nd7 11 Be3 0-0 12 c4 h6 13 Nfd2 f5 14 f3, and now, rather than 14 …c5?! as in Polgar--Spassky, 14 …f4 15 Bf2 g5, as in Trujillo Delgado--Semenova, is obviously thematic and looks OK. I was pleased that I’d reached this conclusion independently, but Spassky’s play in that game did look a bit weird!

One question that might arise out of this is whether, in the Rubinstein Variation after 5 c3 Bd7 6 d4 Nge7, Black can usefully, after either 7 Be3 or 7 0-0, play 7 …g6 rather than 7 …Ng6. After 8 d5 can he launch the same attack just as effectively, or can White by avoiding Rf1--e1 save a useful tempo? (A further sophistication might be 7 …h6 intending 8 …g6, as tried in Anand--Yusupov after 7 Be3.) The question is far from academic, since should these lines be OK for Black, he could reach a Bronstein-Steinitz via a Cozio move order! Does Karolyi say anything about 7 …g6 or 7 …h6, kylemeister?

White’s other main try for advantage is 8 Be3!?. Here NCO gives Topalov--Azmaiparashvili which went: 8 …Nf6 9 Nbd2 0-0 10 de Ne5 11 Ne5 de 12 f3 Ba4 13 Qa4 Qd3 14 Rfe1 with a small edge, but Azmai seemed to draw easily enough after 14 …Rfd8. Pete Tamburro suggests something a lot more exciting: 9 …Ng4!? 10 Bg5 f6 11 Bh4 Qe7!?; he continues with (the never-played?) 12 d5 and claims dynamic equality after 12 …Nb8 13 Bd7 Nd7 14 c4 h5 15 Bg3 0-0-0 16 b4 g5 17 h4 Bh6 18 Nb3 Rhg8. Anyone got any thoughts on this? (11 …g5!?, 11 …h5!? and 11 …Ne7!? have all also been played here -- the latter by Smyslov, though he lost!)
[/quote]

Nah, Karolyi just talks about ...Nge7 followed by ...Ng6 (and White then playing d5).  

I would think that missing out Re1 would be of benefit to White in the King's Indian-ish case.  It's an interesting issue, how these positions compare to the actual King's Indian.  One might think that the Neo-Steinitz versions are just worse for Black, because the light-squared bishops are exchanged (generally unfavourable for Black in such a structure).  But comparing the tempi it seems that Black generally comes off better in the Neo-Steinitz case.  E.g., in the case of ...Nge7, ...Nb8 and ...Nxd7 Black has taken the same number of knight moves to get them to e7 and d7 as in the Classical KID; White took two moves to get his pawn to c4; perhaps Black got in ...a6 "for free," though it apparently has both good (preventing Nb5) and bad (weakening b6) sides.  Then again, it seems that White can generally get in Be3 and Nf3-d2 in the Neo-Steinitz case, which looks like an optimal deployment of those two pieces (and something White generally can't achieve in the KID).  Ah, and I suppose Black often has to use a tempo on ...h6 in the Neo-Steinitz.  So maybe the tempo situation is "a wash," unless White has played Re1, when Black has an extra tempo as "compensation" for the unfavourable bishop exchange?  Preliminary thoughts anyway.

Incidentally, I noticed that Mamedyarov continued to play the N-S at the Aeroflot tournament.  He had at least two games with it there.  One of them involved the same line (basically quick d4 without c3) in which Mamed lost to Leko at Corus (the Aeroflot game was about a 20-move draw, against a strong GM).  Another one followed the line you give as putting the Modern Fianchetto under a cloud, but Mamed's opponent (another 2600+ guy) didn't play 12. a4.    

No you're right that b5 and Be8 are inconsistant, especially because the bishop is not ready to contest the a2-g8 diagonal.

The line you give seems like best play, looks +/- to me the one game I found held out to a draw in a pawn down endgame:
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 d6 5.c3 Bd7 6.d4 Nf6 7.0-0 Be7 8.Re1 0-0
9.Nbd2 Be8 10.Bxc6 Bxc6 11.dxe5 dxe5 12.Nxe5 Bxe4 13.Qb3 Bd5 14.c4 Be6 15.Qxb7 Bd6 16.Ndf3 Re8 17.Qb3 h6 18.Bf4 Nh5 19.Nc6 Qd7
I really dislike this position though it may be drawable and is certainly better than 13 .. Bc6.
So it seems I stopped playing it for a good reason

Iirc that is indeed the problem variation. You could avoid it by flicking in b5 at some point though.
Here's what can happen without Qb3:
[Event "Paris"]
[Site "Paris"]
[Date "1933.??.??"]
[Round "4"]
[White "Znosko Borovsky,Eugene"]
[Black "Alekhine,Alexander"]
[Result "0-1"]
[Eco "C87"]
1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 d6 6.c3 Bd7 7.Re1 Be7 8.d4 0-0
9.Nbd2 Be8 10.Bxc6 Bxc6 11.dxe5 dxe5 12.Nxe5 Bxe4 13.Nxe4 Qxd1 14.Nxf6+ gxf6 15.Rxd1 fxe5 16.Bh6 Rfd8 17.Kf1 f5 18.Rxd8+ Rxd8 19.g3 Kf7 20.Be3 h5 21.Ke2 Ke6 22.Rd1 Rg8 23.f3 h4 24.Bf2 hxg3 25.hxg3 Rh8 26.Bg1 Bd6 27.Kf1 Rg8 28.Bf2 b5 29.b3 a5 30.Kg2 a4 31.Rd2 axb3 32.axb3 Ra8 33.c4 Ra3 34.c5 Be7 35.Rb2 b4 36.g4 f4 37.Kf1 Ra1+ 38.Ke2 Rc1 39.Ra2 Rc3 40.Ra7 Kd7 41.Rb7 Rxb3 42.Rb8 Rb2+ 43.Kf1 b3 44.Kg1 Kc6 45.Kf1 Kd5 46.Rb7 e4 7.fxe4+ Kxe4 48.Rxc7 Kf3 49.Rxe7 Rxf2+ 50.Ke1 b2 51.Rb7 Rc2 52.c6 Kg3 53.c7 f3 54.Kd1 Rxc7 55.Rxb2 f2 0-1

This is real food for thought -- thanks. I notice too that Varga has tried the ...Be7/...Bf6 plan but with ...b5 thrown in before ...Bf6, presumably to prevent the exchange of light-squared Bishops and so potentially strengthen Black's possible kingside attack, esp. if White should play h2--h3. (I'm always a bit apprehensive about launching such attacks if I don't have a light-squared Bishop, but I guess one shouldn't be dogmatic!)

Thanks, Willempie! I really must stop being such a hick and get hold of these essential books, particularly as you can get them quite cheaply! Up here in Durham you never, of course, see any decent chess literature in bookshops -- it's London or nothing if you want to see before you buy. (Thanks once again, Rupert Murdoch!)

I'd no idea the Keckcemet Variation involved the idea ...f6, ...Nd7 and ...Be8, nor that playing ...Be7 before ...Nf6 might be an idea -- interesting. Am I right in assuming we're talking about 5 c3 Bd7 6 0-0 Be7 7 d4 Nf6 8 Nbd2 0-0 9 Re1 Be8!? (instead of Portisch's 9 ...Re8)? This doesn't seem to do too badly! I notice also that 6 ...Be7 7 d4 Bf6!? has occasionally been tried, with reasonable success.

Vlastimil Jansa gave a short lecture on the Modern Steinitz in his interesting book Dynamics of Chess Strategy.  You can read the first fourteen pages of that book (the introduction and a taste of his discussion on the Steinitz) on Amazon.co.uk.  http://www.amazon.co.uk/gp/reader/0713486082/ref=sib_dp_pt/202-4126950-9479801#r....  Perhaps its not comprehensive, but you might find it interesting or insightful.  Hope this helps, good luck. Peace out.

[quote author=Michael Ayton link=1138105012/0#8 date=1138705901]Interesting reflections. The ...Nge7/...Ng6 stuff strikes me as an attractive system if it can be made to work. I guess one attraction might be that it can be reached from a Cozio move order, bypassing the Exchange Variation: 3 ...Nge7 4 c3/0-0 a6 5 Ba4 d6, etc. (Of course, one would need something against 4 d4. Maybe 4 ...ed 5 Nd4 Ng6!?)

I'd be interested to know what lines/assessments Karolyi gives after 8 d5 Nb8 9 c4 h6!?, or 9 ...Be7 10 Nc3 h6!?, if he does.[/quote]

The picture that emerges from the article is that after 8. d5 Nb8 9. c4 Be7 10. Nc3, 10...0-0 is apparently better than 10...h6.  10...h6 is attributed to Keres and was successfully employed by Timman against Motylev, but Timman himself is cited as thinking that 10...0-0 is better (actually they have the evident misprint "10...0-0-0"); apparently Motylev missed a chance to be clearly better.  There are several games/excerpts with 10...0-0, which is portrayed as leading to plus-over-equals with best play.  9...a5 and 9...Bxa4 are also given (with a full game for the former and a couple of citations for the latter) as leading to the same evaluation.

[quote author=Michael Ayton link=1138105012/0#5 date=1138449696]Thanks for this, kylemeister. I'll take a look at the Yearbook. I'm not quite sure which line(s) with ...Nge7 and d4--d5 he's covering in Part I. Are these ...g6 lines? -- maybe you could let me know the move order?

Portisch, in the book you mention, recommends the ...Be7 system I mention in question (3) of my post. He does also recommend the ...g6 line, but with ...Qe7. Davies, however (on ChessPub), suggests ...Qe7 is inaccurate before/unless White plays Nb1--d2, because of the plan d5, Bxd7, c4 and Nc3; instead he suggests 5 0-0 Bd7 6 c3 Nf6 7 d4 g6 8 Re1 b5 9 Bb3 Bg7. The implication is that the line Emms gives as slightly better for White in NCO (p. 334, note 12), namely 10 de Ne5 11 Ne5 fe 12 Bg5, isn't a serious danger for Black -- see e.g. Schekachev--Malaniuk -- and that Keres, who played this system quite a bit, knew this.

In the 5 Bc6 bc 6 d4 variation (C73), Davies also suggests 6 ...ed!? and 6 ...Bg4!? for Black, ascribing both lines to Keres. After 6 ...ed he gives 7 Nd4 (7 Qd4 c5! 8 Qd3 Ne7!) c5 8 Nf3! (Portisch gives this move as best, but without any analysis) Nf6 9 0-0 Be7 10 Nc3 0-0, but here he doesn't mention Emms's 12 Bg5.

As a general remark, I often get the feeling that quite a few people on here neglect the stuff on ChessPub. There's some very interesting material there -- I have the ChessPub shortcut permanently on my desktop!



[/quote]

Wow, you're quite up on this.  Interesting stuff.   Karolyi's point of departure is 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 d6 5. 0-0  Bd7 6. c3 Nge7 7. d4 Ng6 8. d5.  I was thinking that ...Nge7 implies ...Ng6; it seems to me that combining ...Nge7 with ...g6 is "just not done," but I'm not sure why (though some vague possibilites come to mind).  Yes, maybe I should actually look it up ...

It seems unclear to me whether there is a "main line" in this stuff now, regardless of Karolyi's remark that I quoted.  I remember that in the old days (say 1970s-80s) when White played the 5. c3 (rather than 5. 0-0) move order and Black played ...Nge7, the main line seemed to be 5. c3 Bd7 6. d4 Nge7 7. Bb3, forcing Black to spend a tempo on ...h6;  then Black tried to make use of that by means of ...Be7-g5 while delaying castling himself (for reasons pertaining to the possible opening of the h-file) ...




   

Thanks for this, kylemeister. I'll take a look at the Yearbook. I'm not quite sure which line(s) with ...Nge7 and d4--d5 he's covering in Part I. Are these ...g6 lines? -- maybe you could let me know the move order?

Portisch, in the book you mention, recommends the ...Be7 system I mention in question (3) of my post. He does also recommend the ...g6 line, but with ...Qe7. Davies, however (on ChessPub), suggests ...Qe7 is inaccurate before/unless White plays Nb1--d2, because of the plan d5, Bxd7, c4 and Nc3; instead he suggests 5 0-0 Bd7 6 c3 Nf6 7 d4 g6 8 Re1 b5 9 Bb3 Bg7. The implication is that the line Emms gives as slightly better for White in NCO (p. 334, note 12), namely 10 de Ne5 11 Ne5 fe 12 Bg5, isn't a serious danger for Black -- see e.g. Schekachev--Malaniuk -- and that Keres, who played this system quite a bit, knew this.

In the 5 Bc6 bc 6 d4 variation (C73), Davies also suggests 6 ...ed!? and 6 ...Bg4!? for Black, ascribing both lines to Keres. After 6 ...ed he gives 7 Nd4 (7 Qd4 c5! 8 Qd3 Ne7!) c5 8 Nf3! (Portisch gives this move as best, but without any analysis) Nf6 9 0-0 Be7 10 Nc3 0-0, but here he doesn't mention Emms's 12 Bg5.

As a general remark, I often get the feeling that quite a few people on here neglect the stuff on ChessPub. There's some very interesting material there -- I have the ChessPub shortcut permanently on my desktop!

RotGut wrote on 01/27/06 at 23:15:23:


There's a section of an old (1977 or so) book, How to Open a Chess Game, in which (Hungarian GM) Lajos Portisch advocated the Neo-Steinitz for Black.  I'm pretty sure he recommended the fianchetto line, and against 5. Bxc6+ bc 6. d4 he recommended 6...ed instead of ...f6.  As for how it holds up, all I can say is that books such as ECO and Nunn think that with best play it should be "plus over equals." 

Well, there's an article in the latest NIC yearbook about the lines with 6...Nge7, where White castles and plays d5.  It's written by Tibor Karolyi; his overall view is "slightly better for White."  He says that in part 2 he will cover "the main line" 8. Re1 Be7 9. Nbd2, in which he thinks White is slightly better, but Black is close to equalising.