6 .. Nfd7 in Saemish

 Anyone with ideas/opinions/knowledge to share on this line: it is the KID, Saemish, but instead of the standard c5-sacrifice or the other suspects (like a6+c6, or Sc6) there is this obscure idea. Getting the knight out of a possible e4-e5, opening up for the bishop, making a f7-f5 possible (after e7-e5?), or a c7-c5.

Anyone? Would love to hear any thoughts on a (according to me) difficult OTB opening. 

I'd say that anything is possible. But I'm not sure what sixth move white made? But depending on that as to whether or not to go ahead with this move.  Personally, I would either try and fianchetto the queen-side bishop if I weren't going to attack, or, 6...Nbd7. The best thing to do in this position, if you want to use 6...Nfd7 and then the e5 pawn break, is after you use the knight move to prepare the way for the e-pawn, is to play f5 attacking from the flank. A dirrect attack only works if you have superior forces. Also, realize that white is going to continue his development, so don't expect any sudden wins. The best thing in these kind of positions is patience. Don't make any sudden moves and don't get over anxious about your opponents moves. More than likely, its a hard opening for him too! (Just as a side note, Garry Kasparov won quite a few games as black with the positon that arises out of 5...e6.)

Thanks, I am thinking about the standard 6.Le3-variant, where many black players go for c7-c5. Fianchetto the QB seems slow, in terms of white building up the Le3/Dd2-duo + h4-h5 coming?

Do you think f5 can be played without e5 first? Thanks in advance

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.f3 O-O 6.Be3 e5 7.d5 Nh5 8.Qd2 Qh4 9.g3 Nxg3 10.Qf2 Nxf111.Qxh4 Nxe3 

I actually don't like the pawn sac with c5 much and the e5 lines is to theoretical and tactical. however what i do like is the variation where white declines the  c5 pawn sac and push d5. Now a question arose, can black prepare c5 without sacing a pawn and force white to move hes d-pawn?

Of course the first move that comes to mind is 6... Nbd7 but I didn't find this very attractive since blacks knight on d7 is in the way of it's partner on f6. And it more importantly allows white to hold d4 as a strong-point and never commit it to d5. Where i would like to have the b8 knight is on c6.

 So how can black do this? Of course! 6...Nfd7! This move preserves the option of blacks knight coming to c6 and opens up the bishop on g7 which also pressures the d4 pawn, so black is now going to follow up with c5 and Nc6 which will put a lot of pressure to whites d4 pawn and eventually force him to commit it.

Let's make a diagram about this:

I agree with Cymantex and I have studied Geller's book on KID Saemish, Geller also advised to play Nge2 instead of d5. The plan for white was to castle, play Kb1, g4 and h4

I'm not so sure that's better for white, after 12. Rxd2 Nb6 things seems pretty unclear to me. For example if white playes 13. Nxb6 axb6 14. d5 Ne5 15. Nc3 This looks good for white since he has a big center and is ready to push the black knight of e5 soon with f4, but, black of course has this thematic move 15... f5! which is a very effective move since it both has ideas of shutting down f4 forever by playing f4 himself and also sometimes to just take on e4 and when the knight recaptures he can play Bf5 which is often a very good square for the bishop or if the pawn takes the e5 outpost is secured. Gah, to much text again... I should probably have made a diagram directly.

I should probably also point out that the post I made doesn't cover any real analysis. I only wanted to show some of basic ideas, not the best play from each side.

...Nfd7 is very likely to at least be inferior to most other moves. Preferably ...c5 would want to be played immediately, and in my KID book it suggests that 6...c5 is a good sacrifice.

However, In the chess.com videos GM Roman Dzindiaschvilli (can't spell it), apparently the inventor of the sac, claims he refuted it, I believe by giving back the pawn eventually with a favorable position but who knows if that'll hold, but it probably took a lot out of him to refute his own move that so many people started using.

So I would personally avoid ...c5 rather than take a risk at playing a refuted variation. I like the panno (6...Nc6) which delays ...e5 for the moment and prepares ...b5 with ...a6 and ...Rb8, which can be useful for a few things, undermining c4, kicking out a c3 knight, if white castles queenside it attacks the king.

Wow, i had no idea about that, the inventor claiming to have refuted hes own move... Finally I have a good argument to use against those c5 fanatics! 

Many thanks to all of you, for this very interesting discussion.

To Elubas, I want to express my slight sceptisism towards Dzindzi's claim; he seem to display a "high variance" in his work, sometimes having excellent ideas but frequently releasing rather sloppy work. It just doesn't seem plausible that the c5-sacrifice is refuted; it is strongly recommended by many world players (Bologan, Gallegher, "Kazim", others).. Anyway, this is not the main topic, so it doesn't matter. Thanks anyway for the thoughts.

To Cymantex, thanks to your interesting stuff here; I am really going to look into it! i believe it is quite possible that this line in fact is a future main variant. This happened with ...Na6 in the classical KID, being a novelty in the 90s, suddenly being an equally playable alternative as the ...Sc6 (and Sbd7). There is even the possibility to play Na6 a move earlier (prior to e7-e5), avoiding the exchange variant in its most boring form.

Anyone; what is white's best try agains ...Nfd7? Is it the standard plan with Le3 + Dd2 + h4->h5? Any thoughts? Could we work out a few principled lines here?

It is interesting to see that Rybka 2.2 (downloaded it for free...) suggests exactly this line (6...Nfd7) as the best, although it is probably not a significant difference in the evaluation scores between the competing lines.  No! The evaluation just changed ... or, no, its back in the top ...  :)

Well, who believes an engine anyway, in strategical questions and opening theory... The line is certainly not easy to refute, thats for sure.

The question is what kind of positions we get from it. Fun ones (in true, dynamic KID spirit) of solid/technical stuff ...

To eaglex, this line is another alternative, but I believe better for white. And rather extreme, in some difficult-to-define way, IMHO. 

I have actually analyzed this variation earlier and I think i have possibly figured out the most challenging way to play this with white, unfortunately i haven't analyzed anything after finding that position. I would really like help with this.

Here's the analysis btw:

Because I made this analyze originally to myself it has some weird terms I should explain.

Pa = Possible alternative (if the main line end up refuted it's always possible to return to this move)
Tp = Transposition
Pp = Perpetual
Nma = Needs more analysis
=/=+ = Equal or slightly better for black
=/+= = Equal or slightly better for white

Basically whites strategy should be something like this:

1. Develop the Queen and knight before pushing the h-pawn

2. Push the h-pawn to provoke black to play the weakening move h5 before castling.

The point is that if white castles before pushing the h-pawn black can just skip playing the move h5 and focus on an attack over there. But if white keeps the king in the center it will be much harder for black to get hes queenside attack through and is forced to either go in to the slightly worse endgame or play the weakening move h5.

I like playing against the Samisch.

I dont like 6 c5, too theoretical.

I play Nbd7, a6 and c5, if d5 I play b5 switching to a Benko works for me. 

i havent lost to a Samisch in years

Yeah it's sensible to be a little skeptical, but again he's a GM, refuting his own move. I just looked at the video and specifically in the ...c5 line at one point black plays ...Nb4, and everyone thought Kf2 was forced, but it turns out white can sac the exchange and keep his pawn with a huge positional advantage and I think he wins a second pawn for the exchange anyway.

Believe me, I COMPLETELY understand how you don't find it plausible for ...c5 to actually be refuted. Remember this is very up to date analysis, so older books of course will say something different.

Ok, you may be right, I am not arguing against you. Bologan's book from 2009 didn't notice Dzinzi's refutation, however. Interesting. Bologan is one of the strongest players in the world.. But you may be right, it is possible.

Many thanks to Cymantex: I will post my thoughts as soon as I have digested the material.  

Didn't have time for deep analysis yet, but I got some impressions by playing out the position against the GNU chess (meaninless to try to get a playable position against Rybka or Fritz). Here are two attempts that didn't turn out too bad: 

1... Nfd7 2. Qd2 c5 3. Nge2 Nc6 4. O-O-O Qa5 5. d5 Nce5 6. b3 Nb6 (setting up some tactical possibilities versus the strongpoint c4) 7. Nb5 almost the only sensible move for white, but now he can forget about an attack on blacks king (7. h4? Nbxc4! 8. bxc4 Nxc4 black is winning) 7... Qxd2+ 8. Rxd2 Bd7 (possibly a6, see next game) 9. Nbc3 (9. a4!?) 9... f5 10. Bg5 Rf7 11. Kc2 h6 12. Bf4 g5 13. Bg3 f4 14. Bf2 and Black has a good game.

1... Nfd7 2. Qd2 c5 3. Nge2 Nc6 4. O-O-O Qa5 5.d5 Nce5 6. b3 Nb6 7. Nb5 Qxd2+ 8. Rxd2 a6!? I don't think this is the better option 9. Nc7 Rb8 10. Bg5 f6 11. Bh4 Na8 12.Nxa8 Rxa8 13. f4 Nf7 14. Bf2 e6 15. dxe6 Bxe6 16. Ng3 Rad8 17. Bd3 Nh6 18. f5 gxf5 19. exf5 Bc8 somewhere here, white is better. But there is probably improvments for black earlier 20. h3 Rfe8 21. Kb2 Re5 22. Re1 Rde8 23. Rxe5 Rxe5 24. Be4 b6 25. Rxd6 Very sloppy play by myself, in the latter part of the game.

A few games, I actually won. The engine went for a straightforward attack but the machine's timing turned out to be slightly wrong (which, however, was far from being obvious during the play) and blacks counter attack was stronger. In most games, however, I lost, after having had a playable position.

My preliminary conclusions hence are: a potentially useful, playable alternative, which has a character I prefer in front of the c5-variation. 

Another game: let me know if you have improvments to suggest.

1... Nfd7 2. Qd2 c5 3. dxc5 Nxc5 4. Bh6 Nc6 5.O-O-O a6 somehow I feel that black must treaten something through a pawn storm;this seems to be a useful move in many possible lines. An alternative: 5... f5. One of the best moves, according to Rybka 2.2. 6. Bxg7 Kxg7 7. h4 h5 {Fritz 11 & Rybka agree: the attack most be delayed with h5.})

6. Bxg7 Kxg7 7.g4 what is the idea behind this move? To stop h7-h5? So the machine delayes the h4-advance a bit. 7 ...Rb8 can this really be wrong? I don't see the point in e.g., touching the pawns in front of my king (often a bad idea in these kind of positions). Perhaps Be6?  

8. Kb1 a useful move, but really necessary here? Slightly odd. (8. h4 do I then need h7-h6 after all?) 8... Be6 (8... Qa5 is possible)

9. b3 again, an odd move. (9. h4) 9... Qa5 10. Qb2 the machines (F11 & R2.2) like blacks position better, wheareas I thought black was slightly worse. I stopped playing here, looking for improvments earlier.

Possible is 10...f6. Maybe black is OK here, after all. Look at whites underdeveloped pieces. Black has some play with the pawns on the queenside. h2-h4 will come, possibly dangerous, if black plays too solidly.

come on, guys .. No analysis, improvments, suggestions?