Sicilian Variations - What To Study?

I'm curious about expanding my Sicilian Defence knowledge when playing with the black pieces, as it's currently my best scoring opening - but I certainly have major room for improvement.

To anyone who is well experienced in the Sicilian, what variations would you consider putting time & effort into studying?

I'm conscious that some variations are pointless for me personally, as its my own choice weather to play them or not, but some are forced by my opponents... So I'm keen to avoid studying the variations that I wouldn't openly choose to play, and instead focus on what i will encounter.

An example of this in other openings; I always play Nf6 against 1.d4, so to study The Queens Gambit for instance is completely pointless for me... So time may be better spent on the Grunfeld and many others.

I assume the same concept can be applied to the Sicilian.... For instance, if the opening follows the Classical line, I would chose to play the Najdorf over the Dragon, so to study the later would be futile.

From a quick glance, I'm thinking the variations of choice that my opponent could force upon me would be:

Alapin
Smith Mora Gambit
Closed Sicilian

So along with the Najdorf, this would cover the majority of what I'm likely to encounter when playing with the black pieces against 1.e4?

As always, I appreciate everyone's advice and input!

Personally, I think old Sicilian is my answer. Because it's easy

Good question. This is something I've been studying lately. I can share some of my notes with you... since you're also playing the d6 sicilian (Najdorf). And 2... d6 has the easiest time against the anti-sicilians so it's a good choice. No rossolimo, no improved alapin. So here's a non-exhaustive list of the common anti-sicilians you're likely to face -

- alapin - okay
- closed sicilian - good
- smith morra - crap 
- delayed alapin - crap but requires some knowledge to render crap 
- delayed bowdler attack - crap 
- Moscow variation - good
- chekhover variation - borderline crap
- bowdler attack - crap
- tartakower variation - crap 
- mcdonnell attack - kind of crap

Though many of these lines are crap of course you can still lose to them if you're not prepared.

Closed Sicilian - Since white can still transpose into a mainline sicilian your 2nd move response depends alot on what sicilian(s) you play. The move 2... e6 will always be king as far as dealing with Closed Sicilian lines themselves... 2... Nc6 is also quite good. But unless you're willing to learn either the scheveningen (e6) or the classical (Nc6) to allow you to play the favorable closed lines while still transposing... i.e. if you just want to stick with the Najdorf - you should probably play 2... a6. It can be worth it to learn an off-weapon in my opinion since usually Closed players don't transpose, but I digress. After 2... a6 the positions arising are still fine, kind of boring fiancetto positions basically. But I don't recommend 2... d6, this is a wasted move against the Grand Prix... the Grand Prix really is shredding 2... d6 if you look at the results, and many players go for it.

McDonnell attack - this is just the Grand Prix but it's been played too early. Which allows you to always respond e6>d5, the ideal antidote to the Grand Prix. There are a variety of positions that can follow but they go along the same lines as the Grand Prix and if you understand those positions you'll be able to play this fine, generally speaking. I wouldn't focus too hard on these positions since it's not that common people will play this and you already basically cover these ideas via the Grand Prix, for the most part.

Moscow variation - this is the most serious anti-sicilian the d6 player faces. But it's still fine, it's nothing like the Rossolimo. The two main responses are 3... Bd7 and 3... Nd7. In general, 3... Nd7 is alot larger and requires much more theory to get right. And white chooses the line you go into, so it's a bigger burden on you. 3... Bd7 will lead to a trade and much less theory, that's why I prefer it. I like Nxd7 because it does a little better against the maroczy bind setups that can arise after. Also... there's this cheesy attack you can play against the main Re1 line. But overall this is the most testing anti-sicilian and there are many variations you'll have to study -

Smith Morra - just take the pawn and learn the first 8 moves or so, it's a crap opening, pretty much every Smith Morra player plays the same combination of moves for the first 8 moves... just learn how to respond. I like the e6/a6 setup, it shuts down whites bishop and develops the queenside first, doesn't even worry about pushing d5... there's very little white can do, he has no real attack and no real compensation -

Tartakower / Chekhover variations - neither of these are particularly good, but I actually play the anti-Qxd4 move order which avoids both these lines. The reason I play this is if white tries to do anything other than transpose into the open sicilian the game is either equal or black has an advantage. And many players, having never seen this move order, feel the need to mix it up... in like 25% of my games I'd say I just end up with immediate equality due to them trying this. Not only equality, but lines I know which they rarely do. This does require studying to learn these anti-Qxd4 lines... but then again you can circumvent studying for the Tartakower / Chekover, and it's easy equality so why not -

Bowdler attack - it's a dumb line honestly. It's really just not good. You can handle it in alot of ways but I like an early a6, usually white will respond with a4 and you immediately have an advantage, there's a hole for your knight now on b4. You can followup with e6 if you like, I like to followup with Nf6, it leads to this improved french where your bishop is outside the pawn chain. But this is just one way to play it, there are others -

Delayed Bowdler - slightly better than the regular Bowdler but still not great, almost equal... it's really just a bad version of many open sicilian lines where white is castling kingside and his bishop will end up biting granite. Here again I like the the a6/e6 setup... a6 will often bait a4 which leaves a hole on b4 for the knight... they often take that bait. But I wouldn't put too much thought into this line, there's nothing too threatening or unique here, if you have an intuition for playing your open lines you will play these lines fine -

Delayed Alapin - it's completely equal, but requires a little more knowledge to deal with. But definitely play the Nf6 line, it's the only one that gives white trouble. After Nf6 White has a big problem defending his e pawn without making a concession -

For Be2 - make sure not to take the E pawn with your knight due to Qa5+ picking up the knight. Just have faith and fiancetto / castle, white is on a timer - by the time you castle he's gotta find some way to defend the pawn... Qc2 doesn't work since you'd respond with Qc7 and now he can't push d4 since his queen will be left hanging, for example. In the mainline he plays Re1 > Bf1... with that wasted time and passivity you get an easy fiancetto and it's a completely equal position. Generally in these fiancetto positions unless white can attack you early (i.e. yugoslav style) or blockade the center with pawns early enough you're going to have a great game, white can do neither in this case -

if Bd3 his bishop isn't preventing the Bg5 pin, so we play Bg5 and this can prevent him from pushing d5 or lead to some tactics in other lines -

And if h3 white has wasted too much time, here black can actually play e5, sometimes even followed by d5 and black gets this reversed maroczy bind situation. Overall there's alot of theory here but the lines are good you just have to study them -

Alapin - a boring grindy line, usually white doesn't prepare very much but just wants to have a game of chess where he can avoid your prep... so I like to mix it up and make him play sharp. There are actually a number of strong and rare sidelines for this purpose -

To play the Najdorf you need to cover :
1) 2. d4 and 2. f4
2) decide what to do after 2. Nc3 : after 2... e6 and 2... Nc6 there is no Najdorf transposition possible after 3. Nf3 or 3. Nge2, and if 2... d6 then again 3. d4 and 3. f4 are to be considered.

3) pick your Closed Sicilian line

3) 2. c3 and 2. Nf3 d6 3. c3 (they're not the same)

4) 2. Nf3 d6 3. Bb5+

5) 2. Nf3 d6 3. Bc4

6) 2. Nf3 d6 3. d4 cxd4 4. Qxd4

7) 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. f3

In response to the OP, some of the items you say are completely false. For example, the blanket statement that the Queen's Gambit is useless because you play 1...Nf6 is false. Now if you said that you answer 2.c4 or 2.Nf3 with 2...g6, only now would that be a true statement. But take a Nimzo-Indian player. 1.d4 Nf6 2.c4 e6. The idea is to answer 3.Nc3 with 3...Bb4, preventing 4.e4.

However, what about 3.Nf3? Black actually has 4 legitimate responses:

3...b6 - The Queen's Indian

3...Bb4+ - The Bogo-Indian

3...c5 - The Modern Benoni, having avoided the Flick Knife Attack, many Benoni players will only play it this way and play a Nimzo against 3.Nc3 because of that exact variation.

3...d5 - The Queen's Gambit Declined, many of which will only play it under these circumstances, having avoided the exchange lines with Nge2.

As far as the Sicilian you need a line against the Open Sicilian. Then, based on which line you choose, you need a line against each of the following, the last two of which what you play in the open Sicilian matters:

1) The Alapin (2.c3)

2) If you play 1.e4 c5 2.Nf3 d6, then 3.c3 DOES NOT transpose to 2.c3. If you play 2...Nc6 or 2...e6, then it is a transposition

3) Morra Gambit - You can decline the second pawn, if you wish, which would lead to an alapin.

4) 2.a3

5) 2.b3

6) 2.b4

7) The Kopec System (1.e4 c5 2.Nf3 any move 3.Bd3 followed by 4.c3 and 5.Bc2)

8) If 2...Nc6 or 2...d6 in the Open Sicilian, then the Rossolimo or Moscow. If 2...e6, the Kings Indian Attack (3.d3)

9) If 2.Nc3, you need to look at what you play against the Open Sicilian because you can get tricked into it. If you play the Najdorf or Kan, where the Knight goes to d7, then 2...a6, NOT 2...Nc6 because of 3.Nf3 and 4.d4. If you play any other Open Sicilian, then 2...Nc6 and you have the Closed Sicilian (3.g3), the Grand Prix Attack (3.f4), Carlsen's line (3.d4), the Tiviakov (3.Bb5), the Chameleon (3.Nge2) and 3.Nf3 typically transposing to the Open Sicilian.

10) Last but not least, if you play a 2...d6 Sicilian, then you need a line for 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Qxd4 and also 1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.f3 (The Prins)

That will take you a life time to master. That is why I play the French. Aside from a few oddball sidelines, it is really just 5 lines you need to know: 3.Nc3, 3.Nd2, 3.e5, 3.exd5, and the Kings Indian Attack (2.d3).

Yeah it's definitely good to study the dragon as a sicilian player since fiancetto is a good way of dealing with many sidelines, including lines in the Moscow where you actually end up in some fiancetto marcozy bind situations. For maximum flexibility... it's generally good to know just the basic ideas in a number of diffent sicilians.

Sorry about that, you're completely correct, and I did think about this as I wrote it.

I was just trying to make a (somewhat awful) point that I want to focus my time looking at variations that are more forceable by my opponent, rather than studying multiple lines that I can chose from. (just because it's less to learn in the short run!)

Having said that, the information you provide is awesome. So thank you for taking the time!

What is the "old Sicilian"?

That isn't even anything. White still has a wide array of responses other than the Open Sicilian's 3. d4: d3/Nbd2/g3 ideas, 3. Nc3, 3. Bb5, 3. c3.

for white, study the alapin, grand prix, dragon sicilian, and open and closed positions.

as black, study how to counter all of those, the smith morra gambit, and some other advanced theory

just study everything to be real

Even if you memorized all of ECO, you still couldn't predict the rise of h3 in the Najdorf.

If you are committed to Najdorf, I would study in these order:

old mainline (bg5)

rest (non-bg5, non-poisoned pawn. your discretion as to which lines are important as I do not know Najdorf lines.)

alapin

Grand prix

smith-morra

closed sicilian (emphasis on f4 lines)

Bb5 lines (optional)

wing gambits (optional)

would give you a good enough coverage of the opening.

This is a mistake alot of prospective sicilian and anti-sicilian players make, they assume every move white plays in the sicilian leads to a completely new type of position and this is going to just confuse the heck out of black. No, not really... there's alot of overlap in the ideas of different sicilian move orders and positions. If white doesn't apply pressure you can often just play typical sicilian moves and you'll get a good game. Even deep into the midgame the positions will often play very similar to one another, just less pressure on black basically.

I would certainly study this palette of Sicilians: Open: Franco-sicilian, Nimzo-Rollisimo attack, Chekover variation. These are offbeat and may provide a good position if played correctly even though they are dubious.