A Tour of the Sicilian - Choosing a Variation as Black

Starting with the Sicilian can seem intimidating. There's a general feeling that you must know tons of theory to play the sicilian well. This is a misconception... The sicilian is a very rich opening, and it can be played in many different ways. Black also chooses the mainline variation he gets to play, and black has 13 mainline sicilians he can choose from. Some of these require more theory, others are very rare and will get you out of theory quickly. Some are positional, others are tactical; some are passive, others are more aggressive. So it's really up to black how much theory he'll have to learn, what sort of game he's going to get, etc.. 

Furthermore, the same generally holds true against the anti-sicilians - black usually has 3-5 viable options to choose from in every line, leading to different types of games... some of these lines are well known, some lesser known, some very rare. There's a positional aspect to the sicilian as well. Many patterns and ideas recur throughout the lines. And so often if black finds himself not knowing how to play an uncommon position black is able to respond well on principle alone. 

This post is going to examine the 13 viable mainline sicilians black can choose to play against the open sicilian. I will detail what I think the different characteristics of these lines are, what their tradeoffs are, and point out things I think a club player may want to consider when choosing a sicilian as black. I've also tried to list these in rough order of my preference as a club player. Obviously this is a light analysis and if you want something more in depth there are videos available on all these lines. 

Before we talk about the specific sicilians, a bit should be said about the 2nd move options. The common ones are 2... d6, 2... e6, and 2... Nc6.

Of these 3, 2... d6 clearly deals with the anti-sicilians the best. The delayed alapin is terrible against it, black can just play Nf6 and pressure e5, which can't be pushed due to d6. If black proceeds accurately white will be unable to push d4 in one move, and the game has already equalized. Likewise the Moscow is not nearly as testing as the Rossolimo, and Bd7 is a relatively compact response to it. So there is some argument for playing a Najdorf, Classical, or Dragon on account of 2... d6, especially at club level where how you deal with anti-sicilians is even more critical. 

The only real issue with 2... Nc6 is it must face a Rossolimo. This is probably the most objectively testing anti-sicilian white can play. For example, it's objectively better for white than the Sveshnikov. It also leads to a very grindy and unpleasant game. The worst part about it though is white gest to choose to enter this, and will have the opportunity frequently. Thus black has ceded his biggest advantage in the sicilian, namely his ability to choose the major deviation. 

Finally, against 2... e6 the delayed alapin can be quite annoying for black to deal with. His best option is probably to enter an advanced or exchange french. It's not that this position is objectively difficult for black, it's just that black wanted to play a sicilian and now he finds himself in a french, it's not the kind of game he wanted. Furthermore, the Szen variation (5. Ndb5) is also quite sharp if played well. 5. g3, 5. b4 (the wing attack) and 5. b3 (the Westerenin) are also serious lines. While all these can be deal with well, there are just many of them and they can be grindy / still remain good for white. On the upside, the move 2... e6 is probably the best response to the Closed Sicilian. 

1) Taimanov/Kan Hybrid (Nc6 > a6): 

People often debate whether this is a Kan or a Taimanov. I've settled the debate for them, we're calling it a hybrid. Note that you can also reach this from the old sicilian (2... Nc6).

This line can be played in a variety of ways. In some sense you could think of it as the Najdorf of French sicilians. The move a6 is a valuable defensive move in these french sicilians, and prevents many Ndb5 shenanigans which occur in the main lines of the four knights or bastrikov. Instead, black chooses to play in a reactive positional way, and white makes the first move. 

The way I prefer to play this line is for its transpositions. You can followup a6 with Qc7 in many lines, and transpose right back into the lines of the Bastrikov, bypassing the sharper queen harassing / main line defenses white may have had prepared. In general, once you safely reach a stable positional game in the Taimanov you're doing well. 
But instead of Qc7 you can also play Nf6 and get a pseudo four knights position. This can be a nice response to 6. Be3, avoiding whites prepared Bastrikov response (which you'd walk into if you played Qc7 immediately).

The transpositions from this a6 Taimanov line back into the Bastrikov often have some of the best winrates for black I've seen in the sicilian.  

This is also a very under-explored sicilian. The bastrikov has been favored for a long time, probably due to it scoring about 2 centipawns better according to engines. So, rather than playing it for transpositions, you can also play this line for its unique deviations. Often the lines are often positional, though occasionally you will run into unique attacks or sacrifices the engine likes which you won't find played often in the database. But positional lines are the common case, and here black has a comfortable time since his opponent is usually out of book and he can just play Qc7 / Nf6, Bb4 or Bc5, often push d5 or fiancetto b5 > Bb7, and the position plays very naturally. 

White's theoretical response to the Taimanov/Kan hybrid is to immediately take via Nxc6. Here black still has a solid pawn structure in the center, and pushes d5. If white follows up with the main line and plays it sharply, though, blacks king can get displaced. However the position is still objectively fine, and the theory here is not well known or even fully developed. There are opportunities for reaching legitimate and novel positions as early as move 10. Considering this is the main theoretical line of a sicilian I find that pretty surprising. In these lines often both queens will be active and it ends up being quite a sharp and chaotic game. The eval is on par with something like the classical richter-rauzer. But compared with something like the rauzer or Yugoslav I think it's better. For one, if I'm going to deeply study a theoretical line I much prefer it be one the opponent will probably not know, especially in an online setting. I find there are formulaic responses to the four knights and bastrikov which people do know, but that isn't really the case here. These lines get chaotic and are underexplored. So I'm a big fan of this line. 

Since you can mostly play it based on positional principles it's also quite good from a practical standpoint. The only negative thing I'll say about it is, due to the move a6 being played early on, it can't punish white for going wrong as hard as something like the four knights or classical can. But if it's a positional opening you want I think this is the way to go. It's objectively good, and I think the less formulaic nature will make it better against good players... hence I put it at rank 1. 

2) Accelerated Dragon:

More legit than the hyperaccelerated dragon, it's a completely viable alternative to the dragon. Infact it's better objectively. Most importantly it allows you to avoid the yugoslav attack against the dragon. You can push d5 in one move instead to counter it, and this tempo makes a big difference with the Yugoslav.

One biggest downside is you must face the rossolimo to reach the accelerated dragon (unless you play the hyperaccelerated, but that comes with its own issues). The Rossolimo is probably the most legit anti-sicilian, and it's a position that white chooses to enter / will get often, so you can expect white to be well prepared here. White can also respond to the accelerated dragon with a maroczy bind formation, which requires a deep understanding and good prep to deal with properly. Still, neither of these lines are nearly as testing as the Yugoslav. 

One advantage of all these dragon setups is if white doesn't play the most testing lines against them, but just plays a more casual sideline, black just winds up with all active pieces, a ready-made attack on the queenside, and a superior pawn structure long term. So if you have the determination to buckle down and deeply memorize the testing lines you will reap benefits against everything else. I like this at below-master level because generally you can expect players to not know the testing lines very well. 

Generally the dragons reward good early game prep, especially deep prep in specific lines. Often the theoretical lines get sharp and somewhat tactical. If you're a player with a strong early game, who isn't afraid of preparing some sharp lines, this dragon can be very rewarding. 

3) Four Knights - 

This is a very underestimated line. It is a very attacking sicilian. you bring both knights out quickly and you also open up the kings bishop... so you can often push d5 early with alot of initiative after pinning whites knight on c3, and the game equalizes or almost equalizes. You can do this against most moves white makes - there aren't many very good lines against this sicilian, really just 2 - 6. Ndb5 and 6. Nxc6. At below masters level, where players don't know these lines, you will often get an ideal almost equal position very early on, which makes this sicilian very appealing in online and club settings. The positions are also very natural and easy to play, which makes it a good starting place for beginners.

Against Ndb5 you can be forced to either go into a Sveshnikov or to play an objectively suboptimal line such as the Cobra variation (a rare variation that's still scoring very well in practice). On the other hand, the four knights can also be a way of reaching the sveshnikov while avoiding the rossolimo / bypassing the anti-sveshnikov.

With 6. Nxc6 you get a position that's actually pretty dynamic / messy but in a good way. Whites king even gets displaced in the main line, but white has a strong counterattack. Objectively it's on par with the french defense or Rossolimo, about +0.23. But there are still some unexplored deviations in this line which the new engines have unearthed. Compared with the Rossolimo I'd prefer this line, at the very least black will have a greater familiarity advantage here. Thus the Four Knights is considered a perfectly viable way of reaching a Sveshnikov (though you will get it infrequently in a below-masters setting). 

In general this is a great sicilian online and for beginners due to how attacking it is and how often white goes wrong / just hands black an almost equal position. Putting white's theory to the test early is always a good thing at this level. 

4) Classical 

A somewhat uncommon sicilian these days. It's objectively sound and generally underestimated. If white doesn't play the main line rauzer you often just get an ideal setup... your setup is very aggressive due to the knight on c6, and you haven't played any passive moves. It's similar to the four knights in this respect, though not quite as aggressive since you haven't opened up a bishop yet. The move 2... d6 makes this a little more positional and theoretical than a french sicilian such as the taimanov or four knights. You don't usually push d5 early. Generally this is a sicilian where you can leverage theoretical knowledge and punish your opponents mistakes when he goes wrong, due to the more active setup compared with something like the Najdorf or Dragon. For example, the english attack setup does not work against the classical due to the immediate Ng4 followed by Qb6, combined with Nc6 this immediately threatens Nd4 and checkmate if white moves the knight. 

The only testing line against the classical is the richter-rauzer, where white will usualy castle queenside and launch his pawns on the kingside. It's not an objectively bad line for black (in contrast with the yugoslav), but it's generally agreed it's hard to play. So again, as with the dragons / four knights, this is the kind of sicilian where you can study 1 line very deeply and all the other sidelines are quite easy to deal with. This is very beneficial at club level where the player has limited time to study theory, and it makes it very different than something like the Najdorf which must face ~5 or so serious lines. 

In the richter-rauzer lines (Bg5), even though your pawn structure may be compromised or you may be down a pawn (depending on the line you choose), you still have significant counterplay and the position is still objectively fine for black / very unbalanced. There's alot of complexity here, so it rewards careful study and practice. 

A big advantage of the classical is you can reach it via 2... d6, and there is really no challenging 3rd move anti-sicilian you must face. The Moscow is not nearly as challenging as the Rossolimo. 

Overall this sicilian will reward players who like to study theory / play positionally while also using it to punish their opponents. This can also be a good way of getting a theoretical sicilian without playing something that's super-common like the Najdorf.

5) Taimanov bastrikov -

if you want to throw off the opponent and get out of theory this is a very good way to do it. Not only is it an offbeat sicilian but it also brings the queen out, and the variations tend to increase in size complexity as a result. You'll get people out of whatever algorithm they have relatively easily here.

These lines generally play out in a few ways. There are early Ndb5 moves where blacks queen is harassed, and black has to drop back the queen / repel the attack precisely. These lines generally are not very threatening if black does respond precisely. 

If white doesn't play precisely black will often just get a very natural Taimanov setup - a6 / Nf6 / Bc4 or Bb5, and either push d5 early, or play b5 > Bb7 first and build up to d5. When black gets this setup the game has gone into easy mode. 

There is one line where white can trade 3 pieces for blacks queen, leading to quite a chaotic position, and although black has advantage he must play well to realize that.

There are some other testing lines, such as 6. Be3 a6 7. Qf3, where black is still fine, he just has to play precisely... but it's rare to actually encounter these. 

In general I'd say these lines can be either sharp when blacks queen gets harassed, but usually they wind up as easy-mode positional lines where white doesn't know exactly how to proceed. The lines get a little sharper than the Taimanov w/ a 5... a6, but black has a little more initiative in certain lines due to delaying a6. 

Generally I'd recommend this to players who want to avoid the opponents algorithms, but who also don't mind playing sharp occasionally.

6) Lowenthall -

another very attacking sicilian, but this one is very much a tactical line. The lines are also pretty forcing. Good for people who like to memorize sharp and deep lines that don't branch much, and play the sicilian more like an e4/e5 line. It's not as objectively sound as the sveshnikov but it's a little more offbeat. But it can also transpose into the sveshnikov with a little extra effort. But the lowenthall itself is a good way of minimizing the theory you need (since it's very forcing) while still getting a sharp, attacking tactical sicilian.

7) Nimzowitsch -

I generally think of this as an improved alekhines, the objective eval is about the same and many of the positions have a hypermodern tone to them. However, it tends to lead to advantages for black more reliably than the alekhines. Since you've played c5 you can retreat the knight to c7 instead of b6 if it gets tempod, this is an improvement over the Alekhines. People also are far less prepared for this line than they are Alekhines.

People act as if there's a hard counter to this via main line, this actually isn't true... the main line is dynamic and sharp. It's very playable for black, black even has a positive winrate in the line. It's also objectively better than the yugoslav, and yet the nimzowitsch is also only played in 1% of games. You could say this unfair reputation has increased its potency. It is one of my favorite weapons. I played this line for a while - at 2000 elo almost no one knew the mainline, I got it maybe 3 games out of 100. I think I'd be a little hesitant to play this if my opponents were expecting it, like in a club or tournament setting, but in online play or as an off-weapon it's very good and underestimated. The opponent often has no idea how to handle these strange knight maneuvers, and you often just get an edge in the opening as a result.

One of the biggest selling points of this sicilian is it avoids most anti-sicilians, since it begins on move 2. Furthermore, the patterns in the nimzowitsch are very similar to the Nf6 lines in the alapin, and can even transpose with it in some places. Combined with transposing the Smith Morra into the Alapin you've dealt with the 2 most common 2nd move anti-sicilians as well. So this is really the ultimate anti-anti-sicilian. As an off weapon this makes it particularly valuable against players you know intend to play anti-sicilians. 

If black wants to play the sicilian and doesn't want to deal with the anti-sicilians / wants to get his opponent out of book quickly this is a good line to play. 

8) Najdorf -

It's the most commonly played sicilian. It's also a more defensive and positional sicilian, a6 is a defensive waiting move. It keeps the balance and refrains from committing to a setup. As a result of not controlling the center white has 10 or so serious responses he can choose from. So the amount of theory needed to play this line is vast. And while the Najdorf is probably the most equal and objectively sound sicilian... to retain that objective edge there's very little room for deviation. The moment you deviate you're just playing a common, predictable, and relatively passive position which is objectively no better than other lines. So in an online or below-masters setting I'm not a fan of this line. It just makes your play extremely predictable and requires you to grind enormous theory which your opponent will probably know anyway. It also allows white to choose the major deviation, ceding probably blacks biggest advantage in the sicilian - his ability to choose the main deviation. 

9) Scheveningen -

This is a somewhat underplayed line. People assume the g5 pawn advance can't be dealt with but if you study it you'll be able to deal with it. I've found almost no one actually knows the g5 theory here. That said, it's not only the Keres that gives black trouble but also the English attack setup, so you have to study that deeply too. However if you know the theory you will hang on, and you will get some attacking opportunities since this is a sicilian that focuses on development - e6 (in contrast with a6 or g6) develops the bishop and castles relatively quickly, it also prepares d5 more quickly.

Note that there are actually two ways of reaching the Schveneningen - via the french sicilian or via the modern (d6). The french sicilian is more played, here you can delay Nf6 a bit and play Nc6 first, this deals a little better with the Keres. But it's not quite as good against the English attack setup.

I wouldn't write the Scheveningen off but I probably wouldn't use it in a tournament setting as a main weapon, however in online play where your opponent just gets one game against you it's very viable.

Playing it also improves black as a sicilian player - these e6/d6 structures occur frequently throughout the sicilian and knowing them will improve blacks play. 

10) Dragon -

in this line you'll often find the kingside comes under attack. While the yugoslav doesn't appear to be objectively very good for black, the dragon is not refuted by any means, 
and infact black is scoring well in the yugoslav at club level. Probably because the yugoslav is the only really challenging line for the dragon, so black can just focus very heavily on learning the theory and out theorize the opponent. It's similar to the classical in this respect. And this makes it very practical for the club player. It's also used as a winning weapon in tournaments since black has a significant counterattack in the yugoslav. Blacks pawn structure is also good and trades / the endgame favor him long term. For white to maintain the engine edge in the yugoslav it's a very long sequence of moves, black will want to play this precisely until the opponents knowledge of theory is depleted. However, the really challenging part of the Yugoslav isn't necessarily the main engine line, it's all the options white has, and many of these are often very sharp and punishing when black goes wrong. Nonetheless, black is still scoring relatively well in this line at club level, probably just due to superior prep. Fortunately once black knows how to deal with the yugoslav attack everything else is quite easy to deal with for the dragon, dragons always deal well with non-testing lines since all blacks pieces are just active and the pawn structure is very solid.

This variation of the dragon doesn't have to face the Rossolimo (unlike the accelerated dragon), and in general d6 deals very well with sidelines. It also doesn't have to deal with the maroczy bind setup that the accelerated dragons have to deal with. 

As with all the dragons, generally players who are good at opening prep will do well with them, prep is what's essential for playing the yugoslav attack. 

11) Sveshnikov -

a very tactical, attacking sicilian. It's a completely crazy attack, but it's also very theoretical. Here the effort is really to try to out-theorize the opponent in these deep, sharp tactical lines. Allows black to play a very attacking style of chess. However you have to tolerate some very dubious positions especially with your king exposed. Overall it's alot of theory and it puts alot of burden on the opponent to know all that theory. A good option for players who like this kind of sharp tactical but also highly theoretical style of play. 

12) Hyperaccelerated Dragon -

Like the accelerated dragon your goal is generally to transpose into the dragon through another move order while avoiding the yugoslav. Infact, the Accelerated Dragon can be reached from the Hyperaccelerated. 

While this line is entered on move-2, and you might think that's a huge advantage as it bypasses so many anti-sicilians... actually, one of whites most common responses here is c3, and this is essentially an alapin on steroids. There's also a line where white can bring the queen out early and trade down quickly, reaching a very early endgame where he has advantage. This gives white an easy, prepackaged response to the hyperaccelerated. These lines in general require very sharp play from black as well. 

It's not that this is a bad line, it's just that I would rather play a Rossolimo than some of these easy endgames or alapins on steroids, I don't really feel black gains alot by playing the fiancetto early. What he would gain, I suppose, is a factor of surprise - white isn't going to see the hyperaccelerated as often as he sees his Rossolimo. Furthermore, even in the early endgame line there are a couple interesting moves the newer engines have found which may allow black to mix things up. If you like precise play and want to throw the opponent off, especially in an online or below-masters setting, give it a try. 

13) Kan -

some of the lines against the kan are quite sharp on black, and you have to know the theory well. but if you know the theory well, with the right maneuvers, you can fight for equality. What you'll get out of it is a very theoretical sicilian that's not very common, so it puts pressure on white to know his theory. The maroczy bind is the line that gives it the most trouble. Its best line is probably the wing attack. It's worth noting you can also enter just the wing attack from the closed sicilian after 2... e6 if white tries to transpose back into an open line. If I was going to play the Kan that's probably how I'd play it. Unless I just wanted to learn how to deal with maroczy-bind setups better. 

So that's all the variations I think are serious for black. 

As you can see there's a sicilian that suites pretty much any need - if you want to get the opponent out of theory with a focus on online / club-level play there's the french sicilians... if you want to avoid the anti-sicilians there's the nimzowitch... if you want a tactical sicilian there's the lowenthal or sveshnikov... if you want a theoretical line that's capable of punishing mistakes there's the classical... if you want an objectively ideal line and are willing to learn alot of theory there's the Najdorf... if you want a positional line that requires / rewards more focused prep there are the dragons.... etc., etc.. Unlike in most other defenses to 1. e4 it is black who chooses the primary variation, and there are 13 options. It is my opinion that the sicilian is the best defense in chess, and the only response to 1. e4 with no downside for black.