The French has a reputation for solidity and resilience, though it can result in a somewhat cramped game for Black in the early stages. Black often gains counterattacking possibilities on the queenside while White tends to concentrate on the kingside.
The defence is named after a match played by correspondence between the cities of London and Paris in 1834 (although earlier examples of games with the opening do exist). As a reply to 1.e4, the French received relatively little attention in the nineteenth century compared to 1...e5. The first world chess champion Wilhelm Steinitz said "I have never in my life played the French Defense, which is the dullest of all openings". In the early 20th century, Géza Maróczy was perhaps the first world-class player to make it his primary weapon against 1.e4. For a long time, it was the third most popular reply to 1.e4, behind only 1...c5 and 1...e5. However, according to the Mega Database 2007, in 2006, 1...e6 was second only to the Sicilian in popularity.
Historically important contributors to the theory of the defence include Mikhail Botvinnik, Viktor Korchnoi, Aron Nimzowitsch, Tigran Petrosian, Lev Psakhis, Wolfgang Uhlmann and Rafael Vaganian. More recently, its leading practitioners include Evgeny Bareev, Alexey Dreev, Mikhail Gurevich, Alexander Khalifman, Smbat Lputian, Alexander Morozevich, Teimour Radjabov and Nigel Short.
Alternatively or simultaneously, Black may try to break down White's centre, which is cramping his position. Usually playing ...c7-c5 is not enough to achieve this, so Black will often play ...f7-f6. If White supports the pawn on e5 by playing f2-f4, then Black has two common ideas. Black may strike directly at the f-pawn by playing ...g7-g5. The pawn on g5 may also threaten to advance to g4 to drive away a white knight on f3. Or, he may play ...fxe5, and if White recaptures with the f-pawn then Black gains an open f-file for his rook. Then, as White usually has a knight on f3 guarding his pawns on d4 and e5, Black may sacrifice the exchange with ...Rxf3 to further undermine the white centre. On the other hand, if White plays dxe5 then the a7-g1 diagonal is opened, making it less desirable for White to castle kingside. Sometimes, if White is underdeveloped and his king is still in the centre, Black may sacrifice a piece on e5 to destroy White's centre totally and begin an attack. White usually tries to exploit his extra space on the kingside where he can sometimes create a mating attack. White tries to do this in the Alekhine-Chatard attack, for example. Another example is the following line of the Classical French: 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.Bxe7 Qxe7 7.f4 0-0 8.Nf3 c5 9.Bd3 (see right)
White's light-square bishop eyes the weak h7-pawn, which is usually defended by a knight on f6 but here it has been pushed away by e5. A typical way for White to continue his attack is 9...cxd4 10.Bxh7+ Kxh7 11.Ng5+ when Black must give up his queen to avoid being mated, continuing with 11...Qxg5 12.fxg5 dxc3. Black has three minor pieces for the queen, which in theory is a straight swap, but his king is vulnerable and White has good attacking chances.
Apart from a piece attack, White may play for the advance of his kingside pawns (an especially common idea in the endgame), which usually involves f2-f4, g2-g4 and then f4-f5. A white pawn on f5 can be very strong as it may threaten to capture on e6 or advance to f6. Sometimes pushing the h-pawn to h5 or h6 may also be effective. A modern idea is for White to gain space on the queenside by playing a2-a3 and b2-b4. If implemented successfully, this will restrict Black's pieces even more. One of Black's main problems in the French Defence is his queen's bishop, which is blocked in by his own pawn on e6. The bishop can be almost useless for the early part of the game, and unless Black makes some effort to free it (usually with the pawn breaks ...c5 and ...f6), it can remain that way for the whole game. An often cited example of the potential weakness of this bishop is Tarrasch-Teichmann, San Sebastián 1912, in which the position on the left was reached after 15 moves of a Classical French.
Here Black is reduced to complete passivity. The bishop on c8 is a pathetic sight, hemmed in by Black's own pawns on a6, b5, d5, e6 and f7. White will probably try to trade off Black's knight, which is the only one of his pieces that has any scope. Although it might be possible for Black to defend this position and hold on for a draw, it is not easy and, barring any mistakes by White, Black will have no chances to create counterplay. In Tarrasch - Teichmann, White won after 41 moves. In order to avoid this sort of fiasco, Black usually makes it a priority early in the game to find a useful post for the bishop. Black can play ...Bd7-a4 to attack a pawn on c2, which occurs in many lines of the Winawer Variation. If Black's f-pawn has moved to f6, then Black may also consider bringing the bishop to g6 or h5 via d7 and e8. If White's light-square bishop is on the f1-a6 diagonal, Black can try to trade it off by playing ...b6 and ...Ba6, or ...Qb6 followed by ...Bd7-Bb5.
However, despite the symmetrical pawn structure, White cannot force a draw. An obsession with obtaining one sometimes results in embarrassment for White, as in Tatai-Korchnoi, Beer Sheva 1978, which continued 4.Bd3 c5!? 5.Nf3 Nc6 6.Qe2+ Be7 7.dxc5 Nf6 8.h3 O-O 9.O-O Bxc5 10.c3 Re8 11.Qc2 Qd6 12.Nbd2 Qg3 13.Bf5 Re2 14.Nd4 Nxd4 0-1. A less extreme example was Mikhail Gurevich-Short, Manila 1990 where White, a very strong Russian grandmaster, played openly for the draw but was ground down by Short in 42 moves.
To create genuine winning chances, White will often play c2-c4 at some stage to put pressure on Black's d5-pawn. Black can give White an isolated queen's pawn by capturing on c4, but this gives White's pieces greater freedom, which may lead to attacking chances. This occurs in lines such as 3.exd5 exd5 4.c4 (played by GMs Normunds Miezis and Maurice Ashley) and 4.Nf3 Bd6 5.c4. Conversely, if White declines to do this, Black may play ...c7-c5 himself, e.g. 4.Bd3 c5. This idea was employed successfully by Korchnoi, but it is probably best to reserve this risky strategy for must-win situations.
If c2-c4 is not played, White and Black have two main piece setups. White may put his pieces on Nf3, Bd3, Bg5 (pinning the black knight), Nc3, Qd2 or the Queen's knight can go to d2 instead and White can support the center with c3 and perhaps play Qb3. Conversely, when the Queen's knight is on c3 the King's knight may go to e2 when the and the enemy Bishop and Knight can be kept out of the key squares e4 and g4 by f3. When the Knight is on c3 in the first and last of the above strategies, White may choose to castle on the Kingside or the Queenside. Obviously, opposite side castling leads to dynamic checkmate struggles and when the Kings castle on the same side the players usually putz around and don't do much until a complex endgame is obtained. Black may do the exact mirror image of all of the choices above. The positions are so symmetrical that the options and strategies are the same for both sides.
Another way to imbalance the game is for White or Black to castle on opposite sides of the board. An example of this is the line 4.Bd3 Nc6 5.c3 Bd6 6.Nf3 Bg4 7.0-0 Nge7 8.Re1 Qd7 9.Nbd2 0-0-0.
The main line of the Advance Variation continues 3...c5 4.c3 Nc6 5.Nf3. There are alternative strategies that were tried in the early 20th century such as 3... b6 intending to fianciatto the bad bishop or 3...Nc6 intending to just keep the bad bishop on f8 or d7 which is passive and obtains little counterplay. Also, 4...Qb6 5.Nf3 Bd7 intending 6...Bb5 to trade off the "bad" queen's bishop is possible.
After 5.Nf3, the main lines are 5...Qb6 and 5...Bd7. The idea behind 5...Qb6 is to increase the pressure on d4 and eventually undermine the white center. Qb6 also attacks the b2 square, so White's darksquared bishop cannot easily defend the d4 pawn without dropping the b2 pawn.
White's most common replies to 5...Qb6 are 6.a3 and 6.Be2. 6.a3 is currently the most important line in the Advance: it prepares 7.b4, gaining space on the queenside. Black may prevent this with 6...c4 threatening to take en passent if b4 is played. This creates a closed game where Black fights for control of the b3 square. On the other hand, Black may continue developing with 6...Nh6, intending ...Nf5. Nh6 seems strange as White can double the pawn with Bxh6, but this is actually considered good for Black. Black plays Bg7 and O-O and Black's king has adequate defense and White will miss his powerful dark squared bishop. 6.Be2 is other alternative, aiming simply to castle. Once again, a common Black response is 6...Nh6 intending 7...cxd4 8.cxd4 Nf5 attacking d4. When the king's knight reaches f5 from h6, there will be three pieces and a pawn attacking the d4 point while there will only be the Nf3, pawn c3, and Qd1 defending the d4 pawn. As said earlier, White's darksquared bishop cannot come to the rescue because the b2 pawn will fall. Thus, White must plan prophalactically and reply to 6...Nh6 with 7. Na3 so that the White knight can defend the d4 pawn.
5...Bd7 was mentioned by Greco as early as 1620, and was revived and popularized by Viktor Korchnoi in the 1970s. Now a main line, the idea behind the move is that since Black usually plays ...Bd7 sooner or later, he plays it right away and waits for White to show his hand. If white plays 6. a3 in response, modern theory says that Black equalizes or is better after 6...f6! The lines are complex, but the main point is that a3 is a wasted move is the Black Queen is not on b6 and so Black uses the extra tempi to attack the white center immediately. Black may continue 7...Nf5 to attack d4 or 7...Ng6 followed by ...f6 to attack e5.
A general theme in the Advance French is that White would like to put his light squared bishop on d3 where it maximizes its scope. White cannot play this move immediately, however, because the d4 pawn will fall. Clearly, Black cannot take the pawn immediately as 6.Bd3 cxd4 7.cxd4 Nxd4? 8.Nxd4 Qxd4?? 9.Bb5+ wins the Black queen by a discovered attack. Thus, theory says that Black should play ...Bd7 to prevent the discovery. Here White will sacrifice his d-pawn anyway by continuing 8.0-0 Nxd4 9.Nxd4 Qxd4 10.Nc3. This is the Milner-Barry Gambit, named after Sir Stuart Milner-Barry, and is considered to be of marginal soundness.
Another theme is that White wants to expand on the Kingside and attack the Black castled king. White will attempt to gain an attack by kicking the white knight on f5 with g4 or playing h4-h5 to kick the knight when it is stationed on g6. Because of the blocked center, sacrificial mating attacks are often possible. It is said by French players that the classic bishop sacrifice (Bd3xh7) should be evaluated every move. Black, however, often welcomes an attack as the French is notorious for producing stunning defensive tactics and maneuvers that leave Black up material for an endgame. Viktor Korchnoi who along with Botvinnik was the strongest player who advocated the French, talked about how he would psychologically lure his opponents into attacking him so that they would eventually sacrifice material and he would halt his opponents army and win the endgame easily.
The move differs from 3.Nc3 in several respects: it doesn't block the path of White's c pawn, which means he can play c3 at some stage to support the d4 pawn; and it avoids the Winawer Variation because 3...Bb4 can be met with 4.c3 when Black has wasted a move (he has to retreat his bishop).
3...c5 4.exd5 exd5, a staple of many old Karpov-Korchnoi battles, usually leads to Black having an isolated queen's pawn (see Isolated pawn). The main line continues 5.Ngf3 Nc6 6.Bb5 Bd6 7.0-0 Nge7 8.dxc5 Bxc5 9.Nb3 Bb6 with a position where if White can neutralize the activity of Black's pieces in the middlegame, he will have a slight endgame advantage. Another possibility for White is 5.Bb5+ Bd7 (5...Nc6 is also possible) 6.Qe2+ Be7 7.dxc5 to trade off the bishops and make it more difficult for Black to regain the pawn.
3...c5 4.exd5 Qxd5!? is an important alternative for Black. The idea is to trade his c- and d-pawns for White's d- and e-pawns, leaving Black with an extra centre pawn. This constitutes a slight structural advantage, but in return White gains time for development by harassing Black's queen. This interplay of static and dynamic advantages is the reason why this line has become popular in the last decade. Play usually continues 5.Ngf3 cxd4 6.Bc4 Qd6 7.0-0 Nf6 (preventing 8.Ne4) 8.Nb3 Nc6 9.Nbxd4 Nxd4, and here White may stay in the middlegame with 10.Nxd4 or offer the trade of queens with 10.Qxd4.
While the objective of 3...c5 was to break open the centre, 3...Nf6 aims to close it. After 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 (6...b6 intends ...Ba6 next to get rid of Black's "bad" light-squared bishop, a recurrent idea in the French) 7.Ne2 (leaving f3 open for the queen's knight) 7...cxd4 8.cxd4 f6 9.exf6 Nxf6 10.Nf3 Bd6 Black has freed his pieces at the cost of having a backward pawn on e6. White may also choose to preserve his pawn on e5 by playing 4.e5 Nfd7 5.c3 c5 6.f4 Nc6 7.Ndf3, but his development is slowed as a result.
3.Nd2 Nc6 is known as the Guimard Variation: after 4.Ngf3 Nf6 5.e5 Nd7 Black will exchange off White's cramping e-pawn next move by ...f6. However, Black does not exert any pressure on d4 because he cannot play ...c5, so White should maintain a slight advantage.
A fashionable line among top GMs in recent years is 3...Be7!?, an odd-looking move which aims to prove that every White move now has its drawbacks, e.g. after 4.Ngf3 Nf6 5.e5 Nfd7 White cannot now play f4, whereas 4.Bd3 c5 5.dxc5 Nf6 and 4.e5 c5 5.Qg4 Kf8!? lead to obscure complications. Amazingly, 3... h6?!, with a similar rationale, has also gained some adventurous followers in recent years, including GM Alexander Morozevich. Another rare line is 3...a6, the idea being to deny White's light-square bishop use of b5 before playing ...c5.
3...Bb4 pins the knight on c3 to the king, leaving the e4-pawn undefended. White normally moves his pawn into safety with 4.e5, gaining space and hoping to show that Black's b4-bishop is misplaced. Black usually replies 4...c5 followed by 5.a3 Bxc3+ 6.bxc3, resulting in the position below: While White has doubled pawns on the queenside, which Black's counterplay usually targets, they can also help White since they give White more control over the centre and a semi-open b-file. In addition, White has more room on the kingside, where Black is even weaker than usual because he has traded off his dark-square bishop. Combined with the bishop pair, this gives White dangerous attacking chances.
In the diagrammed position, Black most frequently plays 6...Ne7. (The main alternative is 6...Qc7, which can simply transpose to main lines after 7.Qg4 Ne7, but Black also has the option of 7.Qg4 f5.) Now White can exploit the absence of Black's dark-square bishop by playing 7.Qg4, giving Black two choices: he may sacrifice his kingside pawns with 7...Qc7 8.Qxg7 Rg8 9.Qxh7 cxd4 but destroy White's centre in return, the so-called "Poisoned Pawn Variation"; or he can play 7...0-0 8.Bd3 Nbc6, which avoids giving up material, but leaves the king on the flank where White is trying to attack. Experts on the 7.Qg4 line include Judit Polgar.
If the tactical complications of 7.Qg4 are not to White's taste, 7.Nf3 and 7.a4 are good positional alternatives. 7.Nf3 is simply a natural developing move, and White usually follows it up by developing the king's bishop to d3 or e2 (occasionally to b5) and castling kingside. The purpose behind 7.a4 is threefold: it prepares Bc1-a3, taking advantage of the absence of Black's dark-square bishop. It also prevents Black from playing ...Qa5-a4 or ...Bd7-a4 attacking c2, and if Black plays ...b6 (followed by ...Ba6 to trade off the bad bishop), White may play a5 to attack the b6-pawn.
Both sides have many alternatives to the main line Winawer. On the fourth move, some of White's options include:
After 4.e5 c5, White can also try 5. Bd2, again preventing the doubled pawns and making possible 6.Nb5, where the knight may hop into d6 or simply defend d4.
After 4.e5, Black does not have to reply 4...c5 but may try 4...b6 followed by ...Ba6, or 4...Qd7 with the idea of meeting 5 Qg4 with 5...f5. However, theory currently prefers White's chances in both lines. Another popular way for Black to deviate is 4.e5 c5 5.a3 Ba5, the Armenian Variation. Black maintains the pin on the knight, which White usually tries to break by playing 6.b4 cxb4 7.Qg4 or 7.Nb5 (usually 7.Nb5 bxa3+ 8.c3 Bc7 9.Bxa3 and white has the upper hand).
When white plays 7.Nf3 instead of 7.Qg4, it is called the Winawer Advance Variation. This line often continues 7...Bd7 8.Bd3 c4 9.Be2 Ba4 10. 0-0 Qa5 11.Bd2 and the fundamental position occurs after 11.Nbc6 12.Ng5 h6 13.Nh3 0-0-0. Its assessment is very unclear, but most likely black would be considered "comfortable" here.
4...Be7 used to be the main line after 4.Bg5 and is still important, even though the Burn Variation has overtaken it in popularity. The usual continuation is 5.e5 Nfd7 6.Bxe7 Qxe7 7.f4 0-0 8.Nf3 c5, when White has a number of options including 9.Bd3, 9.Qd2 and 9.dxc5. An alternative for White is the gambit 4...Be7 5.e5 Nfd7 6.h4, which was devised by Adolf Albin and played by Chatard, but not really taken seriously until Alexander Alekhine used it to defeat Fahrni in Mannheim 1914. It is today known as the Albin-Chatard Attack or the Alekhine-Chatard Attack. After 6...Bxg5 7.hxg5 Qxg5 8.Nh3 Qe7 9.Nf4 Nc6 10.Qg4 (the reason for 8.Nh3 rather than 8.Nf3), White has sacrificed a pawn to open the h-file, thereby increasing his attacking chances on the kingside. Black may also decline the gambit in several ways such 6...a6 and 6...f6, but most strong players prefer 6...c5. The Alekhine-Chatard Attack is not very popular at grandmaster level (though it is not completely unknown, Garry Kasparov using it successfully against Viktor Korchnoi in 2001, for instance), but is more often seen in amateur games.
A third choice for Black is to counterattack with 4...Bb4 (the MacCutcheon Variation), ignoring the threat of 5.e5, when the main line continues 5.e5 h6 6.Bd2 Bxc3 7.bxc3 Ne4 8.Qg4. At this point Black may play 8...g6, which weakens the kingside dark squares but keeps the option of castling queenside, or 8...Kf8. The MacCutcheon Variation is named for John Lindsay MacCutcheon of Philadelphia (1857-1905), who brought the variation to public attention when he used it to beat Wilhelm Steinitz in a simultaneous exhibition played in Manhattan in 1885.
The Steinitz Variation is 4.e5 Nfd7, when White faces a choice between 5.f4 (most common), 5.Nce2 (the Shirov-Anand Variation; White gets ready to bolster his centre with c2-c3 and f2-f4), or 5.Nf3 (aiming for piece play). After 5.f4 c5 6.Nf3 Nc6 7.Be3 (7.Nce2 transposes to the Shirov-Anand Variation; a trap is 7.Be2 cxd4 8.Nxd4 Ndxe5! 9.fxe5 Qh4+ winning a pawn), Black has several options. He may step up pressure on d4 by playing 7...Qb6 or 7...cxd4 8.Nxd4 Qb6, or he may choose to complete his development, either beginning with the kingside by playing 7...cxd4 8.Nxd4 Bc5, or with the queenside by playing 7...a6 8.Qd2 b5.
2.Qe2 is the Chigorin Variation, which discourages 2...d5 because after 3.exd5 the Black pawn is pinned, meaning Black would need to recapture with the queen. Black usually replies 2...c5, after which play can resemble the 2.d3 variation or the Closed Variation of the Sicilian Defence.
2.Nf3 d5 3.Nc3 is the Two Knights' Variation: 3...d4 and 3...Nf6 are good replies for Black. 2.b3 leads to the Réti Gambit after 2...d5 3.Bb2 dxe4, but Black can also decline it with 3...Nf6. 2.e5 is known as the Steinitz Attack, after the first World Champion who analysed and sometimes played it in the late nineteenth century. It is not considered a challenging response to 1...e6 and, as such, is relatively rare.
After 2.d4 d5 White can also choose the Diemer-Duhm Gambit with 3.c4?!, but this move is considered weak by theory. The most common continuation is 3...dxe4 4. Nc3 Nf6 5. f3, after which 5...Bb4, 5...c5, and 5...exf3 are all possible. Further information can be found on the DDG site
2... Nf6 is known as the Mediterranean Defense, and is very rare.