The only player in the world who can play the French from either side, and castle first, without high risk of being mowed off the board, is Carlsen.>

I used to think the French sucked even back before I ever cracked open a chess book or played through a game other than my own. As a coffee house player in my teens, whenever I say black play 1...e6, I already knew that if they castled short, I was going to play Bxh7+, Ng5+ and Qh5 mating. I thought I was the only person who knew this trick lol

Then, when I was about 20, I got the idea to pick up a chess book from the library, just to see if it contained some information about chess that I didn't already know (lol). I picked up Chess Praxis by Nimzovich and was blown away.

I quickly realized I don't know a damn thing about real chess, but more interestingly, I saw that Nimzovich played the French quite often. He also played 1...Nc6 which turned into something of French often enough. I was interested to see what he had to say about it and how he played it, and why anyone would ever choose to open a game in such a way.

I've been playing the French ever since!

In some ways, I think the French gets to the very essence of chess. In the Winawer, we see all kinds of interesting imbalances. There is white's weak pawn formation (doubled pawns, weak c4 square) vs blacks solid position, but white has more space on the king side. We see the battle of the two bishops versus two knights, but the jury is out until the pawn structure is settled in favor of the bishops or knights!

The French pits the old classic system against the hypermodern system quite naturally as well. Black tries to prove that the pawn on e5 is overextended and white tries to shore it up proving it is an advantage. In some sense, the Advance Variation of the French is a timeless discussion of this them. White neglects development to push the pawn to e5 on the third move, allowing black to undermine the pawn with ...c5, ...Nc6, ...Qb6 and ...f6. Black is often slightly ahead in development after 9 moves or so. Which imbalance wins? Blacks better development and more compact position, or White's space advantage and safe king?

We see in some variations the quintessential exchange sac ...Rxf3!?. This is cool when black can follow up with a kingside attack, but even cooler and more profound when the idea is to undermine the center.

In the French there is room for all kinds of ugly beautiful moves, like ...Nh8 (intending ...f6 and ...Nf7).

Fischer more or less hated the French too, but he famously struggled with it, often changing his opening approach trying to find something that worked. He felt it should be unsound, but said that one day he might have to admit it's sound after all.

Highly creative players have enjoyed the French like Nimzovich, Petrosian, Korchnoi, Botvinnik, Carlsen, Morozevich.

I think the French offers a richness in chess themes like no other and epitomizes the essence of chess in many ways.

In any event, I hate the French Defense!!!>

Umm, that's not true. Wesley So, Caruana, Karjakin, and Morozevich have all played it with great success. Black just misplayed the position. It's not that easu to crack this specific variation as long as you don't move you kingside pawns.>

Unlike Fischer, I don't think it is unsound.

And I didn't say it can't be successful.

What I am saying is that, with few exceptions, the notable games that receive many comments and commentators are characterized by the loser castling first (if anyone bothers to castle at all).

Other than Carlsen games, I only recall seeing 1 well-known French Defense game where the winner castled first.

Hopefully it is obvious that my comment is a bit "tongue-in-cheek", meant to be humorous. And my game sample size is not large, maybe a couple dozen.

But it happens often enough to be noticeable.

Want to bid spatial advantage against play versus IQP? Try Open Tarrasch variation: 2.d4 d5 3.Nd2 c5

Wish to play "standard" French position with positional chances for both sides? There's always Steinitz variation - 2.d4 d5 3.Nc3 Nf6 4.e5

You're interested in wildly tactical slaughterfest where both sides try to rip each other open? Look at 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 Qc7 8.Qxg7

Or maybe you want to release the tension and play symmetrical, if a bit dull, position: so there's Exchange variation - 3.exd5

You like to see highly opposite strategies where e.g. White tries to strangle Black, who in turn tries to rip open White's position - so consider Big Center Variation - 3.Nd2 Nf6 4.e5 Nfd7 5. c3 c5 6.f4 Nc6 7.Ndf3

Perhaps something completely offbeat? 2.b3 ?

All of these lead to completely different types of position, where castling may or may not be a priority, and certainly isn't correlated with winning strategies. Still, if you want a straightforward answer, I quickly found some famous French games where winner castled either first and he was the only one to castle:

Reshevsky vs Vaganian, 1976 - Black castles and destroys White's seemingly safe uncastled King.

Keres vs Capablanca, 1938 - positional struggle where White castled 5 moves earlier than his opponent.

Najdorf vs Gliksberg, 1929 - Black delays castling for too long and gets mated

Nimzowitsch vs Alapin, 1914 - Black gets waaay to greedy, neglects development and gets punished

Kasparov vs Radjabov, 2003 - famous game where both sides delayed castling to mid-game

Charousek vs Maroczy, 1897 - White's plan to delay castling and go for caveman attack backfires

Alekhine vs Fahrni, 1914 - one of best attacking games vs seemingly impenetrable Black's "safe" position

Geller vs Karpov, 1976 - great attacker's dismantling Black's non-castled position

click for larger view

Analysis by Houdini 4: d 25 dpa done

1. + / = (0.53): 17...Qg7 18.Kf1 b6 19.Rh3 Rg8 20.g3 Bb7 21.Qf4 Ng6 22.Qe3 Ke8 23.Re1 Rc8 24.Kg1 Rh8 25.Nd4 Nxd4 26.Qxd4 Ne7 27.Qa4+ Nc6 28.Bb5 Kf8 29.Qf4 h5 30.Bd2 Qg6 31.Rh4 Ne7 32.Bb4 Rxc2

click for larger view

Analysis by Houdini 4