azaris - papi, Gameknot 2005:

1.e4 e5 2.Nf3 d6 3.d4 Nc6 4.Bb5 Bd7

The opening has transposed into the Old Steinitz Defense. If nothing else, Black has avoided the possibility of doubled c-pawns which the Exchange variation brings (with loss of tempo, if Black plays 3...a6 attempting to force the retreat of a bishop which then captures rather than retreats).

5.o-o exd4 6.Nxd4 Nge7 7.Bg5 Nxd4 8.Bxd7 Qxd7 9.Qxd4 Nc6 10.Qd5 f6 11.Be3 g6 12.Nc3 Bg7 13.Nb5 a6 14.Nd4 Nxd4 15.Bxd4 o-o-o 16.Rad1 Rhf8 17.Rad1 ...etc.

Maybe not the best way for Black (or even White) to handle it, but what about the general idea? Cat piss?

If I were White I would be trying to make it hard for Black to finish development and exchange pieces...maybe 7. Bc4 instead of 7. Bg5, for example? Losing a tempo doesn't seem so bad, since Black will certainly have to move the knight at e7 again. A number of Steinitz' opponents responded to ...Nge7 that way.

I do have to question the soundness of all this, though. When does black castle queenside in a King's Indian? More to the point, why would black do that? White has the queenside advantage and can rip open that side of the board when he pleases. I feel that this is a strategic mistake that will be more costly than it may first appear.

But note that White has locked the center and the pawns point toward his kingside; furthermore, it is his kingside that is most evacuated and that is (one supposes) the more likely side for him to castle on. I might seriously consider leaving my options open with a waiting move before setting up fully for castling. If White castles kingside then I think I might try to start a kingside attack beginning with a pawnstorm. So, you might try a waiting move of ...h6 and then if White castles o-o you can immediately try g5 with the idea of g4 to dislodge the knight. White may well play h3 (gosh, a light-squared bishop would come in handy for Black now) and now you have one knight you can put on f6 and another one which can replace it. You have two knights; a queen, a rook, and several pawns on the kingside, as well as a bishop that might have some use; put them all to use at once and even if you have to sacrifice a piece you might be able to mate or else to put White on the defensive and win back material as well as expand and positionally dominate on the kingside.

If all of this sounds hopelessly vague, it is. But it is an idea to consider if you want to try taking the initiative under conditions of a locked center. Strike before White gets set up to attack. If worse comes to worse try to position to exchange pieces and neutralize the possibility of a coordinate White attack on your king (which, incidentally, you might want to get to e7 fairly quickly to make way for your other rook, to form a battery (or simply cover) the h and/or g files you prize open. Obviously this isn't the sort of attack to worry about a pawn or two, *provided* that their capture doesn't offer White the prospect of a counterattack before you can finish your attack.

[Event "www.ChessWorld.net server game"]
[Site "www.ChessWorld.net "]
[Date "2007.2.9"]
[Round "NA"]
[White "darthplutonius"]
[Black "GDAVIS45"]
[Result "1-0"]
[Termination "Black resigned"]
[WhiteElo "1754"]
[BlackElo "1685"]
[Mode "ICS"]
[DateLastMove "2007.2.20"]
[ECO "C62"]
[Board "3373529"]

1.e4 e5 2.Nf3 Nc6 3.Bb5 d6 4.Bxc6+ bxc6 5.h3 Ba6 6.d3 Nf6 7.Bg5 Be7 8.O-O O-O 9.Nc3 Nd7 10.Bxe7 Qxe7 11.Re1 Rab8 12.b3 Bb7 13.a4 a5 14.Qd2 Nf6 15.Nd1 Nh5 16.Qg5 Qxg5 17.Nxg5 h6 18.Nf3 f5 19.c4 fxe4 20.dxe4 Rf7 21.Nh4 c5 22.Nc3 c6 23.Nf5 Rf6 24.Ne2 Bc8 25.Ne7+ Kf7 26.Nxc6 Rxb3 27.Nxa5 Rb4 28.Nc3 g5 29.Nd5 Rg6 30.Nxb4 ♗lack resigned 1-0

4....Nb4 is certainly a bad move, but after 4...Ne7 or ....Nb8 I don't agree that White's advantage is <very large>. He's a little better, that's all.

<ganstaman: I'm playing a game on gameknot right now that began (I'm white) 1. e4 e5 2. Nf3 d6 3. d4 Nc6 4. Bb5 Bd7 5. d5 <I discovered that this isn't so common, but I don't know why> Nce7 6. Bxd7 Qxd7...Looks to me like this is very good for white due to the whole good bishop/bad bishop thing. Sort of now an Old Indian, once I play c4.>

IF I were black I would prefer 5...Nb8 and if 6. Bxd7+ Nxd7. You're right that Black has the worse bishop but he's also traded off a pair of minor pieces -- no small accomplishment in a cramped game. A lot of positions that look tight are not really bad at all once a pair or two of minor pieces have come off the board. Here black has good squares for his knights on d7 and f6 and his bishop at g7 or e7. He can also try to trade off the bad bishop with ...g6 and ...Bh6. He's a little worse, but just a little. Most books that look at positions like this counsel white to withdraw his bishop with 6. Bd3 rather than exchanging it. Implicitly or explicitly, they are suggesting that the cramping effect of leaving the bishops on the board is more valuable than leaving Black with the worse bishop. I think the books are right. Note that after 6. Bd3 the black bishop is occupying the only square the Black QN can sensibly move to. ...Bg4 and ...Bxf3 solves the problem at the cost of leaving Black with a permanent positional disadvantage -- he's traded off his good bishop for a knight. After ...Bc8 and ...Nd7 he's lost time and his bishop is out of play. In other words, after 6. Bd3 Black is genuinely cramped; after 6. Bxd7+ Nxd7 he only looks cramped.

This leaves the question why 5. d5 is not more popular. It may be simply that the Old Steinitz Defense was more popular 100 years ago than now, and in those days open positions were more esteemed than closed ones. But a simpler reason is that, if White leaves his pawn on d4, eventually Black will be forced to play exd4, leaving White with a pawn on e4 against a pawn on d6. This gives him a slight but lasting advantage without giving black the readymade King's-Indian style counterattack that he gets after 5. d5. Or, 5. d5 may be unpopular because of this game:

Lasker vs Blackburne, 1899

Never underestimate the influence of a famous brilliancy on opening fashion.