[Event "It Open"]
[Site "Linares, Spain"]
[Date "1995.??.??"]
[EventDate "?"]
[Round ""]
[Result "½-½"]
[White "Luther, Thomas"]
[Black "Gelpke, Peter"]
[ECO "B17"]
[WhiteElo "?"]
[BlackElo "?"]

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.Qe2 Bd6 8.N1f3 h6 9.Ne4 Nxe4 10.Qxe4 c5 11.Bd2 Nf6 12.Qh4 Qc7 13.dxc5 Bxc5 14.0-0-0 Bd7 15.Rhe1 0-0-0 16.Ne5 Be8 17.f4 Bd4 18.Nf3 Bc6 19.Nxd4 Rxd4 20.Qh3 Rhd8 21.Kb1 Ne4 22.Be3 Rb4 23.Ka1 Qa5 24.Bxe4 Rxe4 25.Rxd8+ Kxd8 26.Qg3 Bd5 27.a3 g6 28.Kb1 Ke8 29.Rd1 Qc7 30.b3 b5 31.Rd3 a5 32.Kb2 b4 33.a4 Qd8 34.Qf2 Qf6+ 35.Kb1 g5 36.g3 gxf4 37.Bxf4 Kf8 38.Bd6+ Kg7 39.Bf8+ Kg6 40.Qxf6+ Kxf6 41.Bxh6 Re2 42.Rd2 Re1+ 43.Kb2 Ke7 44.h4 Kd7 45.Rh2 Bf3 46.h5 Re5 47.Bg7 Rxh5 48.Rxh5 Bxh5 49.c3 Kc6 50.cxb4 axb4 51.Bf8 f5 52.Bxb4 e5 53.Bd2 f4 54.Bxf4 exf4 55.gxf4 Kd5 56.Kc3 Ke4 57.b4 Kxf4 58.Kd4 Kf5 59.Kd5 Bd1 60.a5 Be2 61.Kc6 Ke6 62.Kb7 Kd5 63.a6 Kd6 ½-½

[Event "Ch Netherlands (team)"]
[Site "Netherlands"]
[Date "1996.??.??"]
[EventDate "?"]
[Round ""]
[Result "1-0"]
[White "Hoeksema, Hendrik Pieter"]
[Black "Gelpke, Peter"]
[ECO "B17"]
[WhiteElo "?"]
[BlackElo "?"]

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.N1f3 Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4 c5 11.Bd2 Nf6 12.Qe2 cxd4 13.Nxd4 Qe7 14.0-0-0 a6 15.Rhe1 0-0 16.Nf5 exf5 17.Qxe7 Bxe7 18.Rxe7 b5 19.Re5 Be6 20.Bxf5 Rfe8 21.Bxe6 fxe6 22.Ba5 Nd5 23.b3 Rac8 24.Kb2 Kf7 25.Rd4 Rc5 26.f4 Rec8 27.Re2 g6 28.g4 Kf6 29.Be1 R5c6 30.Bh4+ g5 31.Bg3 a5 32.f5 Nb4 33.Be5+ Ke7 34.Rdd2 exf5 35.c3 Na6 36.gxf5 Nc5 37.Bd6+ Kf6 38.Rd5 Nd7 39.Be7+ 1-0

The first thing I notice is that both games were played after the 1988 Psakhis--Meduna game. Gelpke seems to be a specialist in the 4...Nd7 line of the Caro-Kann, so his opponents may have been ready for it.

Luther, the strongest player (ca. 2600) chose 12.Qh4, playing an early dxc5 to give his queen a little extra breathing room. He castled queenside, and generally went to establish his knight on e5; however, after Black castled queenside, White's queen looked out of place. An extra pawn in the endgame turned out to be purely ornamental.

Hoeksema, "only" rated about 2400, chose 12.Qe2. THe queen turned out to be well-posted here, since after Black's early (possibly premature) 12...cxd4, White's 13.Nxd4 generated some threats with the idea of Nf5. Black stopped these, but only by misplacing his queen on e7. This time, the extra pawn was decisive, especially since Black's knight was not a light-squared bishop.

To see something of what we might be trying to accomplish, look at this game: GM_Kramnik (2753) - GM_Ponomariov (2695) [B17]
MTel Masters Sofia, Bulgaria, 12.05.2005

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Nd7 5.Ng5 Ngf6 6.Bd3 e6 7.N1f3 Bd6 8.Qe2 h6 9.Ne4 Nxe4 10.Qxe4 c5 11.0-0 Nf6 12.Qh4 cxd4 13.Re1 Bd7 14.Nxd4 Qa5 15.Be3 g5 16.Qh3 0-0-0 17.Qf3 Be5 18.a4 Nd5 19.Nb5 a6 20.c3 Nxe3 21.Qxe3 Bc7 22.b4 Qb6 23.Qxb6 Bxb6 24.Nd6+ Kb8 25.Nxf7 Rhf8 26.Nxd8 Bxf2+ 27.Kh1 Bxe1 28.Nxe6 Bxe6 29.Rxe1 Bb3 30.a5 Rc8 31.Rc1 Bc4 32.Be4 Re8 33.Bf3 Kc7 34.Kg1 Re3 35.Kf2 Rd3 36.Be4 Rd6 37.Ke3 Bb5 38.Rc2 Bc6 39.Bxc6 Rxc6 40.Kd4 Rf6 41.Ke5 Rf1 42.h3 Kc6 43.Rd2 Re1+ 44.Kf5 Rc1 45.Rd3 Kb5 46.Rd7 Rxc3 47.Rxb7+ Ka4 48.Kg6 Rg3 49.b5 axb5 50.a6 Re3 51.Kxh6 Re6+ 52.Kxg5 Rxa6 53.h4 Ra8 54.g4 Rg8+ 55.Kf4 Rf8+ 56.Kg3 Ka5 57.h5 Ka6 58.Re7 b4 59.h6 b3 60.Re3 White wins 1-0

Yes, this is the 10.0-0 line, but I think it's important to look at the differences. Kramnik chooses 12.Qh4, then plays an early Re1 to control the e-file. This seems a typical idea in this line.

We have a multitude of ways to go wrong.

(1) 13.c3, giving the KB some air and thinking about answering ...cxd4 with cxd4 ourselves, going for control of e5.

(2) 13.dxc5, controlling e5 indirectly by luring away Black's bishop. The problem might be that its new diagonal could be unpleasant for us after ...Qb6. <sd> and I looked at this a bit tonight, and he brought up the idea of 13.dxc5 Bxc4 14.Bf4!? denying c7 to Black's queen. There are interesting complications after 14...Qb6.

(3) 13.0-0-0, throwing the onus back on Black and not fearing 13...cxd4 14.Nxd4.

(4) 13.0-0, but Black can gain a tempo with 13...cxd4 14.Nxd4 Qc7, if that's important.

Of the choices presented my personal favorite is 13.c3 which goes a long way to establishing a superior pawn structure in the endgame, especially since our opponents played 12...a6 without us first having placed our bishop there. It sort of nullifies their prophylaxsis while establishing our goals to the endgame.

The problem with 13.dxc5 is that 14...Qb6 is almost assured based on previous games with positions very close in character to ours. In those cases it becomes very easy for black to dictate or hold the draw if they so desire. I don't know about you guys, but I want to win! See here: Nunn vs Dolmatov, 1992 and J Benjamin vs G Orlov, 1994

13.O-O-O I also like this move a lot. Any move that throws the onus on black is a good one, and they will most likely take on d4 since we won't, so it's good to have a prepatory move that allows our knight to take back. Besides, castling on opposite sides makes for a more interesting game longterm!

As for 13.O-O I don't see what good it does because of the lost tempo and the fact that we will be responding rather than dictating for the next three or four moves.

there are also other ideas after 13.dxc Bxc as well.

im sticking 13.dxc. i guess we'll see what <pb> wants to do.

click for larger view

It's not as bad as I first thought actually. It looks fairly even though (not a complaint necessarily!). Once both sides castle I'm not sure where our next plan of attack comes in.

Its main point is to prepare a good sqaure for the QB on c3 by stopping Black's attack on it by ...Nd5. It also discourages ...b5, prepares to push the queenside pawns, and threatens d5 in some cases.

If 13...cxd4 14.Nxd4 with the idea of 15.Nf5.

13.0-0-0 just seems too committal, especially if we want to push the queenside pawns.

I haven't decided on the relative merits of 13.c3 and 13.dxc5--both have good and bad points.