Archival Quality

by Jon Edwards

In this information age, we have come to believe that access to good information requires a computer and a network connection. In our world of correspondence chess, we have come to believe that success depends more on the size of one's database and the power of one's computer than inherent experience of the player or the number of volumes in one's library. Or does it?

In this important game, the key reference, Reshevsky-Botvinnik 1948 appeared in my database, but without notes or any sense of its importance. Fortunately, I have ready access just 100 yards away to one of the finest chess libraries in the world, Firestone Library at Princeton University. This little known collection was assembled by Eugene Cook, a composer of problems best known today for having lent his name to the term "to cook" a chess composition.

Opening manuals appear to have forsaken Reshevsky's fine opening and middlegame ideas in this game in no small part I suspect because he lost the game in time trouble. But the notes that I found in the periodicals of the day made it clear that Black has a tough row to hoe in this variation. There's a lesson there somewhere. Our reliance upon the computer can take us just so far. In the end, like a fine wine, there's no meaningful replacement for a seminal book , a master's glance, or just an old, well-written periodical.

Edwards,J (2470) - Pedersen,N (2420)

[C18 French: Winawer]
NAICCC VIII corr, 1996

1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Qc7

Deviating from our last encounter: 6...Ne7 7.Qg4 0-0 8.Bd3 f5 9.exf6 Rxf6 10.Bg5 Rf7 11.Qh5 g6 12.Qd1 Nbc6 13.Nf3 Qa5 14.Bd2 c4 15.Be2 Nf5 16.0-0 Nd6 17.Ng5 Re7 18.f4 Bd7 19.Bf3 Rf8 20.g4 h6 21.Nh3 Ref7 22.Bg2 Kg7 23.Qe1 Qd8 24.Qg3 Bc8 25.Nf2 Ne7 26.a4 b6 27.Rae1 Nb7 28.h4 Qd6 29.Nh1 Rh8 30.Re5 Qa3 31.f5 exf5 32.h5 g5 33.gxf5 Qd6 34.Bxg5 hxg5 35.f6+ Rxf6 36.Rxe7+ Kh6 37.Qxd6 Rxd6 38.Ng3 Na5 39.Bxd5 g4 40.Bf7 Bd7 41.d5 1-0


In my view, given the concessions that White has already made on the queenside, Qg4 is essential.


This move is so logical that I am surprised that I haven't had to face it before this game. By playing ...Qc7 first, White cannot now capture on g7. For Black, that's the good news. The bad news is that ...f5 seriously weakens the dark squares on Black's kingside and Black will no longer be able to strike back at the White center with ...f6..


Maintaining the pressure on ...g7 and covering c3. There's no serious alternative. 8.Qh5 g6 9.Qd1 Nc6 10.Nf3 Bd7 11.dxc5 h6= Spassky-Portisch, 1980.

8...cxd4 9.cxd4 Ne7 10.Bd2 0-0 11.Bd3 b6

With the idea of trading off the bad bishop with ...Ba6.

12.Ne2 Ba6 13.Nf4 Qd7

Amateurs frequently play f4 in positions like this. The pawn move would be a serious error. In order to progress, White must take full advantage of the dark-square weaknesses on the kingside with moves like Nf4, h4-h5-h6, and with Bg5-g6 should Black have to advance the g-pawn.

14.Bxa6 Nxa6 15.Qd3 Nb8

We're still following Reshevsky-Botvinnik, 1948. The knight has no better alternatives, and 15...Qc8 only invites 16.Nxe6


In many lines, the h-pawn will push to h6 in order to loosen up Black's kingside pawns. Of course, the move also frees up the possibility of Rh3-g3.


Botvinnik continued with 16...Nbc6 17.Rh3 Rac8 18.Rg3 Kh8 19.h5 (with the threat of Ng6+) 19...Rf7 20.h6 (perhaps 20.Kf1 first) 20...g6 when 21.Kf1 would have given White a tangible opening advantage. Pedersen is hoping to save a tempo for his development by speeding the rook to c4, but it is simply not clear what the rook can accomplish on the natural square for a knight.

17.Rh3 Rc4 18.Rg3 Nbc6 19.c3

The threat on the d-pawn forces White to solidify the structure with c3. The battle lines are now clearly drawn. Black will push his queenside pawns in an effort to undermine the c3-pawn. White will have to attack the Black kingside in a close-quarter drill, with Bg5 (following a retreat of the Nf4) and with h4-h5-h6.



A curious maneuver, essentially switching places with the king's rook. Pedersen correctly saw the mounting kingside pressure and counts on ...Rf7 to cover the key squares on seventh rank.


This may be an error, though I am not sure how Black can punish me for it. I remain convinced that I should first have played 20.Kf1 with the idea of Re1. Ironically, the correct move order here might have dissuaded Pedersen from trying 23...Kf7 later because his best replay, I believe, would have been 20...Rf7.

20...b5 21.h6 g6 22.Ne2

These moves look so obvious in retrospect, but they required a huge expenditure of time. The idea here is Bg5-f6, Nf6, Qd2 with many sacrifices (Nxg6, Nh5, Nxe6, Nxd5) in the air.

22...a5 23.Bg5 Kf7

In many lines, Black prepared for Bg5-f6 by playing ...Kh8 and ...Ng8. Here, Pedersen has tried to push his queenside quickly and does not have time now for the usual defense. He therefore attempts to guard the f6-square with an unusual but very dangerous posting of the King.

24.Qd2 Ng8

or 24...Rc8 25.Bf6 Ng8 26.Qg5 with the idea of Nf6-g6.


White's preparations are nearing completion. The immediate threat is Nxg6. I remain impressed that White can succeed here without first having played Kf1 and Re1.

25... Nce7 26.Bf6

In a similar position, Reshevsky also played Bf6. There unfortunately, it did not work. But I certainly credit him with giving me the idea for my play here. If he declines to capture the Bishop, I will continue with Bg7 and probably Nh3-g5.

26...Nxf6 27.exf6 Kxf6

At the cost of a pawn, the Black King is now badly exposed. The Rook on c4, posted there more than ten moves ago with an uncertain purpose, now becomes a liability. [diagram]


The obvious threat is 29.Qg5+ Kf7 30.Ne5+ winning the Queen. Black's reply is forced.


Stopping the threat by covering e5, but the Rc4 is now trapped. Black seeks activity, but White now has a clear path to simplification.

29.Nb2 e5 30.Nxc4 dxc4 31.dxe5+ Nxe5 32.Qxd7 Nxd7 33.Rd1

White's first tough decision for quite a while. I could have my King on c1 here for free, but it seemed more appropriate on the kingside. The rest is a moderately difficult endgame that provide with little comment, save that the White King does indeed belong near the kingside pawns.

33...f4 34.Rf3 Ke6 35.g3 g5 36.gxf4 gxf4 37.Rh3 Nf6 38.Rd4 Kf5 39.a4 Re8+ 40.Kf1 bxa4 41.Rxc4 Re4 42.Rc5+! Re5 43.Rxe5+ Kxe5 44.Rd3 Ne4 45.c4

All with the idea of making sure that I place the rook behind Black's queenside passer.

45...a3 46.Rd5+ Kf6 47.Rxa5 Nd2+ 48.Ke2 Nxc4 49.Ra4 a2 50.Rxa2 Kg5 51.Ra6 Ne5 52.Ra5 Kf6 53.Ra7 Kg6 54.Re7 Nf7 55.Kf3 Nxh6 56.Kxf4 Nf7 57.Re6+ Kg7 58.Kg4 Nh6+ 59.Kg5 Nf7+ 60.Kh5 Nd8 61.Ra6 Nf7 62.f4 Kf8 63.f5 Ke7 64.f6+ Kd7 65.Rb6 Ke8 66.Rb7 1-0

The final position is zugswang. Both King moves lose the N/f7. All Knight moves lose the h-pawn. ...h6 permits the White King to enter the position at g6.