The Kibitzer 
by Tim Harding

THE KIBITZER AND THE EVEN MORE COMPLETE
NAJDORF

The recent arrival of a new book by grandmaster John Nunn ("The
Complete Najdorf: 6 Bg5" from Batsford/ICE, 320 pages)
inevitably prompts musings on bygone days. In England they say
that expertise at snooker (a game of the billiards family, the
grandaddy of pool) is a sure sign of a mis-spent youth; the same
could be said about the Najdorf Sicilian, 1 e4 c5 2 Nf3 d6 3 d4
cxd4 4 Nxd4 Nf6 5 Nc3 a6.

In the days when the Kibitzer was a schoolboy player, in the
mid-1960s, the Najdorf was all the rage. At our school, the chess
club was run by an enthusiastic young math teacher and thanks to
him the school library took out a subscription to Chess Archives,
the English edition then published by Euwe in the Netherlands.
This was a good source of information about current master
practice, with lots of games and analysis you couldn't find
anywhere else in English at that time. It took the form of loose-leaf
sheets with punched holes that you fitted into a binder so that
articles on the same opening could be kept together.

For a few years I played the Poisoned Pawn but I kept an eye on
developments in other lines too. In the 1970s I gave up playing the
Sicilian with Black and with White 6 Bg5 became my weapon of
choice. I especially remember a few encounters in the London
League where competition was very keen. The Richmond club, led
by their young star (later grandmaster) Michael Stean, was a
hotbed of Najdorf players and as Athenaeum team captain I
generally managed to ensure I had White against Richmond. I
never met Stean as I wasn't playing top board but my opponents
were generally around the 2100-2300 mark.

It's a curious thing but I usually found it easier to win against the
higher-rated Najdorf players, maybe because they went all out to
win, whereas on the rarer occasions that a player below my rating
chose the Najdorf they might get a draw. The higher-rated player
may opt for a dangerous continuation, hoping to induce an error or
unleash a novelty, where the average player may stick to the safer
sub-variation.

The most memorable of these games came in the 1973 match
against Richmond. My opponent was one M.J. Lightfoot and the
game was played a few days after I had returned from competing in
a Wijk aan Zee Reserve Master Group, so I was on fairly good
form.

I probably won't fully succeed in recreating my thoughts as I was
then a much stronger OTB player than I am now, but I did make a
few notes at the time.

Before looking at the game, I want to digress by saying a little
about Nunn's book and its prehistory, although this isn't strictly
speaking a book review. The Batsford openings series began in the
late 1960s after a couple of years of the publication Chessman
Quarterly edited by International Master R.G.Wade. Bob Wade
used to lecture on chess from time to time at both Oxford and
Cambridge Universities and recruited many of the young masters
and experts there to write for first the magazine and then for
Batsford; part of his reasoning was that we had learned at
university how to research and analyse material and present it in an
organised way.

For one of the first Batsford books, however, a "name" author was
required and Belgian grandmaster A. O'Kelly de Galway wrote the
first Najdorf book. It was a big success but Najdorf books date fast
and in 1976 Batsford issued a replacement, Sicilian Najdorf (126
pages), by Michael Stean. I still have my copy in which the author
has inscribed: You'll never be able to find all the mistakes.

The next generation came in 1982, Sicilian Defence: Najdorf
Variation, essentially an updating by Nunn of Stean's book, and 60
pages longer. In 1988 Nunn first flew solo with Najdorf for the
Tournament Player (298 pages) and now we have a bigger book
still, dealing with only 6 Bg5. A second book on the rest of the
Najdorf will follow. So by making a much larger book (in two
volumes) it is possible to make some claim to 'completeness' for
the first time, and Nunn certainly aims to live up to his title! This is
a very thorough book and what follows should be read as an
addendum not criticism.

One of the problems, for a practical player, of investing much time
in studying a variation like the Najdorf, is you spend a lot of time
on positions arising at move 15, 18, 20 or later; all this work can be
rendered irrelevant by an improvement (or just a fashionable
alternative) a couple of moves earlier on; for a correspondence
player this is aggravated by the factor that the novelty or fashion
change may arise while the game is in progress, after you are
committed but before the line you are hoping for has been reached. 

One area where Nunn's book is very useful is indicating that, for
example, analysing line Y is a waste of time because three or four
improvements in various sub-variations would be required to
resuscitate it, whereas line Z may be a good bet as only one or two
novelties are required. The example in this column is of the line Y
variety, I fear; if I thought it was a type-Z I might keep it secret for
as few years more! 

Nevertheless many type-Y byways in the Najdorf have intrinsic
interest. Indeed I told Stean about the following game (reminded
him, I should say, as he was present when it was played) after
White's stunning 21st move below failed to appear in his 1976
book, but space was tight and he (or Nunn) evidently did not
consider White's idea sound enough to be worth a mention in the
next edition. But I still wonder The line has resurfaced because of
the far greater detail in Nunn's new book, thanks in part to taking
into a consideration a large number of correspondence games
which CC-grandmaster Maurice Johnson made available to him.

The Harding-Lightfoot game began, as Najdorfs generally do, by
rattling out a stream of theory, interrupted by pauses as each player
wonders which option to take, so that by the time the diagrammed
position was reached we had each probably used up half our time,
while still being in known territory. 1 e4 c5 2 Nf3 d6 3 d4 cxd4 4
Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Be7 I had given up the Poisoned
Pawn in the late 1960s after one particularly bad mangling in the
line 7...Qb6 8 Qd2 Qxb2 9 Rb1 Qa3 10 f5. It was much more fun
when opponents chose 10 e5, naively thinking they were going to
smash you up. Sometimes I had played 7...h6 8 Bh4 Qb6 but
objectively it's inferior as Nunn's new magnum opus confirms. If
Black doesn't want to play the main line with 7...Be7 there's the
Polugaevsky, 7...b5, which Nunn says is playable. But in 1973 that
had not really percolated down to the level where I was playing. 8
Qf3 Qc7 9 0-0-0 Nbd7 Another main division occurs here. This
game was played less than six months after the Spassky-Fischer
match in Reykjavik. Spassky had played 10 Bd3 here in Game 15,
sending theory off in a whole new direction, and I reckoned that
Richmond would have concentrated their recent attention on that
move. Also I had invested many hours in studying the old 10 g4
without being able to put it into practice before so I stuck to the old
ways. 10 g4 b5 11 Bxf6 Nxf6 Black has three ways to recapture
but I never met 11...gxf6, on which Nunn devotes eight pages to
proving an advantage for White. It is interesting to note that Stean's
1976 book says: "As always in these lines, 11...Bxf6? allowing 12
Bxb5 axb5 13 Ndxb5 followed by 14 Nxd6+ would be very good
for White." I won a game as White that went this way, against a
man called van Dop at the Amsterdam IBM Reserves Masters in
1973; this was never published and Nunn quotes two 1990s
examples instead. The two 1980s Nunn books just give 11 Bxf6?
12 Bxb5 in parentheses, so an indication of just how more
complete the new book is that Nunn now admits that "This
sacrifice does not have the lethal effect once supposed." He gives a
column of analysis to show that 12...Rb8! is unclear, so that White
might do better with 12 g5. So in the Najdorf you can never be sure
that a door has closed forever! 12 g5 Nd7 In this position Nunn
says that the main winning try is the pawn offer 13 f5 which indeed
I was aware of since those Archives articles, but I never
investigated it deeply. 13 a3 was the older move and I chose that.
13 a3 Rb8 14 Bh3 If 10 g4 was one step away from the prevailing
fashion of 1973, and 13 a3 the second, then with this move I took a
third and even more definite step away from the trendy path. 14 h4
had been introduced circa 1969 but I stuck with the lines I had
studied in my schooldays and which I then believed to be superior
to their theoretical reputation. In Nunn's book, 14 Bh3 receives just
under two pages. Since White threatens to sacrifice his bishop on
e6, Black has little choice and the play to move 17 is virtually
forced. 14...Nc5 15 Rhg1 b4 16 axb4 Rxb4 17 f5 (See Diagram)
17...Qb7 The 1982 book calls this "safest," a comment which has
been carried forward into the subsequent Nunn books but I wonder
has he really looked closely at the diagram position? The
alternatives are 17...Qa5 (comparatively recently introduced),
17...Bd7?! (analysed in the 1988 book only) and 17...Qb6 which
invites a fork on d5 (after 18 fxe6 fxe6 19 Nxe6) so that White can
get rook and pawn for two minor pieces. About two weeks after the
present game I reached the same position again in the League but
my lower-rated opponent preferred 17...Qb6 when the 1980s
Nunn/Stean assessment "with perhaps an edge for White, but most
likely a draw" is a fairly accurate description of what happened
after two sessions of play. As Stean wrote, 17...Qb7 is less forcing
but avoids the combination and allows the queen to defend along
the second rank. Its advantage for White is that it gives White time
to cook up something: 18 f6 gxf6 19 gxf6 Bf8 20 b3 a5 This was
an O'Kelly suggestion, threatening a4, and it's where Stean stopped
in his 1976 book. Instead 20 Bh6+ 21 Kb2 Bd7 22 Qh5 Be3 23
Qxh7! Rf8 24 Rg8! +- Buljovcic-Bertok, Yugoslav Championship
1965, a reference which resurfaces in Nunn after 25 years
forgotten. He also mentions 20...Qb6!? (See Diagram) 21 Nd5!!? I
have waited 21 years to see this move in a theory book;
presumably Nunn found out about it in the CC database he
received from Johnson. ECO B gives only 21 Qe3 (unclear) and 21
Rge1? (plain bad). I found 21 Nd5 at the board, by what mental
process and at what expenditure of time I cannot recall, but later I
searched Bob Wade's library thoroughly for a precedent.
Somewhat to my disappointment, I found in the March 1970 issue
of Fernschach that it had twice been played in German CC
tournaments of the late 1960s. Karl-Heinz Maeder (later a World
Correspondence Championship finalist) lost both of them. After
that, and the drawn game, mentioned above, I never again tried to
play this line but I would be interested to find out whether or not
its soundness (or otherwise) can be established. Naturally if you
give the position to a computer it will tell you White lacks
compensation for the sacrificed piece, but with the long-term
exposure of Black's king it's not easy to come to a final conclusion
about it. 21...exd5 Black has to accept. Breum-Maeder, corr
(BdF/M/176) 1967, went 21...a4 22 Qc3 Rxd4 23 Rxd4 exd5 24
Bxc8 Qxc8 25 Rxa4 d4 26 Rxd4 Bh6+ 27 Kb1 Bf4 28 Rd5 (clear
White advantage says Nunn) 28 Qe6 29 Qc4 Be5 30 b4 Nb7 31
Qb5+ Qd7 32 Rxe5+ dxe5 33 Qxe5+ Qe6 34 Qxe6+ fxe6 35 Rg7
Nd8 36 b5 1-0. 22 exd5!? This was actually the first new move of
the game! Whether it's best I am not sure but no doubt I gave it a
lot of thought at the time. W.Frey - Maeder, 11th West German CC
Ch prelims 1968-71, went instead 22 Bxc8 Qxc8 23 exd5 Qb7? 24
Rde1+ Kd7? (Frey wrote in Fernschach that Black must play
24...Kd8.) 25 Rg7 Kc8 26 Re8+ Kd7 27 Rb8!! (27 Qe2 Qxd5) 27
Qxb8 28 Rxf7+ Ke8 29 Qh5 1-0. The critical point here is Black's
23rd move. Frey wrote that 23 Rxd4 is the alternative and Nunn's
new book then gives 24 Rxd4 Nd7 unclear. My notes show that
when I demonstrated this game to Stean, he suggested 23...a4 24
Rde1+ Kd7 e.g. 25 Qg4+ Kc7 26 Ne6+ Qxe6 27 Qxb4 Qxd5
unclear. My move avoids this line and, by postponing the bishop
exchange, gives Black more to think about, a good practical point
against the clock. 22...Bxh3 Nunn also cites another
correspondence game Fechner-August (East German Ch 1975)
which went 22...Rxd4 23 Rxd4 Nd7 24 Qc3 Nb6 25 Re1+ Kd8 26
Bxc8 Bh6+ 27 Kb1 Qxc8 28 Qxa5 with an edge to White
according to Nunn. The remaining moves were 28 Qc5 29 Qa1
Bg5 30 Rd3 Re8 31 Rxe8+ Kxe8 32 Rg3 Qxd5 33 Qc3 Kd7 34
Rg1 Bd2 35 Qh3 + Kc7 36 Rd1 Qe4 37 Qh5 Bc3 38 Qxf7+ Nd7
39 Qd5 Qxd5 40 Rxd5 Nxf6 41 Rd3 Be5 42 Rf3 Kc6 43 c4 Kc5 44
Kc2 Kd4 45 h3 Nh5 46 Rf7 h6 47 Rc7 Nf4 48 Rh7 h5 49 Rh6 Kc5
50 Rh7 draw. 23 Qxh3 a4? This is where Black finally goes wrong,
overlooking the reply no doubt. He doesn't have time to launch a
queenside counter-attack but must look to the defence of his king. I
analysed this game with Wade later and he came up with 23...Qd7
24 Rde1+ Kd8 25 Ne6+! Kc8 26 Qc3 unclear/White advantage
(Fritz4 disagrees); maybe 24 Qe3+ is better. 24 Nc6! This takes
away the d8 flight square and covers e7 so Re1+ becomes a deadly
threat. Black's reply is forced after which White regains his piece
with a winning position since Black's king's rook and bishop
remain idle bystanders. 24 Re4 25 Rde1 Qd7 26 Rxe4+ Nxe4 27
Qe3 Qf5 28 Re1 Qxd5 29 Nb4 Qf5 30 Qxe4+ Qxe4 31 Rxe4+ Kd7
32 bxa4 Rg8 Here, as was the way with League games in those
days (the time control was at move 30), the game was adjourned
and Black resigned without resuming (1 0). 