Professional Documents
Culture Documents
Chernev Most Instructive Critique PDF
Chernev Most Instructive Critique PDF
As a teenager, having read Irving Chernev’s Logical Chess: Move by Move (1957), I was glad to
come across The Most Instructive Games of Chess Ever Played (1965) not long afterward. It was
a natural progression, the former book being aimed at beginners and neophytes, the latter at
somewhat more advanced players. Both greatly aided my chess development.
However, while Chernev was a good teacher, he was never a great player, probably around 2200
strength at best. And having already done computer-assisted analysis of works by various greats
— Lasker, Alekhine, Euwe, Tartakower, Najdorf and others — and finding a surprisingly high
level of error in some, I wondered how Chernev’s notes would stand up under scrutiny by a
lidless silicon eye. So the games and notes in TMIGoCEP were examined with Komodo 11.2.2
and Stockfish 8, both rated about 3400 Elo, running through ChessBase 14 on a Dell Inspiron 17
7000 Series with an Intel Core i7-7500U CPU at 2.90 GHz with 16 GB RAM and a 64-bit
operating system. The results are presented here.
Rather than the usual, somewhat vague annotation symbols (i, o, y, u etc.), I have generally
used the engines’ numerical assessment to indicate the status of a position, as I consider this
more precise and informative. A position where White is up knight for pawn, and another where
he’s up a queen, would both get a “i” in the usual symbology, but there is obviously a big
difference.
The numbers represent the engines’ evaluation of the position to the nearest hundredth of a
pawn, e.g. a difference of exactly one pawn, with no other relevant non-material differences, has
the value +1.00 when in White’s favor, or -1.00 when in Black’s. A position where White is
considered better by 3½ pawns (or the equivalent, such as a minor piece) would get the value
+3.50, the advantage of a rook +5.00, etc. These numbers should not always be taken entirely at
face value, especially to the right of the decimal point, and they may vary some from one
machine to another, or with the time allowed for analysis, but they are generally valid and
reliable, and serve as useful shorthand for assessments and comparisons that would otherwise
require extensive detailed explanation. When the “royal we” is used, I am speaking for both the
engines and myself.
I have tried to find and highlight only mistakes and improvements that were significant,
especially those that would change a game’s result. There were many cases, especially in the
latter stages of games, where Komodo improved on a Chernev variation, say, finding a forced
mate where his note won a rook, but since both were clear wins I did not bother to point it out. In
some games Komodo’s evaluation was up to +/-4.00 or more well before resignation finally
came, and frankly I paid little attention, or even ended the analysis, once such a stage was
reached. Perhaps some howlers lurk in those denouements, but I doubt it. I was not concerned
with changes in opening theory since 1965.
This supplement is best used along with a copy of the book, since no complete games are given
here, only the relevant fragments. However, diagrams are plentiful, so the reader should find it
easy to set up a given position on his board and play through the analysis. Or if (unlike this
writer) he’s blessed with good powers of visualization, even without a board.
I hope you find the work interesting. Frankly, I did not, unlike previous similar projects. Chernev
proved to be too easy a target, and detailing mistake after mistake became a tedious chore rather
than a voyage of discovery. Therefore I did not proceed past Game 22 of the 62 Chernev
presented in TMIGoCEP. I may eventually return to the task, but at this point I would say the
probability is low. Any futher work I do in this line will involve books by better players.
At Black’s 33rd move, a very important alternative goes unmentioned by Chernev. It is given by
Reinfeld in The Immortal Games of Capablanca, and by Golombek in Capablanca’s 100 Best
Games of Chess, but their supporting analysis, borrowed from Tartakower himself, is inaccurate.
Rather than 33...Ne3-f5 as actually played, they say Tartakower originally intended 33...Nd1,
cuuuuuuuuC
{wDwDwiwD}
{0w0wDwDR}
{w0rDwDpD}
{DwDpDw)w}
{PDw)w)wD}
{Dw)BdKDw}
{wDwDwDwD}
{DwDnDwDw}
vllllllllV
but changed his mind in view of 34.Rh6 Kg7 35.f5 Nxc3 36.Kf4! and White wins. However,
rather than 35...Nxc3?? (or the even worse 35...Rxc3?? as given by Alekhine in the tournament
book), Black has the saving 35...Nb2!:
cuuuuuuuuC
{wDwDwDwD}
{0w0wDwiw}
{w0rDwDp$}
{DwDpDP)w}
{PDw)wDwD}
{Dw)BDKDw}
{whwDwDwD}
{DwDwDwDw}
vllllllllV
If now:
(a) 36.Ke2 Nxd3 37.Kxd3 Rd6=;
(b) 36.f6+ Rxf6 37.gxf6+ Kxh6=;
(c) 36.Bb1 Rxc3+ 37.Kg4 gxf5+ 38.Bxf5 Nc4, and Stockfish can see no more than a miniscule
advantage for White (about +0.32 out to 35 ply).
Game 2, Tal-Lisitsin, 23rd USSR Ch, Leningrad 1956: Chernev marvels at Tal’s king walk and
gives the impression that his victory was predestined from move 22 on, but this was a much
more complicated game with better chances for Black than Chernev realized. All of the
following variations escaped his attention.
That Black was not lost is seen at move 27, where instead of the text 27...Bc8-f5, much stronger
was 27...Bc8-g4!:
cuuuuuuuuC
{wDwDwDkD}
{0wDw4pDp}
{wDpDwDwD}
{DwDpDwDw}
{NDwIw0b)}
{)wDBDwhw}
{w)PDwDPD}
{DwDRDwDw}
vllllllllV
This presents White with difficult choices, e.g.:
(a) 28.Rd2?? Ne4o;
(b) 28.Rg1 Ne2+ 29.Bxe2 Rxe2 and the c-pawn cannot be saved, viz. 30.c3?? Re4+ 31.Kc5
Rxa4o, or 30.Kc3 Bf5 etc., or 30.Kd3 Re3+ 31.Kd2 Rg3 32.Ke1 f3 33.Kf1 Rxg2 34.Rxg2
fxg2+ 35.Kxg2 Bd1 etc.;
(c) 28.Rb1 Bf5 29.Nc3 (not 29.Bxf5? Nxf5+ 30.Kc5 Re2 31.Kxc6 Ne3 -2.24) 29...Be4 30.Rg1
Nf5+ 31.Kc5 Nxh4 32.Bxe4 dxe4 33.Kxc6,
cuuuuuuuuC
{wDwDwDkD}
{0wDw4pDp}
{wDKDwDwD}
{DwDwDwDw}
{wDwDp0wh}
{)wHwDwDw}
{w)PDwDPD}
{DwDwDw$w}
vllllllllV
and Black has much the more dangerous passed pawn;
(d) 28.Ra1 Bf5 29.Nc3 Be4 30.Re1 [30.Rg1 transposes to (c)] 30...Re6 31.Nxe4 dxe4 32.Bc4
Re8, when Komodo says -2.28 and Stockfish -2.93.
At move 28, where Black played 28...Re7-e6, Chernev remarks “Obviously, to go after the h-
pawn.” But this is not at all Black’s best plan.
cuuuuuuuuC
{wDwDwDkD}
{0wDw4pDp}
{wDpDwDwD}
{DwDpDbDw}
{NDwIw0w)}
{)wDBDwhw}
{w)P$wDPD}
{DwDwDwDw}
vllllllllV
Much better for Black to emulate Tal and bring his own king out, viz. 28...Kg7 29.Nc3 Kf6
30.Rf2 Bxd3 31.cxd3 Kf5 32.Kc5 Re6,
cuuuuuuuuC
{wDwDwDwD}
{0wDwDpDp}
{wDpDrDwD}
{DwIpDkDw}
{wDwDw0w)}
{)wHPDwhw}
{w)wDw$PD}
{DwDwDwDw}
vllllllllV
and Black is secure (-1.12).
At move 32,
cuuuuuuuuC
{wDwDwDkD}
{0wDwDpDp}
{wDpIwDwD}
{DwHpDwDw}
{wDwDw0w4}
{)wDPDwhw}
{w)w$wDPD}
{DwDwDwDw}
vllllllllV
it looks like Black, instead of playing 32...Rh6+, could practically force a draw now with
32...Nf5+! 33.Kxc6 Rh2!
cuuuuuuuuC
{wDwDwDkD}
{0wDwDpDp}
{wDKDwDwD}
{DwHpDnDw}
{wDwDw0wD}
{)wDPDwDw}
{w)w$wDP4}
{DwDwDwDw}
vllllllllV
34.Nb3 Defending the rook is forced; if 34.Kxd5?? f3o. 34...Ne3 35.Nd4 Rxg2 36.Rxg2+
Nxg2 37.Kxd5 h5,
cuuuuuuuuC
{wDwDwDkD}
{0wDwDpDw}
{wDwDwDwD}
{DwDKDwDp}
{wDwHw0wD}
{)wDPDwDw}
{w)wDwDnD}
{DwDwDwDw}
vllllllllV
and despite the extreme asymmetry, Stockfish calls it even.
Black’s last chance probably came at move 34. There, instead of 34...Nf5-d4, he needed to play
34...Rh6-h2:
cuuuuuuuuC
{wDwDwDkD}
{0KDwDpDp}
{wDpDwDwD}
{DwHpDnDw}
{wDwDw0wD}
{)wDPDwDw}
{w)w$wDP4}
{DwDwDwDw}
vllllllllV
Now if 35.Kxc6?? or 35.Kxa7?? f3o. Forced therefore is 35.Rf2, when the engines see best play
along the lines of 35...Kg7 36.Rxf4 Kg6 37.Rg4+ Kf6 38.Kxc6 h5:
cuuuuuuuuC
{wDwDwDwD}
{0wDwDpDw}
{wDKDwiwD}
{DwHpDnDp}
{wDwDwDRD}
{)wDPDwDw}
{w)wDwDP4}
{DwDwDwDw}
vllllllllV
Now if 39.Ra4 Rxg2 40.Nd7+ Kg5 41.Ne5 (if 41.Kxd5 h4 etc.) 41...Rc2+ 42.Kxd5 Ne7+
cuuuuuuuuC
{wDwDwDwD}
{0wDwhpDw}
{wDwDwDwD}
{DwDKHwip}
{RDwDwDwD}
{)wDPDwDw}
{w)rDwDwD}
{DwDwDwDw}
vllllllllV
43.Kd4 (not 43.Kd6?? Nc8+ 44.Kd5 Nb6+) 43...f6 44.Nf7+ (if 44.Nf3+ Kf4 45.Ng1 h4 etc.)
44...Kf4
cuuuuuuuuC
{wDwDwDwD}
{0wDwhNDw}
{wDwDw0wD}
{DwDwDwDp}
{RDwIwiwD}
{)wDPDwDw}
{w)rDwDwD}
{DwDwDwDw}
vllllllllV
45.Ra6 (of course not 45.Rxa7?? Nc6+) 45...Nf5+ 46.Kd5 Rxb2 47.Rxa7 h4 and wins.
Therefore (going back to the position after 38...h5 above) virtually forced is 39.Nd7+ Ke6 and
White has nothing better than 40.Nc5+ Kf6 41.Nd7+ etc. with a draw.
Game 3, Boleslavsky-Lisitsin, USSR Ch, Leningrad 1956: Chernev gets off to a bad start,
erroneously placing this game in Moscow. The game features one outright howler, and the notes
more, including a double howler.
We can overlook the way Black misplayed his Dragon Sicilian; in 1956 much of today’s
extensive theory was terra incognita. Chernev’s first error, one of omission, comes at Black’s
14th move:
cuuuuuuuuC
{rDw4wDkD}
{0pDwDpgp}
{wDw0bhpD}
{1NDw0wDw}
{wDwDPDwD}
{)wDwGPDw}
{w)P!wDP)}
{DKDRDBDR}
vllllllllV
Here 14...Qa4 deserved at least one ?, or better two. Correct was 14...Qxd2 15.Rxd2 d5 16.Nc7
Rac8 17.Nxe6 fxe6 when White has some advantage but Black gets counterplay. Chernev
instead goes off on a tangent on what would happen if, after 14...Qa5, White nabbed the
seemingly vulnerable d-pawn with 15.Nxd6:
cuuuuuuuuC
{rDw4wDkD}
{0pDwDpgp}
{wDwHbhpD}
{DwDw0wDw}
{qDwDPDwD}
{)wDwGPDw}
{w)P!wDP)}
{DKDRDBDR}
vllllllllV
Chernev says “this would be the consequence: 15...Ne8 16.Bc5 Nxd6 17.Bxd6 Bf8 18.Qb4
Rxd6! 19.Rxd6 Qxb4 20.axb4 Bxd6 and Black has won a piece.”
This line has two howlers. First off, instead of 15...Ne8??, necessary first is 15...a6!,
cuuuuuuuuC
{rDw4wDkD}
{DpDwDpgp}
{pDwHbhpD}
{DwDw0wDw}
{qDwDPDwD}
{)wDwGPDw}
{w)P!wDP)}
{DKDRDBDR}
vllllllllV
preventing both 16.Bb5 and a potential later Nd6-b5. Then if:
(a) 16.Qb4 Qxb4 17.axb4 Ne8 18.Bc5 Nxd6 19.Rxd6 Rxd6 20.Bxd6 Rd8 winning the bishop
since moving it allows mate;
(b) 16.Bg5 Rd7 17.Qc1 Qc6 18.Nxf7 Bxf7o;
(c) 16.Bc5 Bf8 17.Qc3 Qc6 etc.;
(d) 16.Bb6 Rd7 17.Qc3 Bf8 18.Bc5 Qc6 19.Qxe5 Rad8 20.Bb4 Bxd6 finally winning the
piece.
The problem with 14...Qa4 was not Boleslavsky’s 15.c4, praised by Chernev as a “brilliant
sacrifice” but evaluated by Komodo at just +0.50, not even among its top ten recommendations.
It is not surprising that Chernev missed the best move, but to see it missed by Lisitsin, then an
IM of about 2500 strength, and especially by Boleslavsky, then 2600+ and one of the world’s top
20, definitely is surprising. What they all missed is that instead of 15.c4, immediately winning
was 15.Nc7!:
cuuuuuuuuC
{rDw4wDkD}
{0pHwDpgp}
{wDw0bhpD}
{DwDw0wDw}
{qDwDPDwD}
{)wDwGPDw}
{w)P!wDP)}
{DKDRDBDR}
vllllllllV
Black then has no time for 15...Rac8/Rab8, because of 16.Bb5! winning his queen. Therefore
he would have to play 15...Qc6 allowing 16.Nxa8 and White wins the exchange.
After 15.c4, an interesting and probably better alternative for Black goes unmentioned. Instead
of accepting White’s sacrifice by 15...Bxc4, Black had a promising counter-sac in 15...Rac8!?:
cuuuuuuuuC
{wDr4wDkD}
{0pDwDpgp}
{wDw0bhpD}
{DNDw0wDw}
{qDPDPDwD}
{)wDwGPDw}
{w)w!wDP)}
{DKDRDBDR}
vllllllllV
Two sample variations then:
(a) 16.Nxa7 Rxc4! 17.Bxc4 Bxc4 18.Qb4 (if 18.Qc2 Ba2+! 19.Kc1 Qa6=) 18...Qxb4 19.axb4
d5,
cuuuuuuuuC
{wDw4wDkD}
{HpDwDpgp}
{wDwDwhpD}
{DwDp0wDw}
{w)bDPDwD}
{DwDwGPDw}
{w)wDwDP)}
{DKDRDwDR}
vllllllllV
and Black has considerable counterplay, e.g. 20.Kc2 d4w, or 20.exd5 Nxd5 21.Bc5 f5=.
(b) 16.Nc3 Qb3 17.Rc1 Bxc4 18.Bxc4 Rxc4 19.Bg5 Rdc8
cuuuuuuuuC
{wDrDwDkD}
{0pDwDpgp}
{wDw0whpD}
{DwDw0wGw}
{wDrDPDwD}
{)qHwDPDw}
{w)w!wDP)}
{DK$wDwDR}
vllllllllV
and if White recoups his pawn by 20.Qxd6 then 20...Rxc3! 21.Rxc3 Rxc3 22.Bxf6 Bxf6
23.Qxf6 Rc5=, or if, as in the actual game, he aims at establishing a knight outpost on d5 by
20.Bxf6 Bxf6 21.Nd5 Bd8, he does not have any comparable kingside attacking prospects
(rated -0.54 Komodo, -0.72 Stockfish).
Chernev might have noted that the text 25...g5?? was a blunder. Whatever hope remained lay in
25...Qg7 and if 26.g5 h5 27.gxh6 Qxh6, though then White is still winning.
An interesting footnote to all this: On 9 November 2018 I received an e-mail from British master
and historian Bernard Cafferty, who lives in Hastings, England. Croatian GM Bogdan Lalic is a
regular at the chess club there, and he hit upon 15.Nc7! himself independently. He told Cafferty
that the discovery shook his belief in the strength of Boleslavsky!
Game 4, Tarrasch-Thorold, Manchester 1890: Chernev calls this the most instructive rook-and-
pawn ending ever played. It may not quite deserve such a superlative, but it is certainly very well
played overall by Tarrasch, and Chernev does a good job of explaining the correct strategy.
However, significant lapses by both players pass unnoticed, and Chernev contributes several
howlers.
At move 18, where in response to 18.Ra1-c1+ Black played 18...Kc8-b8, Chernev examines the
alternative 18...Kc8-d7:
cuuuuuuuuC
{wDw4wDn4}
{0pDkDw0p}
{wDwDwDwD}
{DwDw0wDw}
{wDw0wDwD}
{DwDPDwDw}
{P)wHwDP)}
{Dw$w$wIw}
vllllllllV
We give his line, with our punctuation in red: 19.Rxe5 Kd6?? 20.Re4?! Kd5? 21.Rf4?! “and
White wins the d-pawn.” The last four moves range from dubious to dreadful. First off, after
18.Rxe5,
cuuuuuuuuC
{wDw4wDn4}
{0pDkDw0p}
{wDwDwDwD}
{DwDw$wDw}
{wDw0wDwD}
{DwDPDwDw}
{P)wHwDP)}
{Dw$wDwIw}
vllllllllV
rather than 19....Kd6??, better 19...Nf6 when White’s advantage is not great (+0.72 per
Stockfish). However, if Black is so foolish as to play 19...Kd6??, then the correct reply is not
20.Re4?! but 20.Nc4+!,
cuuuuuuuuC
{wDw4wDn4}
{0pDwDw0p}
{wDwiwDwD}
{DwDw$wDw}
{wDN0wDwD}
{DwDPDwDw}
{P)wDwDP)}
{Dw$wDwIw}
vllllllllV
giving Black a choice of three ways to lose:
(a) 20...Kd7 21.Rb5 Kc6 22.Na3+ Kd6 23.Rxb7 (+4.00);
(b) 20...Kc7 21.Rb5! (threatening 22.Ne5+ Kd6 23.Nf7+) 21...Nh6 22.Na5+ Kd7 (if
22...Kb8?? 23.Rxb7+ and mate shortly) 23.Rxb7+ etc. (+4.02);
(c) 20...Kc6 21.Na5+ Kb6 22.Nb3 a6 (if 22...Rd6 23.Rc8 Rc6 24.Rb8 Kc7 25.Ree8i)
23.Re6+ Ka7 24.Rc7 h5 25.Nc5 Rh6 (not 25...Rb8?? 26.Rxa6#) 26.Rxb7+ Ka8 27.Rxg7
(+4.64).
A potentially critical mistake, one rather hard to see, goes unnoticed at White’s 37th move:
cuuuuuuuuC
{wDwDwDwD}
{DwDkDwDw}
{p0wDw4pD}
{DwDwDwDp}
{PDw)wDwD}
{DwDKDwDw}
{w)wDRDP)}
{DwDwDwDw}
vllllllllV
Here Tarrasch needed to play 37.b3, but instead played 37.b4?. While the difference may seem
trivial, it turns out to be crucial. If after 37.b3, play then proceeds as in the game with 37...Rf1
38.Re5,
cuuuuuuuuC
{wDwDwDwD}
{DwDkDwDw}
{p0wDwDpD}
{DwDw$wDp}
{PDw)wDwD}
{DPDKDwDw}
{wDwDwDP)}
{DwDwDrDw}
vllllllllV
White need not be concerned about the black rook getting behind his queenside pawns, viz.
38...Rb1 39.Kc2, or 38...Rd1+ 39.Kc3 Rc1+ 40.Kd2 Rb1 41.Kc2, the king being able to
defend the b-pawn.
However, after 37.b4? Rf1 38.Re5, Black failed to exploit Tarrasch’s error, playing 38...Rf2?.
Instead he needed to play 38...Rd1+!:
cuuuuuuuuC
{wDwDwDwD}
{DwDkDwDw}
{p0wDwDpD}
{DwDw$wDp}
{P)w)wDwD}
{DwDKDwDw}
{wDwDwDP)}
{DwDrDwDw}
vllllllllV
If now:
(a) 39.Kc2/Ke2 Rxd4=, or
(b) 39.Kc3 Rc1+ 40.Kb2/Kd2 Rc4=, or
(c) 39.Ke3 Re1+ 40.Kf3 Ra1 41.a5 Rb1 42.axb6 Rxb4=, or
(d) 39.Ke4 (39.Kc4 Rc1+ 40.Kd5 works out the same) 39...Re1+ 40.Kd5 Rb1 41.Re6 Rxb4
42.Rxg6 Rxa4 43.Rxb6 and Stockfish sees only a miniscule advantage for White (+0.24).
In every case the vulnerability of the pawn on b4 makes the crucial difference.
But with 38...Rf2? Black missed his chance to draw. Tarrasch’s play from then on was
exemplary.
Many side variations are possible in the above lines, but it seems clear that against best play
Black cannot make his extra piece count.
A note at White’s 15th move reaches this position after 15.Nf3 Na5 16.Qb5 Qxb5 17.cxb5 Nc4
18.Bc1:
cuuuuuuuuC
{rDwDw4kD}
{0wDwDp0p}
{w0w0phwD}
{DP0wDwDw}
{wDn)wDwD}
{Dw)wDN)w}
{PDwDP)w)}
{$wGwDRIw}
vllllllllV
Here Chernev gives 18...Nd5 “and the c-pawn falls.” But then the simple 19.Nd2 forces
19...Nxd2 20.Bxd2 and the pawn is safe. Correct is 18...Ne4!, which prevents 19.Nd2 and wins
the c-pawn and probably more.
Only one minor improvement to be noted here. The note at move 27 says that in the event of
27...Nh5,
cuuuuuuuuC
{rDwDrDwi}
{DwDbDwDw}
{wDpDp)w0}
{0pDpDwDn}
{wDw)wDw)}
{Dw)BDwDw}
{PDwINDwD}
{DwDwDR$w}
vllllllllV
White wins by 28.Bg6. True, this is a good, winning move, but even stronger is 28.f7!,
cuuuuuuuuC
{rDwDrDwi}
{DwDbDPDw}
{wDpDpDw0}
{0pDpDwDn}
{wDw)wDw)}
{Dw)BDwDw}
{PDwINDwD}
{DwDwDR$w}
vllllllllV
when if 28...Re8-f8/d8/c8/b8 29.Rg6 forces mate in 11 at most, therefore forced is 28...Rg8
29.fxg8Q+ Rxg8 30.Rf7 Rxg1 31.Nxg1 (+10.03).
First a factual correction. Chernev claims that Alekhine’s Defense, 1.e4 Nf6, “was first played in
serious tournament chess in 1921, when Alekhine beat Steiner with it at Budapest.” ChessBase
however shows several earlier games: Hanham-Delmar, New York 1893; Walbrodt-W. Cohn,
Berlin 1897; and Nimzovich-Albin, Vienna 1905. All of these were serious master-level
tournaments.
As for the game, Chernev’s notes have some major errors of omission, including a failure to
identify the actual losing move.
Interestingly, even in the actual game continuation, where Black played 20...Rae8 instead of
20...Bh8,
cuuuuuuuuC
{wDwDr4kD}
{0pDwDwgp}
{wDw1w0pD}
{Dw0PDbDN}
{wDPDwDwD}
{DPDwDwDw}
{PDwGwDP)}
{$wDQDRIw}
vllllllllV
Komodo likes 21.Rxf5 (+0.96) better than Fischer’s 21.Nxg7 (+0.62).
At move 29 Chernev says “Things go at a rapid pace from now on. Every one of Fischer’s moves
is a hammer blow!” This is yet another instance where Chernev’s all-too-frequent and all-too-
fervent hero worship befogs his critical faculties, so that he misses a critical defensive resource
Black had.
At move 32,
cuuuuuuuuC
{wDwDrDkD}
{0pDwDwDp}
{wDw)w0pD}
{Dw0w1wDw}
{wDPDbGPD}
{DPDwDwDP}
{PDwDw!wD}
{DwDRDwIw}
vllllllllV
Black played 32...Qc3??, perhaps hoping for 33.Qxc5?? Qxh3o. Chernev makes no comment,
but this was in fact the losing move, as Fischer proved with 33.d7!. However, Black could have
saved himself with 32...Qe6!:
cuuuuuuuuC
{wDwDrDkD}
{0pDwDwDp}
{wDw)q0pD}
{Dw0wDwDw}
{wDPDbGPD}
{DPDwDwDP}
{PDwDw!wD}
{DwDRDwIw}
vllllllllV
White’s only try for advantage then is 33.Qxc5 Bc6,
cuuuuuuuuC
{wDwDrDkD}
{0pDwDwDp}
{wDb)q0pD}
{Dw!wDwDw}
{wDPDwGPD}
{DPDwDwDP}
{PDwDwDwD}
{DwDRDwIw}
vllllllllV
when the two main lines are:
(a) 34.Qf2 h5!= (if 35.gxh5?? Qxh3o). Another example of the equalizing effect of the
opposite-color bishops, a factor Chernev never mentioned.
(b) 34.Qe3 Qxe3+ 35.Bxe3 Rxe3 36.d7 Bxd7 37.Rxd7 Rxh3
cuuuuuuuuC
{wDwDwDkD}
{0pDRDwDp}
{wDwDw0pD}
{DwDwDwDw}
{wDPDwDPD}
{DPDwDwDr}
{PDwDwDwD}
{DwDwDwIw}
vllllllllV
and either 38.Rxb7 a5 39.Ra7 h5 40.gxh5 Rxh5, or 38.Kg2 Rc3 39.Rxb7 a5 40.Ra7 Rc2+
41.Kf3 Rxa2, Stockfish seeing both as likely draws (about +0.35 at 36 ply). Even if White can
squeeze out a win in a rook ending, this was far preferable for Black to the immediate crash-and-
burn of 32...Qc3??.
Game 10, Smyslov-Reshevsky, The Hague-Moscow WCh, 1948: Some superficial notes here
bear mentioning.
At move nine,
cuuuuuuuuC
{rDw1kgw4}
{Dp0bDp0w}
{pDn0wDn0}
{DwDw0wDw}
{wDw)PDwD}
{DB)wDNDw}
{P)wHw)P)}
{$wGQIwDR}
vllllllllV
where Smyslov played 9.Nc4, Chernev says “A mechanical move, such as 9.0–0 would allow
the reply 9...Nf4 followed by 10...g5, and Black has seized the initiative.” Nonsense. If 9.0-0
Nf4?!, then 10.Nc4 simply forces the knight back to g6,
cuuuuuuuuC
{rDw1kgw4}
{Dp0bDp0w}
{pDn0wDw0}
{DwDw0wDw}
{wDN)PhwD}
{DB)wDNDw}
{P)wDw)P)}
{$wGQDRIw}
vllllllllV
since if 10...g5? 11.dxe5 etc. winning the e-pawn.
Variation (2) in the note at Black’s 24th move comes to a superficial, incorrect conclusion.
cuuuuuuuuC
{rDwDwDwi}
{DpDw1p0w}
{pGn0nDw0}
{DwDw0wDw}
{wDwDPDQD}
{DB)wDw)w}
{P)wDw)w)}
{DwDRDwIw}
vllllllllV
After 24...Ncd8 25.Rd2 f6 26.Bxe6 Nxe6 27.Qd1 Chernev says “and the pawn falls.”
cuuuuuuuuC
{rDwDwDwi}
{DpDw1w0w}
{pGw0n0w0}
{DwDw0wDw}
{wDwDPDwD}
{Dw)wDw)w}
{P)w$w)w)}
{DwDQDwIw}
vllllllllV
Really? After 27...Ng5! White must either pause for 28.f3 Nf7, in which case the pawn does not
fall, or play 28.Rxd6 Nxe4, in which case one of White’s pawns falls too.
And anyway, White can do much better in this line by varying at move 26.
cuuuuuuuuC
{rDwhwDwi}
{DpDw1w0w}
{pGw0n0w0}
{DwDw0wDw}
{wDwDPDQD}
{DB)wDw)w}
{P)w$w)w)}
{DwDwDwIw}
vllllllllV
Instead of 26.Bxe6?!, it is much better to keep both bishops with 26.Qd1 Nf7 27.h4, and the
black knights (or for that matter also the queen and rook) have hardly any scope (+1.80 per
Komodo).
The note at move 15 is a travesty. In the variation beginning 15...Ke6 (actually Black’s best),
cuuuuuuuuC
{rDwDwDw4}
{0p0wDpgp}
{wDwDkDwD}
{DwDwDwDw}
{wGwhwDwD}
{DwDwDPDw}
{P)wDw)w)}
{DwIRDBDR}
vllllllllV
Chernev overestimates White’s advantage and fails to see the best line, which is 16.Bc4+ Kf5
17.Bd3+ Kf6 18.Bc3 Rad8, with a definite though not huge plus (+0.86). Instead he gives only
16.Bh3+, and in reply he considers only two dreadful moves, 16...f5?? and 16...Nf5??. Correct is
the implausible-looking but sound 16...Kd5!,
cuuuuuuuuC
{rDwDwDw4}
{0p0wDpgp}
{wDwDwDwD}
{DwDkDwDw}
{wGwhwDwD}
{DwDwDPDB}
{P)wDw)w)}
{DwIRDwDR}
vllllllllV
when the best White has is 17.f4 c5 18.Bg2+ Kd6 19.Bc3 Rab8 and his advantage is
microscopic (+0.25).
Game 12, Reshevsky-Najdorf, Dallas 1957: Chernev’s notes have Black making too many
reflex recaptures and blunders.
After 17.Nb5-d6, Chernev illustrates what he believes White’s main threat is,
cuuuuuuuuC
{rDw1w4kD}
{DwDwDp0p}
{nDwHpDwD}
{0w)pDwDw}
{w)wDwDwD}
{)wDwDwDw}
{wDw!w)P)}
{$wDwDRIw}
vllllllllV
giving 18.Nb7 Qc7 19.Nxa5 Nxc5 20.Rfc1 Ne4 21.Rxc7 Nxd2 22.f3 Rfc8 23.Rac1 Rxc7
24.Rxc7 Nb1 25.b5 h6 26.b6 Rxa5 27.b7 Rb5 28.Rc8+ “and White wins.” But 24...Nb1?? is a
blunder. Instead Black can play 24...g5,
cuuuuuuuuC
{rDwDwDkD}
{Dw$wDpDp}
{wDwDpDwD}
{HwDpDw0w}
{w)wDwDwD}
{)wDwDPDw}
{wDwhwDP)}
{DwDwDwIw}
vllllllllV
both giving his king Luft and intending ...g5-g4. White cannot then play 25.b5?? Rxa5. Instead
25.a4 Nb1 (now playable) leaves White with a considerable advantage (+1.57), but nothing like
what Chernev’s blunder allows.
The note at move 18 has both Black and White playing very bad moves.
cuuuuuuuuC
{rDwDw4kD}
{DwDqDp0p}
{nDwHpDwD}
{0w)pDwDw}
{w)wDw)wD}
{)wDwDwDw}
{wDw!wDP)}
{$wDwDRIw}
vllllllllV
Chernev gives 18...g6 19.f5 exf520.Rxf5 gxf5 21.Qg5+ Kh8 22.Qf6+ Kg8 23.Nxf5 “and Black
must give up his queen to prevent mate.” This is very sloppy. In the first place, Chernev fails to
mention that best for Black is 18...axb4 19.axb4 Nc7 intending ...Nb5 or ...Ne8 to trade off the
strongly posted knight.
Secondly, after 18...g6 19.f5 exf5, Chernev has White playing not the correct 20.Nxf5 but
20.Rxf5?, a serious mistake that throws away his winning chances.
cuuuuuuuuC
{rDwDw4kD}
{DwDqDpDp}
{nDwHwDpD}
{0w)pDRDw}
{w)wDwDwD}
{)wDwDwDw}
{wDw!wDP)}
{$wDwDwIw}
vllllllllV
Black replies not with the reflex recapture 20...gxf5?? but 20...axb4! 21.axb4 Nxc5! 22.Rxa8 (of
course not 22.bxc5?? Rxa1+) 22...Rxa8 23.bxc5 gxf5 24.Qg5+ Kf8:
cuuuuuuuuC
{rDwDwiwD}
{DwDqDpDp}
{wDwHwDwD}
{Dw)pDp!w}
{wDwDwDwD}
{DwDwDwDw}
{wDwDwDP)}
{DwDwDwIw}
vllllllllV
This shows how having this additional escape square (instead of only h8) makes the crucial
difference. Now the best White can do is draw by perpetual check.
Game 13, Porges-Lasker, Nuremberg 1896: Again Chernev’s notes show poor tactical
awareness, and have the loser making needless blunders.
At move 13,
cuuuuuuuuC
{rDb1rDkD}
{0nDwgp0p}
{wDp0wDwD}
{DwDwDwDw}
{wDwDwDwD}
{DPDwDNDw}
{PGPHQ)P)}
{$wDwDRIw}
vllllllllV
while Chernev is correct that the text move 13.Rfe1 was best, his claim that 13.Qd3 Nc5
14.Qc4 “loses the exchange after 14...Ba6” is wrong.
cuuuuuuuuC
{rDw1rDkD}
{0wDwgp0p}
{bDp0wDwD}
{DwhwDwDw}
{wDQDwDwD}
{DPDwDNDw}
{PGPHw)P)}
{$wDwDRIw}
vllllllllV
White plays 15.Qg4 (threatening mate) and 16.Rfe1, avoiding any material loss. It’s strange that
Chernev would miss this, since it’s clear he was aware of the potential mate threat in the sub-
variation he gives, 13.Qd3 Nc5 14.Qc3 Bf6 (preventing mate) 15.Nd4 Qb6, when he says
“Black threatens 16...Ne6.” But White can say “So what?” and play 16.Nc4 forcing 16...Qc7
and White is OK.
The note at move 18 can be improved at the end. After 18.Qxb7 e4 19.Qa6 Bb2 20.Rab1,
cuuuuuuuuC
{rDw1rDkD}
{0wDbDw0p}
{QDpDwDwD}
{DwDpDwDw}
{wDwDpDwD}
{DPDwDNDw}
{PgPHw)P)}
{DRDw$wIw}
vllllllllV
Chernev gives 20...Bc3, which is good, but better is simply 20...exf3 21.Rxe8+ Bxe8
cuuuuuuuuC
{rDw1bDkD}
{0wDwDw0p}
{QDpDwDwD}
{DwDpDwDw}
{wDwDwDwD}
{DPDwDpDw}
{PgPHw)P)}
{DRDwDwIw}
vllllllllV
with a much bigger plus, since if 22.Rxb2?? Qg5 22.Nxf3 Qc1+ (-6.30).
At Black’s 20th move, Chernev gives the variation 20...Be6 21.Nc5 Rac8 22.Qe3 Qg6,
cuuuuuuuuC
{wDrDw4wi}
{0pDwDnDp}
{wDwDbDqD}
{DwHp)p0w}
{wDwDw)wD}
{DPDw!w)w}
{PDwDwDB)}
{$wDwDRIw}
vllllllllV
concluding with 23.Rad1 “and White wins a pawn.” Why not simply the obvious 23.Nxb7,
taking a pawn already en prise (+1.78)? Besides, 23.Rad1? gets White in trouble after 23...gxf4
24.gxf4 Rg8 (threatening mate) 25.Rf2 d4 26.Qxd4 (not 26.Rxd4? Rxc5) 26...Rgd8,
cuuuuuuuuC
{wDr4wDwi}
{0pDwDnDp}
{wDwDbDqD}
{DwHw)pDw}
{wDw!w)wD}
{DPDwDwDw}
{PDwDw$B)}
{DwDRDwIw}
vllllllllV
and while White has won the d-pawn he loses either the knight (27.Qa1 Rxc5) or his queen for
rook and bishop, 27.Nxe6 Rxd4 28.Nxd4, and his advantage has evaporated (+0.19).
Game 18, Lasker-Tarrasch, St. Petersburg 1914: A much more complicated game than Chernev
thought, with many tactical finesses that go unmentioned.
Game 19, Kupferstich-Andreassen, Denmark, 1953: In such a wild game as this, it is not
surprising that Chernev missed some tactical shots.
White in turn blew it at move 15. Just as Black at move 13 need not have worried about his
hanging rook, so White could have disregarded his imperiled knight.
cuuuuuuuuC
{rDwDkDrD}
{0p0qDpDp}
{wDn0wGpD}
{DNDw0wDw}
{bDw)wDwD}
{DwDwDQDw}
{P)PDw)P)}
{$wDwIwHR}
vllllllllV
Instead of 15.Na3?, he could have pressed on boldly with 15.dxe5! Bxb5 16.0–0–0 Qe6 17.exd6
cxd6 18.Qf4! forcing 18...g5 19.Bxg5:
cuuuuuuuuC
{rDwDkDrD}
{0pDwDpDp}
{wDn0qDwD}
{DbDwDwGw}
{wDwDw!wD}
{DwDwDwDw}
{P)PDw)P)}
{DwIRDwHR}
vllllllllV
If now (a) 19...Qxa2 20.Re1+ Kd7 21.Qf5+ Kc7 22.Qxb5 (+2.23); (b) 19...Ne5 20.Nf3 f6
21.Qb4 0–0–0 22.Nd4 Qxa2 23.Nxb5 fxg5 24.Nxd6+ Rxd6 25.Qxd6 (+2.20); or (c) 19...Qe5
20.Nf3 Qxf4+ 21.Bxf4 (+2.01).
After making no comment on moves 10-16, Chernev finally says something at Black’s 17th.
cuuuuuuuuC
{rDwDkDrD}
{0p0wDpDp}
{wDw0wGpD}
{DwDw0wDw}
{bDwhwDwD}
{HwDwDwDN}
{P)PDw)P)}
{$wDwIwDR}
vllllllllV
Of 17. Nxc2+ he says “Black tries to collect a few pawns for the piece he has lost.” An accurate
comment would be “A terrible blunder that immediately gives Black a lost game.” Correct was
17...Bc6!, when Komodo sees best play as something like 18.Rg1 g5 19.Bxg5 (if 19.Nxg5?
Rg6) 19...Ne6 20.Be3 Rxg2 21.Rxg2 Bxg2 22.Ng5 Nxg5 23.Bxg5 Kd7:
cuuuuuuuuC
{rDwDwDwD}
{0p0kDpDp}
{wDw0wDwD}
{DwDw0wGw}
{wDwDwDwD}
{HwDwDwDw}
{P)PDw)b)}
{$wDwIwDw}
vllllllllV
White is up a knight for two pawns, but Komodo thinks Black has decent drawing chances
(+0.46). In any event, this better than 17...Nxc2+??, which suicidally opens the c-file for White.
White could have made use of that file sooner than he did. At move 20,
cuuuuuuuuC
{rDwDkDrD}
{0p0wDpDp}
{wDw0wGpD}
{DwDw0wDw}
{wDwDbDwD}
{DwDwDwDN}
{P)wDw)P)}
{Dw$wIwDR}
vllllllllV
much better than the text move 20.Ng5, which Chernev erroneously praises as gaining a tempo,
was 20.Rxc7!. If then 20...Bxg2?? 21.Ng5! Bxh1 22.Re7+ Kd8 23.Nxf7+ Kc8 24.Nxd6+ Kb8
25.Bxe5 etc. much like the actual game. Relatively best is 20...h6, but then after 21.f3 Bc6
22.Nf2 White is clearly winning (+3.57).
The problem with 20.Ng5 is that Black, instead of getting fatally greedy with 20...Bxg2??, could
have sealed off the c-file with 20...Bc6!:
cuuuuuuuuC
{rDwDkDrD}
{0p0wDpDp}
{wDb0wGpD}
{DwDw0wHw}
{wDwDwDwD}
{DwDwDwDw}
{P)wDw)P)}
{Dw$wIwDR}
vllllllllV
White then still stands better (about +1.45), but does not have a clear win, as he does after
20...Bxg2??.
At Black’s 25th move, Chernev is correct that the text move 25...Rd5 is better than nabbing a
pawn by 25...Qf7xa2:
cuuuuuuuuC
{rDw4wDkD}
{0bDwDwgw}
{w0wDw0wD}
{Dw0wDP0p}
{wDwDwDwD}
{Dw)wGw!w}
{q)wDwHP)}
{DwDR$wIw}
vllllllllV
However, he gives an incorrect refutation, 26.Qc7 Rxd1 27.Rxd1 Bd5 28.Qd7 “and White
might become obstreperous.”
cuuuuuuuuC
{rDwDwDkD}
{0wDQDwgw}
{w0wDw0wD}
{Dw0bDP0p}
{wDwDwDwD}
{Dw)wGwDw}
{q)wDwHP)}
{DwDRDwIw}
vllllllllV
But in that case comes 28...Bb3 29.Qc6 Rf8 30.Re1 (if 30.Rd7 Ba4) 30...Qxb2 and the
obstreperous one is Black. The truly obstreperous move is 26.h4!,
cuuuuuuuuC
{rDw4wDkD}
{0bDwDwgw}
{w0wDw0wD}
{Dw0wDP0p}
{wDwDwDw)}
{Dw)wGw!w}
{q)wDwHPD}
{DwDR$wIw}
vllllllllV
when White gets counterplay, e.g. 26...gxh4 27.Qxh4 Qf7 28.Bf4=, or 26...g4 27.Qc7 Bd5
28.Ne4 Qa4 (or 28...Rac8 29.Qe7 Rf8 30.Nd6y) 29.Ng3=.
Game 21, Tartakower-Domenech, Sitges 1934: Quite a few tactical oversights here, by both
Chernev and the players.
Chernev makes no comment next move, where stronger than the text 17.a4 was 17.Qh3,
cuuuuuuuuC
{rDwDw4kD}
{0p1wDp0p}
{wgnDwDwD}
{DwDw0wDw}
{w)wDBDwD}
{Dw)wDwDQ}
{PDwDw)P)}
{$wGRDwIw}
vllllllllV
forcing 17...g6 18.Bg5
cuuuuuuuuC
{rDwDw4kD}
{0p1wDpDp}
{wgnDwDpD}
{DwDw0wGw}
{w)wDBDwD}
{Dw)wDwDQ}
{PDwDw)P)}
{$wDRDwIw}
vllllllllV
and Black will have to give up the exchange by 18...Rd8 Bxd8 to avoid disaster by Bg5-f6 and
Qh3-h6, since if 18...f6 19.Rd7i, or 18...f5 19.Bd5+ Kh8 20.Be6 Nb8 21.Qh4 Qg7 22.Rd5
h5 23.Rad1 etc. (+4.13).
The problem with 25...Bb8??, which neither Tartakower, Domenech nor Chernev realized, is
26.Bxf7+!!:
cuuuuuuuuC
{wgw4wDkD}
{DpDw1BDp}
{pDnDwDpG}
{)wDwDwDw}
{w)wDp!w)}
{Dw)wDwDw}
{wDwDw)PD}
{DwDRDwIw}
vllllllllV
This wins in all variations, viz.
(a)26...Qxf7 27.Rxd8+ Nxd8 28.Qxb8i;
(b) 26...Kh8 27.Rxd8+ Qxd8 28.Qg4 (not 28.Qd2 Qxh4) and:
cuuuuuuuuC
{wgw1wDwi}
{DpDwDBDp}
{pDnDwDpG}
{)wDwDwDw}
{w)wDpDQ)}
{Dw)wDwDw}
{wDwDw)PD}
{DwDwDwIw}
vllllllllV
(b1) 28...Be5 29.Bg5 h5 30.Qe2 Qd6 31.Bb3 (+3.65);
(b2) 28...Ne5 29.Bg5! Nxg4 30.Bxd8 Be5 31.Be6 (+3.54).
This game was effectively decided at move 15. Chernev botches that crucial point. Black had
been preparing the advance ...f7-f5 for several moves, and now was the time to do it.
cuuuuuuuuC
{rDb1w4kD}
{0phwgphp}
{wDw0wDpG}
{DwDP0wDw}
{wDwDPDPD}
{DwHBDNDP}
{P)QDw)wD}
{$wDwDRIw}
vllllllllV
Chernev claimed it was unfeasible, giving 15...f5 16.Bxg7 Kxg7 17.gxf5 gxf5 18.exf5,
cuuuuuuuuC
{rDb1w4wD}
{0phwgwip}
{wDw0wDwD}
{DwDP0PDw}
{wDwDwDwD}
{DwHBDNDP}
{P)QDw)wD}
{$wDwDRIw}
vllllllllV
“and White has won a pawn.” Briefly, yes, but he can’t hold it, viz. 18...Na6 and after either
19.Kh2 Nb4 20.Qd1 Nxd3 21.Qxd3 Bxf5, or 19.Qe2 Nc5 20.Bb1 Bxf5, or 19.a3 Nc5 etc.,
Black gets it back with a freed-up position.
Rather than Chernev’s line, there were two ways White might try for advantage after 15...f5:
cuuuuuuuuC
{rDb1w4kD}
{0phwgwhp}
{wDw0wDpG}
{DwDP0pDw}
{wDwDPDPD}
{DwHBDNDP}
{P)QDw)wD}
{$wDwDRIw}
vllllllllV
(a) 16.gxf5 gxf5 17.Kh2 [if 17.exf5 Qe8! (or 17...Bxf5 18.Bxf5 Rxf5 is quite playable too)
18.Ne4 Rxf5 19.Ng3 Rxf3 20.Bxh7+ Kh8 21.Be4 Rf7 22.Qxc7 Bxh3=] 17...Rf6 18.Bxg7
Kxg7 19.exf5 Na6 (intending 20...Nb4) 20.Qe2 Nc5 21.Bc2 Bxf5=.
(b) Probably better is 16.exf5 gxf5 17.Kh1 and now not 17...e4?! 18.Nxe4 fxe4 19.Bxe4 when
White has good attacking prospects against the black king’s suddenly scanty protection, but
17...Rf6,
cuuuuuuuuC
{rDb1wDkD}
{0phwgwhp}
{wDw0w4wG}
{DwDP0pDw}
{wDwDwDPD}
{DwHBDNDP}
{P)QDw)wD}
{$wDwDRDK}
vllllllllV
when Black has a difficult position but better prospects than he had in the actual game (about
+0.55 per Stockfish, +0.70 Komodo, as opposed to +1.14 after the text move 15...Nce8). The
position seems to resist definitive analysis. That Black already is having such a hard time is due
to his choice of a poor opening, the Hungarian Defense (1.e4 e5 2.Nf3 Nc6 3.Bc4 Be7).