The Kibitzer by Tim Harding

"He could have been a contender..."

 I VISITED RIGA, home town of Mikhail Tal, in September for
the annual Congress of ICCF, the International Correspondence
Chess Federation. While I was there I learned a lot about some of
the great Latvian players who helped to build up the tradition from
which Tal benefitted. I was especially pleased to receive as a
present the handsome little book "Pari Savam Laikmetam" (by V.
Kirilovs, published by Sahs of Riga in 1994). This deals with the
tragically short career of Hermanis K. Matisons, the first FIDE
Champion who died of tuberculosis in 1932, shortly before his
38th birthday.

You may find him mentioned in books as "Mattison", which could
make you think he was an American, but "Matisons" is correct; all
Latvian masculine names end in "s". His birthdate was December
28, 1894 - in fact December 16 in the old calendar used in the
Tsarist empire at that time. He died on November 16, 1932.

A real artist of the chessboard, in different circumstances, Matisons
might have been a contender for the world championship had
circumstances enabled him to play in his late teens and early
twenties. His historical Elo is reckoned to be about 2510 and
nowadays he would at least be an International Master as a player
and probably a Grandmaster of Endgame Composition. It is for his
endgame studies that he is now principally remembered. It is less
well known that he also composed mate problems - some before
the First World War and another batch in the 1920s.

Matisons learned chess at the age of 7 and began composing in his
teens. As a player, he first made his mark in 1913 when he met
Capablanca in a simultaneous and beat him with a fine piece of
calculation in a king and pawn ending.

Capablanca-Matisons Riga, 1913 1 e4 e5 2 Nf3 Nc6 3 Bb5 Nf6 4 0
0 Be7 5 Nc3 d6 6 d4 exd4 7 Nxd4 Bd7 8 Nde2 0 0 9 Ng3 Re8 10
Re1 Bf8 11 Bg5 h6 12 Bxf6 Qxf6 13 Nd5 Qd8 14 Qh5 g6 15 Qd1
Bg7 16 c3 Ne5 17 Bxd7 Nxd7 18 f4 c6 19 Ne3 Nc5 20 Nc4 d5 21
exd5 Rxe1+ 22 Qxe1 cxd5 23 Ne5 Qb6 24 Qf2 Re8 25 Nf3 Qb5
26 Re1 Nd3 27 Rxe8+ Qxe8 28 Qd2 Qb5 29 b4 Qb6+ 30 Kf1 Qb5
31 Ne2 a5 32 bxa5? Qb1+ 33 Ne1 Nxe1 34 Qxe1 Qxa2 35 Nd4
Bxd4 36 cxd4 Qc4+ 37 Kf2 Qxd4+ 38 Qe3 Qb2+ 39 Kg3 d4 40
Qe8+ Kg7 41 Qe5+ Kh7 42 f5 Qb3+ 43 Kg4 Qd1+ 44 Kh4 Qh5+
45 Kg3 d3 46 Qd5 Qxf5 47 Qxf5 gxf5 48 Kf2 (See Diagram)

Anyone who has played or watched a simul knows that in such
circumstances it is usually the grandmaster who outplays the
amateur but Matisons wasn't fazed. He played 48...f4! and after 49
Ke1 Kg6 50 Kd2 Kf5 51 Kxd3 Kg4 52 Kc4 h5 53 Kb5 h4 54 Kb6
h3 55 gxh3+ Kxh3 56 Kxb7 f3 57 a6 f2 58 a7 f1Q White resigned.
If 59 a8Q Qf3+ 60 Kb8 Qxa8+ 61 Kxa8 f5-+ and the f-pawn runs.

Matisons was not yet 19 and had already published some problems
and studies. 

In his great book about endgame studies, "Test Tube Chess", A.J.
Roycroft groups Matisons with Rti, saying that both these
masters' "positions are at least as simple to look at as those of the
classics [of Troitzky etc.], but they contain in addition a depth of
thought amounting to a philosophy. They are the foreunners of the
strategic master composers Liburkin, Bron and Kasparyan."

Here is one of his early efforts, composed when he was 16 years
old. White to play and mate in two. Published in the "Rigasche
Rundschau" 23 July 1911. (See Diagram)

White: Kh1, Qe3, Bf3, Bg5, Nd6; pawns - c5, e4 Black: Ke5, Ra4,
Rc7, Na2, Ng8; pawns - c3, e6, h5

Solution at the end of the column.

The year after he beat Capablanca, World War I began and the
Tsarist Empire went into upheaval. Eventually Latvia became an
independent nation, for two brief decades, but it was 1924 before
Matisons appeared on the world chess stage as a player. What
could have been his golden decade was lost.

In the few years remaining to him, he certainly made an impact.
Chess players from many countries assembled in Paris in July 1924
and FIDE was formed. There the first (unofficial) olympiad was
held, in the form of a tournament for individuals with team prizes.
Players from the same country might have to meet each other.

The tournament director was Alekhine. The winner was to be
recognised as the World Amateur Champion and the winner was
Matisons! Moreover, another Latvian was runner-up but there was
no Latvian team as such. At this time, I learned, there was just a
Riga chess club but when he went home Matisons set about
transforming it into a national association so that Latvia could join
FIDE

For a good account of the Paris event, you can obtain the recently
published pamphlet entitled "The First Amateur Chess Olympiad"
by Val Zemitis, which is Supplement 22 to the new "Latvian
Correspondence Chess and Latvian Gambit" magazine (details
about this at the end of the article). You can also seek out Arpad
Foldeak's book on Chess Olympiads, which has the crosstable of
the final.

The 54 Paris entrants were arranged into nine preliminary and final
sections. Zemitis writes: "The playing conditions were not ideal;
the weather was sultry and often players had to play not only two
games in one day but also in the evenings had to finish their
adjourned games." Several of the favourites such as George
Koltanowski and the Czech master Hromadka (one of the chief
pioneers of the Modern Benoni) failed to reach the final. Only the
section winners qualified. 

"In the final," writes Zemitis, "Matisons exhibited the steadiest
chess" and he won the gold medal with +4 =3 1, the loss being to
Euwe, while the draws were with Colle, Vajda and his countryman
Fricis Apsenieks (sometimes known as Franz Apscheneek) who
finished second and beat Euwe.

Matisons' win with Black against Hromadka is the game usually
quoted in this context, so I will publish a different one. Matisons
was hard on the tail-enders and the following was his last round
game.

Matisons-Kornel Havasi (Hungary) Paris ol final (8), 1924 1 d4
Nf6 2 Nf3 d5 3 c4 c6 4 cxd5 cxd5 5 Nc3 Nc6 6 Bf4 e6 7 e3 Bd6 8
Bg3 0 0 9 Bd3 Re8 10 Rc1 a6 11 0 0 Qe7 12 Bh4 Bd7 13 Bb1 h6
14 Qd3 Qd8 15 a3 Kf8 16 e4 White breaks open the centre with
advantage. 16...dxe4 17 Nxe4 Be7 18 Bxf6 gxf6 19 Ng3 f5 (See
Diagram)

To prevent Qh7, but now the pressure is intensified in the centre.
20 d5 Na7 21 Rfd1 Bb5 22 Qb3 Now Black's queen is in a very
awkward position vis-a-vis the white rooks, and cannot protect the
weaknesses at e6 and b7. 22...Bd6 23 dxe6 fxe6 24 Ba2 Rc8 25
Rxc8 Nxc8 26 Nd4 Bd7 27 Ndxf5! A simple exploitation of the
light square weaknesses; Black could have spared himself the rest.
27...Qc7 28 Nxh6 Bc6 29 Qc4 Qg7 30 Qh4 Bc7 31 Re1 Bd7 32
Ng4 Bd8 33 Qh5 Nd6 34 Ne5 Nf5 35 Ng6+ 1 0

In the field of endgame composition, Matisons' understanding of
chess reached its highest level. Hartston (in Golombek's
"Encyclopaedia of Chess") describes him as the "composer of more
than sixty published studies of great artistic merit". I am not sure if
that figure is correct; there are 53 endgame studies by him in the
Kirilovs book so maybe Hartston's figure includes the 12 mate
problems also.

Take a look at the diagram on the book cover. This is study 42 on
page 30 of the book and shared first prize in a "Shakhmatny
Listok" competition in 1927. I'll talk you through it. (See Diagram)

White: Kd3, Rd5 Black: Kb1, Na2, Bh2; pawn - e2 White to play
and draw!

The black pawn threatens to queen and is indirectly protected by a
knight fork - 1 Kxe2? Nc3+; checking on b5 first does not
eliminate the fork. The bishop prevents Re5, so it is not hard to see
that White's first move must be 1 Kd2.

Now 1...Bg3? is easy for White as 2 Kxe2 Nc3+ 3 Kf3 Nxd5 4
Kxg3 leaves Black with insufficient mating material. The cleverer
1...Bf4+ 2 Kxe2 Nc3+ is also too simplistic, met by 3 Kf3 Nxd5
when temporarily the N protects the B but 4 Ke4 and Black loses
one of his pieces. This is the first glimmering of a solution, but of
course Black can do better and the real point of the study is quite
different.

1...Nc1! The start of the main line. 2 Rb5+ Since Black threatens
2...Bf4+ 3 Ke1 Bg3+ and e1Q there is evidently no choice. 2...Ka2
3 Ra5+ Of course not 3 Kxc1?? e1Q+. 3...Kb3 Or if 3...Kb2 4 Ra4!
(4 Rf5? Nb3+ 5 Kxe2 Nd4+; 4 Rb5+ Ka3 5 Ra5+ Kb4) 4...Bc7 (or
4...Bg3 ) 5 Rb4+ and 6 Re4 to eliminate the pawn. 4 Rf5 Bd6 If
4...Bg3 5 Rf3+ and 6 Rxg3; this is why White wanted the King on
the third rank. Instead 4...Bc7 5 Rf3+ Kb2 6 Rf2 Ba5+ 7 Ke3 leads
to the same finish as in the main variation b) below. 5 Rf3+ Now
there are two principal variations, with a nice symmetry about
them:

a) 5...Ka4 (5...Kc4 ends the same way.) 6 Re3 (6 Rf2?? Bb4+)
6...Bf4 (6...Bb4+ 7 Kxc1) and now 7 Ke1!! again threatens an
immediate draw by Rxe2 that can only be prevented by 7...Bxe3
stalemate.

b) 5...Kb2 6 Rf2 The rook attacks the pawn from this direction
because the black K is on the same rank. Of course not 6 Re3??
Bb4+. 6...Bb4+ 7 Ke3 The e-pawn is pinned. 7...Bc5+ 8 Kd2
Again threatening Rxe2. Black checks lead nowhere and 8...Bxf2
is stalemate!

Stalemate featured in some other famous studies by Matisons and
in about 60% of his studies the task was to draw.

Now for a win study that you might like to try to solve for
yourself. Remember that with studies, as with longer problems, it
is not sufficient to spot the key move; you are expected to find the
variations that beat the principal defence or defences, and to
identify why plausible tries fail. Roycroft wrote in his book,
"Without pawns, and about to lose a bishop, White seems to have
nothing better than 1 Rxd3 but he has an extraordinary resource."
(See Diagram)

White: Kh3, Be6, Bh6, Rd7 Black: Kh5, Re5; pawn - d3

White to play and win. Solution at the end of this column.

Although primarily a study composer, Matisons produced a few
more mate problems also. This one was published in 1925. You
can see the OTB master in the setting - it looks as if it is a Sicilian
Defence in which White has sacrificed his queen for a king hunt
with Black desperately hoping that the f-pawn will provide
counterplay. (See Diagram)

White: Kc1, Rf1, Rd4, Bh1, Ne3, Ne4; pawns - b2, b4, c2, d2
Black: Ke2, Qc7, Rf7, Nf8; pawns - b5, e6, f2

White to play and mate in four; solution at the end of this column.

Now we return to Matisons' career as a player. He played a master
tournament at Debrecen, Hungary, in 1925 but found the step up to
the professionals hard; the field included Gruenfeld, Tartakower
and Kmoch. He only scored about 30%, even losing to Havasi this
time, but he did beat Johner, Przepiorka and Seitz. The experience
he had gained showed the following year in a stronger field at
Budapest, when, after a poor start he finished strongly and beat
Rubinstein, Tartakower, Yates and Gruenfeld. He lost to Rti and
some lesser masters. His other major tournament was the Carlsbad
(Karlovy Vary) 1929 marathon in which he scored 50% (+7 =7 -7)
in a field consisting of all the world's top players except Alekhine
and Lasker. 

Matisons was missing from the Latvian team of the 1928 and 1930
Olympiads but the following year he did play at Prague and
defeated the reigning world champion in round 16. 

Matisons-Alekhine Prague ol 1931 1 e4 e5 2 Nc3 Nc6 3 Nf3 d6 4
d4 Bd7 5 Bc4 exd4 6 Nxd4 Nf6 7 0 0 Be7 8 Nf5 Bxf5 9 exf5 0 0?

Black should have played 9...Qd7! and if 10 g4 Ne5 11 Be2 g5!
when if 12 Bxg5 then 12...Nfxg4 13 Bxe7 Rg8 14 Kh1 Qxe7 etc..
This surprising line had been analysed with Karlis Betins at the
Riga Seniors Chess Club, but Matisons gambled that Alekhine
would not find it over the board.

10 g4 Nd7 11 f4 Nb6 12 Bd3 f6 13 a3 

Matisons decided to prevent ...Nb4 but he said afterwards that 13
Be4! would have been stronger.

13...d5 14 h4 Nd4! 

Now Alekhine has counterplay. 

15 Be3 c5 16 Bxd4 cxd4 17 Ne2 (See Diagram)

17...Nc4! 

This pawn sacrifice equalises. 

18 Bxc4 dxc4 19 Qxd4 Qxd4+ 20 Nxd4 Bc5 21 Rad1 Rad8 22 c3
Rd5 

If 22...Rfe8 23 Kg2 Bxd4 24 cxd4 with a passed pawn that could
prove awkward for Black.

23 Kf2 Rfd8 24 Ke3 h5 

Black thought he had the initiative, but underestimated White's
next move. 

25 Ke4! hxg4 26 Rg1 R5d7 27 Rxg4 Re7+ 28 Kf3 

In my database the continuation was given as 28...Rd5 29 Rg2 b5
(following Arpad Foldeak's book) but we shall follow the move
order in the Latvian book which is more likely to be authentic. 

28...Rd5

On 28...Red7 Matisons again intended 29 Rg2.

29 Rg2 b5 30 Rgd2 Kf7 31 h5!?

White now threatens Ne6, but is this the best move? In his notes,
Matisons mentions the trap 31 Nxb5 Re3+ 32 Kg4 Rxf5 when 33
Kxf5?? loses to 33...g6+ 34 Kg4 f5+ 35 Kg5 Be7+ 36 Kh6 Rh3.
However, it looks to me that 33 Rd7+ 33...Kg6 34 Rxg7+ Kxg7 35
Kxf5 would perhaps win for White.

31...Bxd4 32 Rxd4 Rxf5 33 Rd5 Ree5 34 Rd7+ Re7 35 R1d5 Rxd5
36 Rxd5 a6 37 Rd6 (See Diagram)

Here Alekhine was in time trouble and made a fatal blunder.

37...Re6?

The King and Pawn ending is lost, whereas 37...Re1 would have
led to a draw.

38 Rxe6 Kxe6 39 Ke4 Kd6 

Or 39...f5+ 40 Kd4 Kd6 41 a4.

40 Kf5 Ke7 41 Kg6 Kf8 42 f5 Black resigned. 

Sadly, Matisons "only" scored 50% on top board in Prague+3 =8
-3. However, his other wins were against Akiba Rubinstein and Dr
Milan Vidmar, so he only beat top class opponents. He suffered a
rather short loss to Mir Sultan Khan. At this stage he was probably
already suffering from the TB which killed him 16 months later.

In his book "Solving in Style". GM John Nunn writes that
Matisons "was a strong player in addition to being a fine study
composer and it is unfortunate that he is remembered mainly for
losing a famous rook ending to Rubinstein". That was at Carlsbad,
but Matisons got his revenge with another instructive ending as
follows.

Matisons - Rubinstein, Prague ol, 1931 1 e4 e5 2 Nc3 Nf6 3 f4 d5
4 fxe5 Nxe4 5 Nf3 Nc6 6 Qe2 Nc5 7 d4 Ne6 8 Be3 a6 9 Nd1 Be7
10 c3 0 0 11 Qc2 f5 12 Bd3 Ng5 13 Nxg5 Bxg5 14 0 0 g6 15 g3
Bxe3+ 16 Nxe3 Be6 17 Ng2 Ne7 18 Nf4 Qd7 19 Rf2 c6 20 Qb3
Rab8 21 Re1 Kg7 22 Qb4 Ng8 23 Kg2 Rbe8 24 a4 Re7 25 Qa3 g5
26 Nxe6+ Qxe6 27 Ref1 Ref7 28 b4 Ne7 29 Qc1 h6 30 Bb1 Ng6
31 Qc2 Kg8 32 Rxf5 Nf4+ 33 R1xf4 gxf4 34 Rf6 f3+ 35 Kf2 Qh3
36 Qg6+ Rg7 37 Rxf8+ Kxf8 38 Qf5+ Qxf5 39 Bxf5 Ke7 40 Kxf3
b6 41 Kf4 Rg8 42 g4 a5 43 bxa5 bxa5 44 h4 (See Diagram) 

As a study composer, Matisons was now in his element. 44...Rb8
This lost quickly 45 g5 hxg5+ 46 Kxg5 Rb3 46...Rb2 47 h5 leads
to play similar to variations we will discuss shortly. 47 h5 Rxc3 48
h6 Rc1 49 h7 Rh1 50 Bg6 Threatening 51 Bh5. 50...Kf8 51 Kf6
Rf1+ 52 Bf5 Black resigned.

Returning to the last diagram, Euwe (according to the Latvian
book) claimed to have found a draw for Black with 44...Rg7,
analysing 45 g5 hxg5+ 46 hxg5 Rg8 47 g6 Rh8 48 Kg5 Rh1 49
Bg4 Rg1 50 e6 Kf8 51 Kf4 Kg7 52 e7 Re1 53 Bd7 Rxe7 54 Bxc6
Rc7 55 Bxd5 Rxc3 56 Bf7 Ra3 57 Be8 Kf8=.

However, White can improve on this. Matisons demonstrated 45
Kg3! Rg8 46 g5! (See Diagram)

Now if a) 46...hxg5 47 h5! Rb8 48 Kg4 Rb2 49 Kxg5 Rh2 50 h6
Rh1 (50...Kf7 51 Bg6+ Kf8 52 Kf6! Rxh6 53 e6+-) 51 h7 Kf7 52
e6+ Kg7 53 Bg6 Re1 (53...Rg1+ 54 Kf5 Rxg6 55 h8Q+) 54 h8Q+
Kxh8 55 Kf6+- and Black must soon give up the rook. Or if b)
46...h5 47 Kf4 when b1) 47...Rb8 48 g6 Rb3 49 Kg5 Rxc3 50 Kh6
Rg3 51 g7 Kf7 52 e6+ Kf6 53 e7 Kxe7 54 Bg6; b2) Finally, iIn the
variation 47...Kf7 48 g6+ Kg7 49 Be6 Rb8 50 Kg5 Rb3 51 Kxh5
Rxc3 Black creates his own passed pawn, but it is not quick
enough52 Kg5 Rg3+ 53 Bg4 c5 54 h5 c4 55 h6+ Kg8 56 e6 c3 57
e7 Re3 58 Kf6 and White wins.

To sum up, Matisons never achieved consistent success as a player
but on his day he was able to beat the best. As a study composer,
he was truly great and I recommend you to seek out more examples
of his work and try to solve them. You will find them truly
rewarding.

Solution to the mate in two: 1 c6 Threat Qc5. If 1..Kxd6 2 e5. 

Solution to the mate in four: 1 Kb1 Threat 2 c4; White wants to
give a N check on either c3 or g3. If 1 c4? Qxc4+ delays the mate
by two moves; 1 b3? Qxc2+ delays it one move. Also inadequate
are 1 Rd6? Rf3 and 1 Rc4? Qe5. After 1 Kb1, the main variation is
1...Rf6 to pin the N after 2 c4 Qh7. So White plays 2 b3 and now
since 2...e5 allows mate in one by 3 Ng3, the Rook must interfere
with the queen on either f7 or f5. Or if the N moves, it either
prevents ...Qh7 or blocks the diagonal from h7 to b1. The finish
could be 2...Rf5 3 c4 Qxc4 4 Ng3 mate.

Solution to the second endgame study: 1 Rd5! Rxd5 Not 1...d2 2
Bxd2. 2 Be3!! Threatening Bf7 mate. It is easy to see that 2 Bxd5?
fails to 2...Kxh6= and that if 2 Bc1? then 2...d2 blocks the mate. It
is quite another thing to spot that the answer to 2 Bd2? is 2...Rf5!!
when there is nothing better than 3 Bxf5 stalemate. That doesn't
work now because the d-pawn has a move. The Latvian book gives
another try, not mentioned by Roycroft, which adds to the
precision of the solution - 2 Bf4? Rg5 3 Bf7+ Rg6 when 4 Kh2
Kg4 attacks the bishop, which is why it has to go to e3. Computers
try to win this line by 4 Be8 (instead of 4 Kh2) 4...d2 5 Ba4 but
don't understand that after 5...Ra6 6 Bd1+ (6 Be8+ Rg6 just repeats
moves.) 6...Kg6 two bishops can't beat the rook even if Black loses
his pawn. 2...Rg5 3 Bf7+ Kh6 3...Rg6 trying to set up a stalemate
is defeated by 4 Kh2 when Black must abandon the rook. 4 Be8
The final point is that White must avoid 4 Kh4?? Kg7=. 4...d2 5
Bxd2 and wins the rook. You can see the practical player, the
problemist and the endgame wizard all coming together in this
mature study by Matisons, one of his last.

I am grateful to Correspondence Chess Women's GM Ingrida
Priedite for telling me the story of Matisons, to FIDE GM Igors
Rausis for the book and to Val Zemitis for a copy of his article
about Paris 1924. The English-language magazine Latvian
Correspondence Chess and Latvian Gambit is published six times a
year with the help of John Donaldson from Inside Chess magazine.
Please support this venture by sending US$18 (or equivalent) for
surface mail, or $24 airmail to circulation manager Val Zemitis,
436 Citadel Drive, Davis, CA 95616 USA.