We close out the selection of BCE positions from Tarrasch's games with the notorious bishop+rook pawns rook ending. While this is generally a theoretical draw, it can be difficult to hold in practice. Despite having looked at this ending in preparing this post, I was unable to hold it recently with little time on the clock vs. GM Zapata.

In 1911, Tarrasch played a match with Carl Schlechter, who was coming off
of his drawn World Championship match with Lasker. Tarrasch took the lead
3 times, but Schlechter was able to answer each time and the 16-game match ended
in a draw. The 9th game was a critical one, a Schlechter win sandwiched between
two Tarrasch wins. It was a blunder-filled affair with each player winning at
various points. They eventaully reached a rook ending after
**58.Rxe3 **

**58...Kxf6 59.Ra3 h5 **
In
the book on the match,
Tarrasch was critical of this move and the next one,
awarding them both question marks.
Instead, he argues that Black should wait with his
king in front of the pawns, which shields him from checks from behind.
However, objectively the position is still a fairly easy draw with the
king behind the pawns. **60.Kg2 g5?! **
This is the move that makes the job more difficult.
When the a-pawn reaches a7, the Black king will have to wait on g7 or h7,
which leaves the kingside pawns undefended.
This allows White to trade the a-pawn for the g and h-pawns
leading to the f+h pawn rook ending. Instead, 60...g6 would be simpler.
**61.Ra8 ** White could put his rook behind the passed pawn with
61.Rf3+ Kg6 62.a4 Ra5 63.Ra3 but 63...g4 gives Black sufficient counterplay
**61...Rd3 62.a4 Ra3 63.a5 h4 64.a6 Kg7 65.f3 Kh7 66.Kf2 Kg7 67.Ke2 Kh7
68.a7 **
White ties down the Black rook before heading towards the queenside.
68.Kd2?! Rxf3 69.Rb8 Ra3 70.Rb7+ Kg6 71.a7 g4 is fine for Black
**68...Kg7 69.Kd2 Kh7 70.Kc2 Kg7 71.Kb2 Ra6 72.Kc3 Ra1 73.Kb4 Ra2
74.Kb5 Ra1 75.Kc6 Ra2 76.Kd6 Ra1 77.Ke6 Ra5 78.Re8 Ra6+! **
78...Rxa7? 79.Re7+! Rxe7+ 80.Kxe7!
White has the opposition, so he wins the pawn ending.
**79.Kf5 Rxa7 80.Kxg5 Ra3 81.Kg4 Rb3 82.Re4 Kg6 83.Re6+ Kg7
84.f4 Rg3+ 85.Kxh4**
Tarrasch thought this ending was a win for White and gave no further notes.
**85...Rg1 **Finally reaching the starting position of BCE-352
**86.Re3 Kf6 87.Rg3 Ra1 88.Kg4 Ra8 89.Rb3 Rg8+ 90.Kf3 Rg1 91.Rb6+ Kf5
92.Rb5+ Kf6 93.Rb8 Rf1+ 94.Kg4 Rg1+ 95.Kh5 Rg7 96.Rf8+ Ke7 97.Rf5 Ke6
98.Rg5 Rh7+? **
Fine shows the way to the draw here with
98...Rf7! 99.Kg4 Ra7 100.h4 Ra1 101.Rg6+ Kf7 102.Rb6 Kg7
* (*The Black rook must remain flexible to check from the side or
from behind. The passive *102...Ra7? *loses *
103.Kg5! Ra5+ 104.f5! Ra7 105.h5 Kg7 106.h6+ Kh7 107.Re6 Rb7 108.Re5 Ra7
109.f6 Ra1 110.Re7+ Kg8 111.h7+ Kh8 112.f7) *103.Kf5
* (*Now *103.Kg5 *is met with *103...Rg1+!) *
103...Ra5+ 104.Kg4 Ra1 105.Kg5 Rg1+! 106.Kh5 Rf1]
**99.Kg6! Rxh3 100.Re5+! Kd6 101.Re1 Rh8 102.f5 Rg8+ 103.Kf7 Ra8
104.f6 Ra7+ 105.Kg6 Ra2 106.Rg1 1-0**

This BCE position also features a bonus game. Instead of Tarrasch's
**95...Rg7**, Fine's main line continues with **95...Kf7** and soon
reaches a position from the game Marshall-Rubinstein, San Sebastian 1911.
This tournament was played several months before the Schlechter-Tarrasch
match and Tarrasch's annotations to the game appear in
the tournament book.
I found it very interesting that Tarrasch thought that this ending was
drawn, while he thought the similar position he had against Schlechter was
lost. I'm not sure when his annotations were made, which might help explain
the differences.

Against Marshall, after
**49.Kb2 **

Rubinstein allowed the rook ending with **49...Nb6? **
keeping the knights on the board offered more winning chances.
**50.Nxb6 Kxb6 51.Kb3 Rb4+ 52.Kc3 Kb5 53.Rb8+ Ka4 54.Rc8 Rb3+
55.Kc2 Rb5 56.Rh8 **
This is BCE-352 with colors reversed after the 10th move
**56...Kb4 57.Rh1? **
Fine gives this move an exclam, but it should lose.
Instead, White draws with Fine's alternative line
57.Rh4+ c4 58.Rh8 a4 59.Kb2! a3+ 60.Ka2! Rd5 61.Rb8+! Kc3
62.Rh8 Rd1 63.Rh7 Kc2 64.Kxa3! c3 65.Rh2+ Rd2 66.Rh1=
**57...a4! 58.Kb2 a3+? **
See the BCE link for the win beginning with 58...Ka5+!
**59.Ka2! Ka4 60.Rc1! Ra5 61.Rb1 c4 62.Rb8 Rc5 63.Ra8+! Kb4 64.Rxa3 c3
65.Rb3+! Kc4 66.Rb8 1/2-1/2**

I've added my 2005 encounter with Alex Fishbein to the GM games section. I briefly checked my earlier notes on this game with an engine, mostly to confirm the equalizing opportunity at move 24.

Returning to the series from Tarrasch's games, we have a position from
a tournament closer to the end of his career. Teplitz-Schoenau was a strong
tournament, but lacked the star power of players such as Capablanca, Alekhine,
and Lasker. Despite a number of famous names, Tarrasch was the only player
in the field who ever played a World Championship match. However, this was
not a great tournament for Tarrasch as he only had one win and finished
in a tie for 11th-13th in the 14 player field, well off the pace set by the
winners, Reti and Spielmann. The only player who finished with a lower score
was his opponent in today's game, Paul Johner, who even managed more wins
than Dr. Tarrasch. After **30...Ke7**

Tarrasch went for the rook ending with **31.Nxf6+ **
The tournament book
also looks at taking the b-pawn
31.Nxb6+ Ke6 The pin looks dangerous,
but there doesn't seem to be a way to exploit it if White just waits with
32.g4 since 32...Bd4 * (*Black could also try *32...g6 *
when it doesn't look like White has anything better than
*33.f5+ gxf5 34.gxf5+ Kxf5 35.Nc4 Rxb7 36.Nd6+ Kf4 37.Nxb7 a4 38.Nc5 Be7
39.Nxa4 Bxa3 *with an easy draw despite White's extra pawn.*) *
33.f5+ makes the Black king walk into a check allowing White to trade rooks,
but Black should not have any problems holding after
33...Kf6 34.Nd7+ Kg5 35.Rxb2 Bxb2
The tournament book looks at some ways White can get in trouble
if he tries to break the pin tactically
**a) **32.f5+? Kxf5 33.Nc4 Rxb7 34.Nd6+ Ke6 35.Nxb7
Now, a4! (instead, the tournament book's 35...Bc3?
lets the knight escape via d8 36.a4 Kd5 37.Kf1 Kc6
* (37...Kc4 38.Ke2 Kb4 39.Kd3 Be5 40.Nd8 Kxa4 41.Kc4 Ka3
42.f4 Bf6 43.Ne6 a4 44.Kb5 Be7 45.Nd4) *
38.Nd8+ Kd7 39.Nf7 h6 40.f4 Ke7 41.Ne5)
36.Nc5+ Kd5 37.Nxa4 Kc6-+ and the knight is trapped;
**b) **32.Na4? Rxb7 33.Nc5+ Kd5 34.Nxb7 a4 35.Kf1 Be7;
**31...Kxf6 **The starting position for BCE-397
**32.a4 h5 33.Kg2 g6 34.h3 Kf5 35.Rc7 Rb4 36.Rc6 Rb3 37.Rd6 Rb4
38.Rc6 Rb3 39.Kf1 Rb2 40.Ke1 Rb3 41.Rd6 Rb2 42.Kf1 Rb4 43.Ke2 Rb3
44.Kd2 h4 **
This move is condemned in the tournament book and by Fine,
but Black should still be holding.
**45.gxh4 Rxh3 **
45...Kxf4 also holds.
The tournament book has White winning with
46.Rf6+ Ke4
47.Rxg6 Rxh3 48.Ke2 Rxh4
49.Rxb6 Rh8 50.Rb5 Ra8 51.f3+ Kf4 52.Kf2, but Black holds with
52...Rh8 53.Ke2 Re8+ 54.Kd3 Ra8=
**46.Rxb6 Kxf4? **
46...Rxh4 is the topic of the BCE correction;
46...Ra3 also draws 47.Rb5+ Kxf4 48.Rxa5 Kg4 49.Rg5+ Kxh4 50.Rxg6 Rxa4
**47.Rf6+! **a key zwischenzug forcing the Black king to a bad square
Black escapes after 47.Rxg6? Rxh4 as White can't pick up the a-pawn*
(47...Ra3 *is also sufficient *
48.h5 Rxa4 49.h6 Ra2+ 50.Kc3 Ra3+ 51.Kc4 Rh3 52.Kb5 Kf5 53.Rc6 Kg5=) *
**47...Ke4 48.Rxg6! Rxh4 **
Black loses all his pawns after 48...Ra3 49.Rg4+ Kf5 50.Rg5+
**49.Rg5! Kf3 50.Rf5+ Kg2 51.Rxa5 Kxf2 52.Ra8 **
Fine stops here in BCE with the note
and White wins by marching the Pawn on to the seventh.

However, Dr. Tarrasch didn't follow this recipe and allowed a draw.
**52...Kf3 53.a5 Ke4 54.a6! Rh7 55.Kc3? **
55.a7! wins. Now, Black can use his rook to shelter his king's
approach to the pawn.
**55...Kd5! 56.a7 Rc7+! 57.Kb4 Kc6! 1/2-1/2**

I seemed to have jumped the gun on starting the Tarrasch series as I had one more position left for the Horowitz trifecta. So I'll complete this round of Horwitz studies this week and pick back up with Tarrasch next week.

Today's position is a very strange knight ending. The original seems to be from Horwitz' column in The Chess-Monthly which had the following position set as White to play and win.

The main line of the solution runs **1.Nb5 Kg7 2.Nc7 Kh7 3.Ne8 Nh6
4.Nxf6+ Kg7 5.Ne4** and wins. One interesting point about this ending
is that illustrates the fact that the knight cannot lose a move. The
position after 3.Ne8 is one of mutual zugzwang. If Black could pass, White
could not break down the fortress. Horwitz explains it another
way stating of the starting position
White wins in this position, if the Kt is orignially
on a black square; he draws only, if the Kt is on a white
square.

This is consistent with my explanation of the
position after 3.Ne8. If it is White to move, we have the starting position, but
with the knight on a light square.

The main problem with this study is that the construction is very
artificial. It is hard to figure out how this position, with the Black
pieces and the White king both stuck behind the pawn chain, would arise in
from an actual game. It appears that Berger tried to solve this problem
in his book,
which has the BCE starting position and is attributed to Nach Horwitz

which translates to after Horwitz

indicating that the position is
not the original. Benko completely removed this position from the revised
edition. I think it could have been kept as the correction to the
Berger version does illustrate domination of a knight. I think the problem
Berger faced was that his change to the position introduced a new winning
method that he was blind to since he was trying to keep the Horwitz zugzwang
solution.

This week begins a trio of BCE positions from the games of Siegbert Tarrasch, one of the giants of chess. Besides being a top player, he was also a prolific writer and theoretician. However, the top prize eluded him. His shot at the title came in 1908 against Lasker. Lasker won 4 of the first 5 games and won the match 8-3 with 5 draws. After each of Tarrasch's 3 wins, Lasker answered with a win in the very next game.

Lasker also topped Tarrasch in the tournament from which today's example comes.
Nuremberg 1896 featured most of the top players of the time and Lasker won with
13.5/18 ahead of Maroczy at 12.5 and Tarrasch and Pillsbury at 12. Tarrasch's
opponent in this game, Emanuel Schiffers was in the middle of the pack with 9.5.
In the position after **37.Bc6**

Schiffers allowed a transition to a pawn ending with
**37...Ne4+ **
In the tournament book,
Tarrasch gave 37...Nd1 as better,
but thought White was still better because of the bishop against the knight.
However, Black should be holding here too.
For example, 38.e4 fxe4 39.Bxe4 Nf2 40.Bc2 a5 41.f5 gxf5 42.Bxf5 h6
43.Kb5 a4 44.Kxa4 Ne4 45.Bxe4 Kd7!
with a draw since the White h-pawn queens on the wrong colored
square for his bishop.
**38.Bxe4 fxe4! **
The starting position for BCE-88a **39.Kc6 h6? **
Tarrasch said that the subject of the BCE correction
39...Kc8! was the toughest defense,
but implied that White still wins although he did not give any variations.
**40.h4 h5 **
Now 40...Kc8 allows a breakthrough
41.h5 gxh5 42.d7+ Kd8 43.Kd6 and mate in 4
**41.Kd5 Kd7 42.Kxe4 Kxd6 43.Kd4! Ke6 44.e4! Kd6 45.Kc4! Kc6 46.Kb4 Kb6
47.Ka4 Kc5 48.Ka5! Kd4 49.e5! Kd5 50.Ka6 Ke6 51.Kxa7 Kf5 52.Kb7 Kxf4
53.e6! Kg4 54.e7 Kxh4 55.e8Q g5 56.Qe1+ Kg4 57.Qg1+ Kf4 58.Qh2+ Kg4 59.Kc6 1-0**

Archives