This week's BCE position comes from the 1928 tournament at Bad Kissingen. This
tournament was a big triumph for Bogoljubov who finished clear first at 8 out of 11
a full point clear of Capablanca. Tartakower and Nimzowitsch finished in
the middle of the pack. If Tartakower had converted today's endgame they would
have both finished with even scores. Tartakower was well on his way to victory
after **35...Rxd4 **

Black is only a pawn down, and will soon pick up the b-pawn,
but he has no passed pawns and the White a-pawn is able to get far advanced while Black is collecting the b-pawn.
**36.a5 Rb4 37.a6 Rxb2 38.Rb7 Ra2 39.a7 **
The threat of Rb8+ forces Black to abandon the f-pawn.
**39...Kh7 40.Rxf7 Kg6 41.Rb7 **
The starting position for BCE-373
**41...Kf6 42.Ke1 g6 43.Kd1 Ke5 44.Kc1 Kd5 45.Kb1 Ra6 46.Kb2 Kc6 47.Rg7 Kc5
48.h4 **
White can't go after the kingside pawns yet since
48.Rxg6 Rxa7 49.Rg5+ Kd4 50.Rxh5 Rf7 picks up the f-pawn and draws.
**48...Ra5 49.g4? **
Tartakower tries to take advantage of the position of the Black king and
rook to create a passed pawn on the kingside with a tactical stroke,
but Nimzowitsch has a counter.
The patient approach with 49.g3 is the BCE line
**49...hxg4 50.h5 g3! **
Black creates his own passed pawn, which gives sufficient counterplay.
**51.fxg3 e3! 52.Rxg6 **The other choices also lead to a draw.
52.h6 e2! 53.Re7 Rxa7 54.Rxe2 Rh7
or 52.hxg6 e2 53.Re7 Rxa7 54.Rxe2 Rg7 55.Re6 Kd5 56.Ra6 Ke5
**52...Rxa7 53.Kc2 Ra2+ 54.Kd1 Kd4 55.h6 e2+ 56.Ke1! Ke3 57.Re6+! Kf3! **
It isn't too late to go wrong with 57...Kd3? 58.Kf2! and the e-pawn is contained.
**58.Rf6+ Ke3 59.Re6+! Kf3! 60.h7 Ra1+ 61.Kd2 Rd1+ 62.Kc2 Rd8 63.g4 Rh8 64.Kd2 Rxh7 65.Rxe2 Rd7+ 66.Ke1! Ra7 67.Rf2+ Kg3 68.Rf8 Kxg4 1/2-1/2**

We're coming down the home stretch in the Olympiad/Yearbook 128 series. Luis Rodi presented a third survey on the Advance Variation of the Caro-Kann. It had an interesting premise that even though todays engines are super strong, they still are not perfect, so positions that have an evaluation of 0.00 may not be equal if you go deep enough. Unfortunately, the position he focused on did not get any testing during the Olympiad. Instead, I picked a game from the Moldova-Georgia match between Dimitry Svetushkin and Levan Pantsulaiathat that featured a theoretical sideline.

**1.e4 c6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.Be2 Nd7 **
5...c5 6.Be3 Qb6 7.Nc3 Nc6 8.Na4 Qa5+ 9.c3
was the subject of Rodi's survey.
**6.0-0 Bb4!? **

An odd looking move.
White often plays for a queenside bind with c3 and b4 in this variation,
and the text seems to walk right into that.
However, Black reckons that his bishop is better placed to help attack
White's center from c7 than it would be on e7.
Magnus Carlsen has even tried the idea a couple of times,
although he moved the bishop to b4 on the previous move.
**7.Nbd2 **
This move and 7.c3 have the same number of games in the database.
White has scored heavily in both lines.
**7...Ba5 8.Nb3 Bc7 9.Bg5 **
9.Ne1 was the move in the big upset by Sanan Sjugirov over Carlsen in the 2010 Olympiad.
**9...Ne7 **
9...f6 is the main alternative, but it has scored even worse than the text
**10.Nh4 Be4 11.f3 Bg6 12.f4**

**12...f6 **
12...Be4 as used successfully by Dubov against Volokitin in the
2016 World Rapid Championship might be better than starting to
open the position. The arrangement of the Black pieces makes it
hard for the king to exit to the queenside.
**13.exf6 gxf6 14.Bh6 Qb8 15.Bd3 Kf7 16.Qe2 Re8 17.Rae1 Nf5 18.Nxf5 exf5 **
18...Bxf5 19.Bxf5 exf5 20.Qh5+
**19.Qf3 Nb6 20.h4 Nc4 21.h5 Nd6**

**22.g4 **
White could pick up the bishop immediately
22.hxg6+ Kxg6 23.Qg3+ Kxh6 24.Kf2 with a devestating Rh1+ coming.
However, the bishop isn't going anywhere
**22...Rh8 **
22...fxg4 23.hxg6+ hxg6 24.Bxg6+ is decisive
**23.Re2 Qc8 24.g5 Ne4 25.Rg2 Qg8 26.Qh3 Qd8 27.Nc5 Bb6 28.Nxe4 fxe4 29.f5 **

White is hunting more than just the bishop.
He is willing to give up Bd3 in order to open the f-file for his rook.
**29...Bxd4+ 30.Kh1 exd3 31.fxg6+ hxg6 32.hxg6+ 1-0**
It's mate after
32...Kxg6 33.gxf6+ Kf7 34.Rg7+ Kf8 35.Re7+ Rxh6 36.Qxh6+ Kg8 37.Qg7# 1-0

I discovered another pawn ending that I misplayed with the assistance of the engine. This one comes from the 5th edition of Fred Gruenberg's Put the Fun Back into Chess tournament. Many players might know Mr. Gruenberg as the organizer for many years of the National Open in Las Vegas, but his organizing career started in Chicago and Put the Fun Back into Chess was his trademark tournament. It had everything you could want in a tournament, good prizes (15 GP points plus many class prizes) with a low entry fee ($15) plus free lunch, free snacks and raffles for prizes. The 1984 version drew in 170 players including 2 GMs among the 15 masters.

In the second round I had Black against Keith McDaniel after **34...Kb6**
Despite the number of White pawns fixed on light squares,
White can simply shuffle his bishop back and forth between d1 and f3
(or e2 if Black brings his bishop to g2) and Black is unable to take
advantage of any of those weaknesses. Even the attempted pawn break
with ...a4 can just be ignored. However, without realizing the danger, White
kept the bishop on the open f1-a6 diagonal.

**35.Bc4?**35.Bd3? and 35.Bf1? are met with the same reply. In those cases
White can stay in the bishop ending, but then the bishop can get to the squares
it needs to in order
to start winning pawns.
For example, 35.Bf1? Ba6 36.Bh3 Be2 37.Ke3 Bd1 38.Kd4 (he still needs to keep c5
covered) 38...Bf3 39.Kc4 Be4 and Black gets the a2 pawn.
**35...Ba6!**Black forces a winning pawn ending
**36.Bxa6 Kxa6! 37.Ke3 Kb6 38.Kd3 Kc5 39.Ke4 a4 40.bxa4 **

**40...Kc4? **This seems simple enough, to save his d-pawn,
White must distract Black with the a-pawn.
But after Black collects the a-pawn he dissolves the queenside pawns and
outflanks White to win the d-pawn after all.
Indeed, that is how the game played out. The road to victory was to
retreat and collect the a-pawn first 40...Kb6! 41.Kd4 Ka5
**41.a5? **
It turns out the d5 pawn just gets in the way of White's counterplay.
He could hold the draw with 41.Ke3! Kxd5 *
(41...Kc5 42.Kd3! Kb6? 43.Kc4 *and Black doesn't collect the a-pawn*) *
42.a5! Kc5 43.Ke4! Kb5 44.Kd5
which leads to a queen ending that should be drawn,
but White has the better practical chances.
Black has to decide where his king ends up, a6 or a5.
44...Kxa5 *
(44...Ka6 *
may be slightly better so that the king is on a5 at the end of the variation *
45.Kc6 Kxa5 46.Kd7 d5 47.Kxe7 d4 48.Kf7 d3 49.e7 d2 50.e8Q d1Q) *

45.Kc6! Ka6
(the problem is that Black can't go forward 45...Ka4? 46.Kd7! Ka3
* (46...d5 47.Kxe7! *and White will queen with check*) *
47.Kxe7! Kxa2 48.Kxd6! b3 49.e7! b2 50.e8Q b1Q 51.Qa4+ Kb2 52.Qb4+
and White mops up the kingside after the queen exchange.)
46.Kd7 d5 47.Kxe7 d4 48.Kf7 d3 49.e7 d2 50.e8Q d1Q

with the same queen ending as the note to 44...Ka6,
except the Black king is on a6 instead of a5.
**41...Kb5 42.Kd4 Kxa5 43.Kc4 **
43.Kd3 Kb5 44.Kd4 b3 **43...Ka4 44.Kd3 b3 45.axb3+ Kxb3 0-1**

Lessons from this ending. 1. Like in the Newson game, don't discount backwards moves (40...Kb6! and 41. Ke3!). 2. Be alert for chances even in seemingly simplified postions. 3. As always, calculation is key!

This week starts a series of three positions from the games of Savielly Tartakower. Tartakower represented both Poland and France in his career. He was at his peak from about 1920 to the mid-1930s. Sonas has him at #3 in March 1921 behind Capablanca and Rubinstein, so he was no slouch, but I don't think he was ever considered World Championship caliber. In Volume 1 of his Great Predecessors series, Kasparov only gives one game fragment from Tartakower, a famous endgame loss to Capablanca.

Tartakower made a big mark as an author. 500 Master Games of Chess is a classic compendium of historical games. The Hypermodern Game of Chess is also well regarded, but not one I have in my collection. I thought the original German version from 1924 would be in the public domain, but my search came up empty.

Likewise, my search for further information on the source of this week's position was fruitless. NN is commonly used in chess literature for an unknown opponent. Tartakower did play in the Paris championship in 1933 finishing second behind Alekhine, but this position was not from that event. It might have come from a casual game, but not one that Tartakower published in his best games collection. Fine does cite some articles from the Belgian magazine L'Echiquier in the BCE bibliography, so perhaps the position came from there. However, I could not find that publication in the public domain either.

The cook at the end of BCE-300a hinges on two factors, the distance between the two remaining pawns and the active position of the White king. The pawns are far enough apart that the Black king can only block. With the pawns closer together, the opposing king has to spend time going around its own pawn as shown in the following study by Reti, which is also BCE-300b.

**1.Kf2! **
1.Kxg2? is premature 1...Ke4! 2.Kf2 e1Q+! to block the first rank
* (2...Kd3? 3.Ke1!) *
3.Kxe1 * (3.Rxe1+ Kd3!) *
3...Kd3! 4.Ra1 Kc3! 5.Rc1 Kd3! with a draw
**1...Ke4 **
On 1...Kf4 White wins the way he should have won in BCE-300a
2.Kxe2 Kg3 3.Ke3 Kh2 4.Kf2!
**2.Kxe2! Kd4 3.Rg1 Ke4 **
Going to the queenside mirrors the previous note
3...Kc3 4.Re1 Kb2 5.Kd2! Kb3 6.Rc1
**4.Re1** and the Black king doesn't have a good move **4...Ke5 **
Black's king gets cut off if he tries either side
4...Kd4 5.Kd2; 4...Kf4 5.Kf2 Now the fastest win is
**5.Ke3** and Black must cede more ground **5...Ke6 6.Ke4 Ke7
7. Kd3+** and the White king has time to get back to the kingside after
collecting the c-pawn **7...Kf6 8. Kxc2 Kf5 9. Kd3 Kf4 10. Ke2 Kg3
11.Rg1 Kh2 12. Kf2!**

With the king further away, even widely space pawns do not help as shown in the following study by , the king will also be too far away to support its pawns. The following study by Imre Bekey from 1933 illustrates this.

**1.Kb6 Kc8 2.Kc6 **
[2.Rxc2+? lets the Black king become too active.
2...Kd7! 3.Rc1 Ke6! 4.Kc5 Ke5
and the Black king will be able to support the h-pawn
**2...Kd8 **
The Black king can't get to the h-pawn in time after
2...Kb8 3.Re1 Ka7 4.Kb5 Kb7 5.Kb4 Kc6 6.Kb3 a1Q 7.Rxa1 Kd5 8.Kxc2 Ke4 9.Rh1
**3.Re1 **
forcing Black to give up a pawn
**3...a1Q **
3...Kc8 4.Re8#;
3...c1Q+ 4.Rxc1! Ke7 5.Kd5 Kf6 6.Ke4 Kg5 7.Kf3 Kf5 8.Ra1 is similar
**4.Rxa1! Ke7 5.Kd5 Kf6 6.Ke4 Kg5 7.Kf3 Kf5 8.Rc1 **
8.Kg2? Ke4 would throw the win away
**8...Ke5 9.Kg2 1-0**

The final game from the closed openings in my Olympiad/Yearbook 128 series is a heavyweight battle from Board 1 of the Russia-India match between Ian Nepomniachtchi and Viswanathan Anand. In this case, the survey by David Cummings proved to be highly topical and the players delivered a complex game. As this was one of the more prominent matches, this game got a lot of coverage by Sopiko Guramishvili and Ivan Sokolov during the live broadcast. Nepo was definitely pushing, but Vishy defended and earned the draw.

**1.c4 Nf6 2.Nc3 e6 3.e4 d5 4.cxd5 exd5 5.e5 Ne4 6.Nf3 Bf5 **

The starting point of Cummings survey. **7.d3 Nxc3 8.bxc3 c5 9.d4 Qa5 **
This move was introduced by
Ding against Nakmura at 2018 Norway chess.
Coincidentally, Ding was playing on the table next to this game.
Anand spent a lot of time during the opening and it appeared he was
trying to remember his analysis. He would stare off to his right, which
made it appear like he was looking right at Ding. **10.Bd2 Nc6 11.c4 **
Cummings calls this the critical try. Nakamura played
11.Be2 and a complicated game ended in a draw.
**11...Qd8 12.Qb3 Be6 13.Qxb7 Rc8 14.Ng5 Nxd4 15.Nxe6 fxe6 16.Rb1 Be7
17.Bd3 Kf7 **

A surprising move.
Guramishvili and Sokolov had only looked at the more natural 17...0-0
which seems reasonable for Black although like in the game,
the two bishops should give White the long term chances.
The king move gives the e6-pawn and e7-bishop extra support,
but the king looks very exposed on f7
potentially facing danger down the f-file across the 7th rank and on the a1-g8 diagonal.
Generally this is the kind of a computer move that would have been prepared at home.
However, I wonder if Anand had looked at Cummings analysis, and remembered
the move, but not the exact position.
Cummings continued 17...Nc6 18.f4 Kf7
* (*again this move instead of *18...0-0 **) *
19.Qa6 Nb4 with counterplay

.
**18.0-0 Nc6 19.Be2 Nd4 **
19...Nxe5 20.Bf4 opens lines and puts Black under pressure.
**20.Bd3 Nc6 21.Rfd1 Rc7 **
again 21...Nxe5 22.Bf4 Nxd3 23.Rxd3 gives plenty of compensation
**22.Qb3 Nd4 23.Qb2 Rd7 24.cxd5 Rxd5 25.Bc4**
Defending the pawn with 25.f4 runs into 25...Nf3+
**25...Rxe5 26.Qb7 Qa8**

** 27.Bf4 **This might be the point where Nepo lost his advantage.
27.Qc7 keeping the queens on, seems like a better way to pursue the attack.
**27...Re4 28.Bg5 Nf5 29.Bf1 **
29.Bd3 Qxb7 30.Rxb7 Rd8 makes it more difficult for White to regain his pawn.
**29...Rb4 30.Rxb4 cxb4 31.Rd7 Qxb7 32.Rxb7 a5 **

A nice move from Anand, making White have to spend extra time to
regain his pawn and setting up b vs. a on the queenside,
which will allow Black to swap the pawns with Rb8, Nd4 and b3.
White might still have a little something to play for after
32...h6 33.Bxe7 Nxe7 34.Rxb4
although I would expect Anand would be able to hold this as well.
However, I think finesses like this are what separate the super GMs from
the rest.
**33.Ra7 **
33.g4 h6 is the tactical justification of Black's last move
**33...Rd8 34.Bxe7 Nxe7 35.Rxa5 Rb8 36.Bb5 Nf5 37.h4?! **

I'm not sure what this move is about.
White is trying to distract the Black knight from the d4 square,
but a pawn down, he certainly can't have further winning ambitions.
37.g4 or 37.Bc4 are both met by 37...Nd4 and Black
has no problems **37...Nxh4 38.Bc4 Ng6 39.Ra7+ Kf6 **
39...Ne7 is equal per the engine, but Black does have an extra pawn.
**40.Ra6 Nf4 **
40...Nf8 maintains the extra pawn, but it is hard for Black to make progress.
**41.g3 Rc8 42.Bb3 Ne2+ 43.Kg2 Nd4 1/2-1/2**After 44.Rb6, White will regain
is pawn, but Black can trade knight for bishop leading to a drawn rook ending.

The trilogy of Marshall positions is wrapped up with a very subtle and instructive rook ending from his match against Capablanca in 1909. This match was the explosive introduction of the Cuban maestro on the international chess scene. Marshall was one of the top players in the world at that point, while Capablanca very little known. This match changed all that. Capablanca won 7 of the first 13 games with only 1 loss. Marshall. After a series of 9 draws, Capablanca won the final game. On Sonas' rankings, Marshall was #9 on the April 1909 list and Capablanca did not appear. One month later he has Capablanca as equal 2nd with Rubinstein, trailing only the World Champion, Lasker.

Today's BCE position is from the 9th game. This may have been the last critical point in the match. Marshall trailed by 3 points at the start of this game, so maybe things would have been different if he had converted one of the several opportunities he had. Fine gave the game continuation as one variation, but didn't point out many errors, so the comments are all in the correction link this week.

A few months after his disasterous World Championship match against Lasker, Marshall
was back in action at the 6-player quadruple round robin Ostend tournament,
which effectively was a candidates
tournament for the next World Championship match. Tarrasch won a tight race, half a point clear of
Schlechter, with Marshall and Janowski tied for 3rd another half point back.
We previously saw a BCE example
between Schlechter and Janowski. Today, features the other two players from the
leading group. Tarrasch seemed well on his way to victory with Black after **59.Nf3**

However, here he inexplicably played
**59...Qxd5? **
In
the tournament book,
Tarrasch did not award this move a question mark,
but mentioned that Black wins without difficulty with 59...Rxd5.
Perhaps he thought he could just walk his king forward and pick up
the pinned knight.
**60.Qf8+ Kh7 61.Qe7+ Kg6 62.Qe8+ Kf5 **
62...Kf6 63.Qf8+
**63.Qh5+ Ke4 **
this breaks the pin on the knight so White regains the exchange.
**64.Nxd2+ Qxd2 **
The starting position for BCE-573.
The more active king and advanced e-pawn give Black winning chances,
but objectively, the position is a draw. **65.Qg6+ Kd5 **
65...Kf3 66.Qf5+ picks up the e-pawn
**66.Qxa6 e4 67.Qb5+ Kd4 68.Qb6+ Kd3 69.Qa6+ Ke3 70.Qxh6+! Ke2 71.Qh5+ Ke1
72.g4? **
f5 and b1 are key squares for White to deliver perpetual check.
He would draw with 72.Qf5
as shown in the correction link
**72...e3 73.Qc5 Kd1 **
In his annotations, Tarrasch gave
73...Kf1? as a faster win. He had calculated as far as
74.Qf5+! Qf2 75.Qb1+ Qe1 76.Qf5+! Ke2+ 77.Kg2 Qf2+ 78.Kh3!
and abandoned the line as drawn since both sides would get new queens
after a trade on f5. In his notes, he continued the variation with
78...Qf3+ 79.Kh4? Kf2 with an eventual win. In this line
*79...Qxf5 *is much easier as now Black queens first with check.
But instead of walking onto the e1-h4 diagonal, White would draw with
*79.Kh2! Kf2 80.Qc2+! e2 81.Qc5+! Qe3! *else Black gets mated *
82.Qf5+! Ke1 83.Qb1+! *with perpetual check*) *
**74.Qf5 Qf2? **Tarrasch correctly identified this as the
final mistake giving the winning line
74...e2 75.Qb1+ Qc1 76.Qd3+ Ke1 77.Kg2 Qc6+
which Fine extended a few more moves
78.Kg1 Qc5+ 79.Kg2 Qf2+ 80.Kh3 Kf1 81.g5 Kg1 and Black queens
**75.Qb1+ Ke2 76.Qb5+ Kf3 77.Qd5+ Kg3 78.Qe5+ Kxg4 79.Qe4+ Kh3 80.Qe6+ Kg3 81.Qe5+ Kf3 82.Qd5+ Ke2 83.Qb5+ Kd1 84.Qb1+ Kd2 85.Qb2+ Ke1 86.Qb1+ Ke2 87.Qb5+ 1/2-1/2**

This week starts a series of three positions from the games of the legenday American Frank Marshall. Marshall was one of the first players given the title of "Grandmaster". His move 23...Qg3 against Levitsky is one of the most famous moves in chess history with legend having it that the spectators showered the board with gold pieces in appreciation of its brilliance. Among his many contributions to opening theory, his gambit idea in the Ruy Lopez is still one of the absolute main lines of opening theory, despite its unsuccessful debut.

Marshall came close to the ultimate chess summit, playing a match with Lasker for the World Championship in 1907. However, this turned out to be an total rout with Lasker winning the first three, the last four, and one game in between for an undefeated 11.5-3.5 drubbing.

Today's game against Aron Nimzowitsch comes from the super strong quadruple round robin contested at New York 1927. Capablanca was at the height of his powers and dominated with an undefeated +8, 2.5 points clear of Alekhine, with Nimzowitsch a further point behind. Marshall was past his prime at this point and finished last with only one win.

Fine seems to have used Alekhine's notes from
the tournament book for this position. However, Alekhine was
uncharacteristically sloppy in his annotations of this game.
After **29.Rdg2**

**29...Bxe4 **
Alekhine gives 29...g5? as simpler, but 30.fxg5 hxg5 31.Rf2
* (*Instead of Alekhine's *31.Rxg5?) *
and the pin will cause Black to lose the exchange.
**30.dxe4 Rd3 31.Rxg7 Rxe3 32.Rg8+ **
Allowing Black to activate his king.
I think better winning chances were offered by 32.Re7
**32...Kd7 33.R1g7+ Kc6 34.Rg6 **
As Alekhine shows, Black has sufficient counterplay after
34.Rc8 Rxf4 35.Rcxc7+ Kd6 36.Rxb7 Rf2
**34...Rd6 35.e5 Re1+ 36.Kb2 Re2+ 37.Ka3 Rxg6! 38.Rxg6+! Kd5 39.Rxh6 **
The starting position for BCE-400 **
39...a5 40.Rh7 Rc2 **
Alekhine was critical of this winning attempt instead of the solid
40...Kc6 , but nothing is spoiled**41.Re7 **
Fine's line beginning 41.Ka4
is the subject of the first correction.
**41...b5 42.b4 **
The second correction occurs after 42.f5? where Alekhine also gives
42...b4+? instead of the winning 42...c6!
**42...a4 ** This move was criticized as losing by both Alekhine and
Fine. Instead, 42...axb4+ 43.Kxb4 Rc4+ 44.Kxb5 c6+ 45.Kb6 Rxf4 is a simpler draw.
**43.f5 c5? **
The real loser, Black can still hang on after
43...Rf2 44.f6 Rf3+ 45.Kb2 Rf2+ 46.Kb1 Rf1+ 47.Kc2 Rf2+ 48.Kd3 Rxa2 49.Ke3
* (49.f7 Rf2 50.e6 a3) *49...Ra1 50.Kf2 Ra2+ 51.Ke3 Ra1 52.Ke2 Ra2+
53.Kf3 Ra1 54.Kg2 Ra2+ 55.Kg3 Ra1
**44.f6? ** Both Alekhine and Fine point out that
44.e6! wins 44...Rc3+ 45.Kb2 cxb4 46.Rd7+ Kc6 47.Rd8 a3+ 48.Kb1 Re3 49.f6 b3 50.axb3 Re1+
51.Ka2! * (51.Kc2 a2!=) *
51...b4 52.e7 Re2+ 53.Kb1 Re1+ 54.Kc2 a2 55.Ra8 Kd7 56.f7
**44...Rc3+ 45.Kb2 cxb4 1/2-1/2**
Everything is coming off after
46.f7 a3+ 47.Kb1 Rf3 48.e6 Rf1+ 49.Kc2 Rf2+ 50.Kd3 b3 51.axb3 a2 52.Ra7 Kxe6
53.Rxa2 Rxf7

Archives