26 November 2018

Have to back Caruana now

I generally don't have strong personal favorites in World Championship matches, although it's natural to pull a bit for one side.  After Game 12, however, I have to fully back Caruana going into the tiebreak of the 2018 World Championship.  Carlsen seriously violated the mental toughness rules** by offering a draw in a superior position, in what was the decisive classical game.  That certainly made it seem like all he was aiming for was a draw from the start.  He wasn't sick or exhausted or suffering from any other exogenous factors, either.

Caruana in the press conference even looked slightly taken aback, saying that "White could never be better" in the final position, then going on to mention how at least he was more equal at the end than he had been a few moves before, when there was more danger.  There is some justice in the outcome, however, as it seems Caruana could have essentially forced a repetition of moves earlier in the game, but chose not to.

GM Erwin L'Ami's video analysis - click on Round 12 in the Round-up show frame on the Chessbase 2018 World Championship Page - is well worth viewing, for a candid look at the game.

** draw rules:
  • Do not deliberately aim for a draw from the start of a game, regardless of your opponent's rating or your tournament standing.  If you play your best and press any advantages you are able to obtain, you are more likely to achieve what you need and may in fact win.
  • Resolve not to offer a draw to your opponent unless the position on the board is in fact completely drawn.  This will contribute to a winning mindset and to not being afraid to play out any position.

24 November 2018

Annotated Game #202: Breaking the trend

Although my results hadn't been terrible in the previous three games of this open tournament (1 out of 3 versus much higher-rated opposition, including my first-ever win over a master), I hadn't played very well at all.  I consciously understood that I needed to break this trend (see "Streakiness in Chess Performance") and did so effectively with this fourth-round game, facing an Expert.

The most important factor, both psychologically and in terms of chess skills, was that I was able to reach a position-type I knew quite well out of the opening.  Objectively it's balanced for both sides until around move 17, when my opponent essentially wastes a tempo and I am able to seize the opportunity to take control of the a-file.  From there on, I have the initiative and am able to go up a pawn, although after a tense confrontation and tricky sequence in the center, we end up in a draw.

I think that the idea of having better success out of "comfortable" positions - regardless of the objective measurement of equality - has gained a lot of force in recent years, primarily due to Magnus Carlsen's ability to win from seemingly dead equal positions, and his deliberate strategy of choosing solid openings many times that have no theoretical advantage whatsoever.  It certainly worked in this case for me, as I knew what to do in the middlegame transition phase better than my opponent did, so was able to seize the initial opportunity when it appeared on the board.

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "ChessAdmin"]
[Black "Expert"]
[Result "1/2-1/2"]
[ECO "A26"]
[Annotator "ChessAdmin/Komodo 11.2"]
[PlyCount "84"]

{A26: English Opening vs King's Indian with ...Nc6 and d3} 1. c4 g6 2. Nc3 Bg7
3. g3 d6 4. Bg2 Nf6 {going for a King's Indian setup, which was expected, even
with a somewhat unusual move-order.} 5. Nf3 O-O 6. O-O e5 7. d3 Nc6 8. Rb1 a5
9. a3 Re8 10. Bg5 h6 11. Bxf6 {the idea behind the exchange is that White's
dark-square bishop is constrained by all of the pawns on dark squares, so is
best traded for the Nf6, which is normally a good attacking piece for Black on
the kingside.} Bxf6 12. b4 axb4 13. axb4 {everything so far is very standard.}
Bg7 14. b5 Ne7 15. Qb3 {generally speaking, in these types of English
formation, it's sometimes hard to figure out where to put the queen. Here, it
seems clear that b3 is an excellent square, as it supports the b-pawn, doubles
up on the b-file, and lines up on the a2-g8 diagonal. There are no obvious
drawbacks, even with Black lining up his bishop to oppose the queen.} f5 $5 {
an uncommon (only one game in the database), aggressive play by Black, which
however is in keeping with typical KID ideas.} (15... Be6 16. Nd2 c6 17. Rfc1
d5 18. bxc6 bxc6 19. Qd1 Ra7 20. Na4 f5 21. cxd5 cxd5 22. Nc5 Bf7 23. Rb7 Rxb7
24. Nxb7 Qd7 25. Nc5 Qd6 26. e3 d4 27. Nb7 Qd7 28. Nc5 Qd6 29. Nb7 Qd7 30. Nc5
{Georgiev,K (2660)-Akopian,V (2600) Tilburg 1993 1/2-1/2 (41)}) (15... c6 16.
Rfd1 d5 17. cxd5 cxd5 18. d4 e4 19. Ne1 Qd6 20. e3 g5 21. Rdc1 Be6 22. b6 Rec8
23. Na4 Rxc1 24. Rxc1 Rc8 25. Rb1 Rd8 26. Nc2 h5 27. Na3 Bd7 28. Bf1 h4 29. Nb5
Qf6 30. Na7 {Oral,T (2415)-Simacek,P (2255) Prague 1997 1/2-1/2 (42)}) 16. Nd2
{this was played in the only DB game (see below for that continuation). It's a
common idea in the English, unleashing the Bg2 being the main idea.} (16. Nd5
Nxd5 17. cxd5 $11 {is preferred by Komodo. This is another common idea in the
English, where after an exchange on d5, doubled d-pawns are accepted in
exchange for gaining space and cramping Black.}) 16... Kh7 $146 {moving the
king off the diagonal, but now White gets a bit of initiative on the queenside.
} (16... f4 17. Nd5 Nxd5 18. Bxd5+ Kh8 19. Ra1 Rxa1 20. Rxa1 fxg3 21. hxg3 Bg4
22. Bxb7 Bxe2 23. Bc6 Rf8 24. Ra8 Qf6 25. Rxf8+ Bxf8 26. Ne4 Qf5 27. Qc2 Bf3
28. Nc3 Bxc6 29. bxc6 Be7 30. Nd5 Bd8 31. c5 {Seel,C-Eissing,C Pinneberg 1996
1-0 (55)}) (16... c6 $5 17. Ra1 Be6 18. e3 $11) 17. Ra1 {whenever White can
play this move with a Black bishop still on c8, it always has some punch to it,
due to the Ra8 hanging.} Rb8 (17... Rxa1 18. Rxa1 c6 19. Ra8 $14) 18. Ra7 $16 {
this is now a great rook, exerting lateral pressure on the 7th rank while also
dominating the a-file.} e4 {my opponent is willing to sacrifice a pawn for a
bit more space and activity, which I gladly accept.} 19. dxe4 $16 Bd4 {with a
"backwards" attack on the Ra7, which I calmly move back. Although it's a shame
it's no longer on the 7th rank, the pawn more than compensates for that.} 20.
Ra2 fxe4 21. Ndxe4 Be6 {it's understandable that Black wants to develop his
neglected bishop, but I could have reacted more strongly.} (21... Bg7 22. Rd1
b6 23. e3 $18) 22. Nd5 {part of Black's strategic problem has been his
relatively more cramped position and less-effective pieces. This exchange just
helps him, although it's not a terrible move in itself.} (22. Rd2 $5 {the
virtue of this move is that it shifts the rook to what is now a more effective
file on which to exert pressure.} Bg7 (22... Be5 23. b6 {and the idea still
works, by simultaneously threatening to exchange on c7 and undermine the d6
pawn. For example} Nf5 24. bxc7 Qxc7 25. Nd5 $18) 23. b6 $18) 22... Nxd5 $16
23. cxd5 Bf5 24. e3 {I thought for a long time before playing this, as the
continuation is rather complex.} Bxe4 25. Bxe4 Bb6 ({after} 25... Rxe4 26. Qd3
{is now the key move, forking the rook and bishop.} Qe8 27. Ra4 $16) 26. Qd3 (
26. Ra4 $5 {is a more creative idea, protecting the bishop and making the rook
mobile along the 4th rank.}) 26... Qf6 {developing the queen and protecting g6.
At this point I was a pawn up, but didn't have any good ideas on how to make
further progress.} 27. Kg2 Re7 28. Qc2 {I'm starting to make moves that mark
time rather than make threats, so Black is able to recover a bit of initiative.
} (28. h4 h5 29. Bf3) 28... Rbe8 29. Bd3 Kg7 30. Ra4 {finally I figure this
idea out.} Qg5 31. e4 {again, not a terrible move, but not necessarily in the
spirit of the position, either. It protects the d5 pawn, but contradicts the
idea of having the Ra4 laterally mobile and bottles up the Q+B battery. At the
time, I felt it would help shut down any counterplay by Black and it was
intended to help to mobilize my extra pawn.} (31. Qb3 $5 {is more flexible and
would allow the rook to subsequently transfer to f4 or h4, to good effect.})
31... Rf8 32. f4 Qf6 33. e5 (33. h4 $5 $16) 33... dxe5 34. fxe5 Qxe5 35. Rxf8 {
missing the critical continuation. Now the position becomes largely even.} (35.
Re4 $5 Rxf1 36. Rxe5 Rg1+ 37. Kh3 Rxe5 38. d6 cxd6 39. Qc8 {going after the b7
pawn and the soft underbelly of the 7th and 8th ranks, instead of the g6 pawn})
35... Kxf8 36. Bxg6 Qxd5+ 37. Be4 Qxb5 38. Ra8+ {deliberately going for a
drawing line, now that I'm a pawn down.} Kg7 39. Qc3+ Qe5 40. Qxe5+ Rxe5 41.
Bxb7 Re2+ 42. Kh1 Re1+ 1/2-1/2

09 November 2018

Annotated Game #201: The importance of falsifying all your moves

The best thing that can be said about this next game is that it wasn't a miniature, like the previous one.  There are some useful lessons from the opening / early middlegame phase, particularly regarding the move 10 decision to avoid doubled pawns that was not in fact best.  However, the main lesson is to falsify all of your planned moves - meaning, to expend the mental effort necessary to calculate if your opponent can refute them.  This theme was originally introduced in Annotated Game #35: Thou Shalt Falsify.  Sometimes I find it all too easy to slip back into laziness, when a move looks fine (for me), or can be made on "general principles".  It does take additional mental effort to meaningfully check for your opponent's possible responses, rather than simply giving the board a cursory look.  Probably the best piece of advice I have ever received from a martial arts master is "don't be lazy", so I try to replay that in my mind whenever I am tempted to cut corners.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "0-1"] [ECO "C42"] [Annotator "ChessAdmin/Komodo 11.2"] [PlyCount "46"] {[%mdl 8192] C42: Petroff Defence: 3 Nxe5 and unusual White 3rd moves} 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. e3 d5 5. cxd5 Nxd5 6. Bb5 Nxc3 7. bxc3 Qf6 {an unusual move to protect the e-pawn, immediately signaling aggressive intentions.} 8. d4 {White proceeds as usual in this variation.} exd4 9. exd4 { normally it's "dealer's choice" on which pawn to recapture with here, but with the queen on f6 I think it's probably better to use the c-pawn.} (9. cxd4 Bb4+ 10. Bd2 Bxd2+ 11. Qxd2 O-O 12. Rc1 Bh3 13. O-O Qg6 14. Ne1 Rad8 15. Bxc6 bxc6 16. f3 Rd5 17. Nd3 Rg5 18. Nf4 Qh6 19. Nxh3 Qxh3 20. Rxc6 Rc8 21. e4 Qh5 22. Qf4 Ra5 23. Rxc7 Rxc7 24. Qxc7 h6 25. Qc8+ Kh7 26. Qc4 Ra3 27. d5 Qe5 28. Rd1 Qb2 29. d6 Rxa2 30. Qxa2 {1-0 (30) Suba,M (2531)-Ljubarskij,J (2344) Bad Zwischenahn 2008}) 9... Bg4 $146 (9... Bd7 10. O-O h6 11. Re1+ Kd8 12. Ne5 Nxe5 13. dxe5 Qe6 14. Bxd7 Qxd7 15. Qf3 Kc8 16. Be3 Qe6 17. Rab1 c6 18. a4 Be7 19. a5 Rd8 20. a6 b6 21. c4 Kd7 22. c5 b5 23. Qg3 g5 24. Qf3 {Moracchini,F (2270) -Trinh,R (2230) Issy les Moulineaux 1989 1/2-1/2}) (9... Bd6 10. Bg5 Qe6+ 11. Kf1 $14) 10. Be2 {here I was transfixed by the obvious threat to create double f-pawns, so retreated the bishop.} (10. O-O $5 {the engine has no such prejudices against doubled pawns and instead sees what other possible advantages White could get in compensation.} Bxf3 11. Re1+ {this is the key idea, using the open e-file to harass Black's king. Losing the right to castle and having the king in the center is more worrisome than the kingside pawn structure, although White should be careful.} Kd8 (11... Be7 12. Qxf3 Qxf3 13. gxf3 Kd8) 12. gxf3 Bd6 13. Bf1 Qh4 (13... Qg6+ 14. Bg2) 14. h3 $16) 10... Bd6 $11 {Black has an active position, notes Komodo via the Fritz interface. The position is equal, but I don't really have many prospects for progress.} 11. O-O O-O 12. h3 Bf5 13. Bd3 {looking to trade, as Black's bishop seems more active than my own.} (13. Bg5 $5 {is a better idea, as it develops my last piece with tempo.}) 13... h6 {basically a free move for Black, preventing the previous bishop development idea by seizing control of g5.} 14. Rb1 Rab8 15. Qc2 $2 {an example of lazy thinking and not following my thinking progress, which requires *always* to falsify your intended moves.} (15. Re1 b6 $11) 15... Bxh3 {my opponent spots the tactic, which is based on the overloaded g-pawn and the now not sufficiently protected Nf3.} 16. gxh3 $6 {here it would have been better to simply accept the loss of a pawn, rather than disrupting my pawn structure for no good reason.} (16. Re1 $15) 16... Qxf3 $17 17. Bf5 Ne7 18. Bg4 Qf6 {Black is simply a pawn up now and I still have no real threats.} 19. Rb5 {I start becoming desperately aggressive with my plans now, basically trying to force counterplay.} c6 20. Rh5 {unfortunately Black has too many pieces able to defend his kingside, for any sacrificial ideas to work on my part.} Qg6 (20... Nd5 $5) 21. Qb3 (21. Qxg6 $5 {is technically much better, but at the time I didn't think I had a real chance if the queens stayed on the board. In reality, it's worse for White, as we'll shortly see.} fxg6 22. Ra5 $17) 21... Nd5 22. Kh1 $4 {simply worsens the situation. I wanted to break the pin on the Bg4, was the original thinking.} (22. Rxd5 cxd5 23. Qxd5 $19) 22... Qe4+ {now Black has mate threats, thanks to the queen and knight combination.} 23. f3 (23. Kg1 {what else?} Nf4 24. Bxf4 Qxf4 25. Re5 $19) 23... Qe2 (23... Qe2 24. Rf2 Qxf2 25. Re5 Bxe5 26. dxe5 Qf1+ 27. Kh2 Qxc1 28. Qxd5 cxd5 29. h4 Qf4+ 30. Kg2 Rbc8 31. c4 Rxc4 32. Bf5 Qxf5 33. Kg3 Qf4+ 34. Kh3 Rc2 35. a4 Qxf3#) 0-1

04 November 2018

Video completed: The Fort Knox Variation in the French Defence

I recently completed "The Fort Knox Variation in the French Defence" by IM Lawrence Trent, which is a ChessBase 60-minute video download.  I am not in fact taking up the French for its own sake, but using this variation to make sure that if I want to play the Dutch Stonewall by transposition, I can't be tripped up by the move-order sequence 1. d4 e6 2. e4.  Now it's important to note that White on move 3 can avoid the Fort Knox by not playing either 3. Nd2 or 3. Nc3 (either of which allows 3...dxe4, leading to the Fort Knox after 4...Bd7).  However, the less frequently played French lines like the Advance and Exchange variations are not critical and I feel Black can get by at a reasonable level with just some basic familiarization.  This is especially true if you are a Caro-Kann player who responds to the Advance variation with 3...c5, which heads into French-type positions.

Because the Fort Knox is structurally very similar to the Caro-Kann, I believe going this route is easier for me - or at least is more efficient in the near term - in terms of building an opening repertoire without major holes.  The alternative would be having to look in-depth at the non-standard Dutch lines after 1. d4 f5 (although I may do that at some point in the future).


1.  Introduction - this is a very basic intro to the Fort Knox which covers the initial entry into the variation - White has pawns on e4/d4 and Black on e6/d5, then goes 3. Nc3 (or Nd2) dxe4 4. Nxe4 Bd7.  IM Trent then discusses Black's standard development plan (...Bd7 followed by ...Bc6 and usually ...Nd7, with the idea of exchanging the bishop for one of White's knights and then playing ...c6 for a solid pawn structure), and talks a bit about the solidity of the opening and some of its past practitioners.

2.  Variation - 5. c4.  IM Trent says that he likes to start with the sidelines and this is an interesting one, in which White gets the c-pawn forward before retreating the knight with Ne4-c3.  Black sticks with standard developing moves up until a novelty, which gets the bishop off c6 with the idea of following up with ...c5 and then playing actively.

3.  Variation - 5. Nf3 Bc6 6. Neg5 Bd6; IM Trent in the video labels it as "Tricky Neg5".  (Note: there is a typo in the video's table of contents which says "Nge5" instead of "Neg5", which is rather confusing and suggests more editing effort was needed by ChessBase).  This is perhaps White's only/best shot at major tactical complications, with the idea of sacrificing the g5 knight on f7 or e6.  Black can go horribly wrong if he ignores White's tactical threats, for example by playing 6...h6? instead of the recommended 6...Bd6, which covers the e5 square.  But with smart play by Black, White simply doesn't have enough compensation for any sacrificial attacks.  Some memorization is needed in these lines, although not a huge amount.  The main line ends up looking very much like a Caro-Kann.

4.  Variation - 5. Nf3 Bc6 6. Bd3 Nd7 7. Qe2 Ngf6 8. Neg5.  Like the above variation (and also with a table of contents typo), White tries to be a little tricky and aggressive.  However, Black now has time for the standard Fort Knox plan of developing with ...Nd7 followed by ...Ngf6.  IM Trent's recommendation - he doesn't give any alternatives - is to respond with 8...Qe7, which looks awkward but at the same time defends against all of White's sacrifice ideas on f7 and e6.  The queen can later on move off the e7 square to good effect, and/or Black can castle queenside, allowing the follow-up of ...g5 and developing the dark-square bishop to g7.

5.  Variation - 5. Nf3 Bc6 6. Bd3 Nd7 7. O-O [listed in table of contents as Qe2] Ngf6 8. Ned2.  Here White's idea is to reposition the knight to c4, hoping to dominate e5 that way.  IM Trent offers multiple possible ideas for Black in response, including ...Bb5, ...Bd5 - both of which allow for the ...c5 pawn break - or simply ...O-O.  Black is not afraid to exchange the light-squared bishop for a knight, or even in some variations to preserve it on the long diagonal after going ...b6 first.  GM Rustamov is observed to often play ...Bd5 in these situations.

6.  Variation - 5. Nf3 Bc6 6. Bd3 Nd7 7. O-O [listed in table of contents as Qe2] Ngf6 8. Ng3.  This is in fact White's main line, according to IM Trent.  He presents the idea of avoiding the standard plan of an immediate ...Bxf3 reaction by Black, instead using GM Jobava's modern move 8...g6!? which blunts the White light-square bishop, offers up g7 for Black's dark-square bishop development, and takes away the idea of Nh5 for White as a future attacking move.  There is a lot of relatively new territory to explore, but IM Trent shows various plausible scenarios, including the later utility of the ...Bxf3 idea. White's 9. c4 is the most challenging response for White, who sacrifices a pawn on d5 in order to force Black's king to f8, but Black according to Trent should be OK, as it's White who has to prove compensation for the material.

And that's all there is to the opening!  The video concludes with a quiz section of five puzzles, taken from GM-level tournament games, that help show what kinds of positions you can end up with in the Fort Knox.  They're drawn from all stages of the game and include some effective ideas, for example multiple examples of exchange sacrifices.  They also demonstrate that the Fort Knox, while an equal opening, is certainly not devoid of tactical possibilities.

General comments:
  • The Fort Knox is not intended for aggressive counterplay by Black.  IM Trent emphasizes key concepts such as opportunities for multiple exchanges on e4 and the general idea of exchanging off White's active pieces, to arrive at a position where White has nothing.  Black however also has nothing in terms of obvious winning chances in most lines, so it's important to be aware of this.  I think the Fort Knox could be used to win if your opponent is the type who is too aggressive and typically over-presses in equal positions.
  • Each of the variations covered feature a direct link to the analysis in game format, so you can review them directly rather than having to go through the video again.
  • IM Trent usefully cites reference games in his narrative and encourages you to go through the full games, which is always a good idea.  He points you toward top players like GM Rustamov, who has played it frequently, and others like Gelfand and Karpov who have played key games.
  • The typos and misleading text in the video's table of contents (the "text' field of the database) are annoying but not a critical flaw.

03 November 2018

Annotated Game #200: A ghastly little game

This ghastly little game teaches a few things, so perhaps it wasn't a total waste of time.  Firstly, the strategic error of 4...Bg4?!? is not in itself losing, or even bad according to an engine, but it does not fit at all with what Black really needs to be doing in the position (4...e5).  These types of early strategic errors often result in a very narrow path to equality, which is easy to fall off of and into a worse position which slowly (or more rapidly here) leads to a loss.  The other major strategic error, 7...Qxd1, is also useful to illustrate how flawed is the idea of always going for piece exchanges (including the queen) against a much higher-rated opponent.  Just because material is off the board does not mean you are any safer, and in fact can simply heighten your opponent's advantage.  Don't be afraid to keep material on the board and maneuver, in other words, rather than simplify into a lost position.  You can see one of the earlier games on this blog, my simul against GM Yermolinsky (Annotated Game #4), as another good example of this.

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "Expert"]
[Black "ChessAdmin"]
[Result "1-0"]
[ECO "D10"]
[Annotator "ChessAdmin/Komodo 11.2"]
[PlyCount "33"]

{D10: Slav Defence: cxd5 (without early Nf3) and 3 Nc3} 1. d4 d5 2. c4 c6 3.
Nc3 dxc4 {an aggressive continuation unique to the Slav.} 4. a4 Bg4 {the best
that can be said about this move is that it's creative.} (4... e5 {striking in
the center is more to the point, exploiting the now-open d-file for Black.}) 5.
f3 {I confess I had been hoping for a move like this in reaction, which causes
problems for White's development.} Bh5 {a logical retreat, but it doesn't do
anything for Black's development.} (5... Bd7 $5 {with the idea of} 6. e4 {
and now} b5) 6. e4 e5 $146 {the right idea, just played a little late.} 7. dxe5
Qxd1+ $2 {a strategic blunder. I now have a much more difficult time dealing
with White's pawns, while White's king position is effectively no longer weak.
I believe I let myself be dominated by the idea of 'against stronger opponents,
trade down quickly'} (7... Nd7 {should be equal. For example:} 8. e6 (8. g4 $6
Qh4+ 9. Ke2 Bg6 10. Nh3 Qe7 $15) 8... fxe6 9. Bxc4 Qh4+ $11) 8. Kxd1 Nd7 {
again with the right idea, played later than it should have been.} 9. g4 Bg6 {
although Komodo only gives a small edge to White, it's easy to see how White's
space advantage makes it much easier for him to play.} 10. f4 $14 h6 $2 {
unfortunately the correct defense is moving the h-pawn two squares forward,
not one. I was concerned about providing a haven for the bishop on h7, but
also should have recognized the need to break up White's kingside formation,
which now rolls forward and crushes me.} (10... h5 11. Bxc4 (11. f5 {is now
less effective:} Bh7 12. gxh5 $6 (12. g5 Nxe5 $11) 12... Nxe5) 11... hxg4 $14)
11. Bxc4 $16 {the most straightforward winning continuation. I'm now down a
pawn with no compensation and still being squeezed.} Bb4 12. Nge2 O-O-O 13. Kc2
$18 {now White's king is out of danger and he can fully mobilize his forces.}
f6 $6 {a desperate move that hastens my downfall.} (13... h5 14. g5 Ne7 15. Rf1
$18) 14. f5 {good enough to win.} (14. e6 $5) 14... Nxe5 {again accelerating
the loss, but I was done for anyway.} (14... Be8 15. exf6 Ngxf6 16. Be6 $18)
15. Be6+ Kc7 16. fxg6 Nxg4 {why not? Basically wishing that White would miss
the backwards bishop move, with no other hope.} 17. Bxg4 {a quick end to a
ridiculous game.} 1-0