Monday, February 23, 2009

The Tortoise and the ...

Artwork by illustrator Stephanie Smith.

I played in another Open event this weekend and scored a USCF 1500 performance rating. That's solid, but nothing spectacular.

The following game was typical of my battles. I made no simple tactical mistakes, but I consistently used more clock time and my position always seemed to be slightly worse than my opponent's. In two of these battles my opponents found exceptional tactical shots to finish me off. In one game I returned the favor. The last game was a draw.


Below is the critical position from another game, likesforests-careyfan:

(a) What move gives White an advantage?
(b) White actually played 21.Na4. What move wins for Black?


likesforests said...

a) 21.Nxe4 Bxe4 22.Bxe4 23.dxe4 Qxc4 so 21...Qg6! and now: 22.Bxf3 (22.Rfc1!? Bxg2 23.Kxg2 Nxe3+ 24.fxe3 dxe4 25.Qxc6=) 22...Rxf3 23.Rac1 (23.Rfc1?! Nxe3 24.Qxc6 Qxe4 25.Qxa8+ Rf8 26.Qxf8+ Kxf8 27.fxe3 Qxe3+ 28.Kg2 Qxd4 -/+) 23...Nxa3 24.Qa2+/- (24.Qa4?!=)

b) 1.Na4? Bxg2! 2.Kxg2 Qxe3! and 3.fxe3 is met by 3...Nxe3+, so in the actual game I played 3.Qc3 Qf3 4.Qxf3 Rxf3 instead but now the a3 pawn also falls and Black had a huge lead. Note, 3.Qxc4?? fails due to 3...Qf3+.

Anonymous said...

6.e3 Rubinstein's approach is to fianchetto the KB, so that it bears down on d5.

7.Bd3 Doesn't seem quite right - the queen needs clear sight of d4. In fact later you move the B to e2 anyway.

16.- I don't see that you have realistic attacking chances.

19...Qxd1 Instead, black might grab a pawn for his trouble (of exchanging queens) by 19. ... Bxg3 20.Qxh5 Bxf2+ 21.Kxf2 Nxh5.

20.- It's difficult to see how you can successfully attack d5. The lone bishop roughly offsets the isolani.

25.f4 Seems unnecessary to hand over the e4 square. Perhaps instead gradually improve your position by f2-f3, Kf2, etc.

26.Nxd4 Doesn't this lose a pawn to 26 ... Rc3?

28.- The position should be easier to draw if you keep rooks but exchange knights. Rd3 to protect the pawns.

33 ...Nxf2 Instead of this, Nd2-c4 looks good for black.

35- It's not fatal yet. In this situation, count the number of moves needed for Black to promote, versus for you. This tells you your king needs to go west.

36- Ke2 and you can just barely save the a pawn, no?

Have a look at 36...a4 37 e5 fxe5 38 fxe5 axb4 39 axb4 Kd7 Black will improve his pawns, push the d pawn (forcing white to capture) and then Black will win the foot race.

I wouldn't call 38 g5 a mistake - you have to throw the kitchen sink at this point, and you ended up to hit him!

likesforests said...

6.e3/7.Bd3 - Thanks, this is definitely one of the gaps between me and a strong positional player. I played 6.e3 "to support the pawn" but never intended exd4, and while I did consider Bb5 and reject it because I wanted an IQP position, I chose d3 over e2 because it was "more aggressive" than whatever that means. g3/Bg3 or even e3/Be2 are clearly better. Lazy thinking.

16.Nc2 - I meant, I was keeping an eye on my OPPONENT'S attacking chances. On moves 12-16 my opponent had four pieces trained on my kingside. That is enough that I had to calculate out the possibilities of a sacrifice, which ran down my clock a bit.

19...Bxg3! - It's a better move, but I considered it and had no intention of giving up a pawn. 19...Bxg3 20.hxg3 Ng4 21.Nf3 and White holds. (If Rd6-h6 then there's Nh4, too.)

25.f4 - You're right, I didn't even think about the strong outpost this would create. I saw 'all' my pieces were in a bind and lashed out, but f3, Kf2, Rd2 and my position is slowly improving.

"26.Nxd4 Doesn't this lose a pawn to 26 ... Rc3?" - It does, but I thought White gets counter-chances due to back-rank mate threats--27.Kf2 Rxa3 28.Rc1 Ne4+ 29.Kf3 h6 30.Rc8+ Kh7 and 31.Rd8 or 31.Rc7. But I missed that Black has a neat defense in 29...Nd2+ 30.Kf2 Nc4! No idea why I fixated so much on this. Wasted time. 26.Rxd4 was simpler, safer, and faster.

"36- Ke2 and you can just barely save the a pawn, no?"

Nah, one move too late. It takes White 4 tempi to protect it and Black only 3 tempi to win it. 35.Ke2 is safe.

"I wouldn't call 38 g5 a mistake - you have to throw the kitchen sink at this point, and you ended up to hit him!"

This is where computers are amazing! 38.Kd5, 38.Kf5, and 38.h4 draw with best play while 38.g5 loses. In fact, 38.h4 wins if Black is inaccurate. I think with 2-3 minutes to calculate it's possible to see g5 loses and with 5 minutes I could find a draw. But only my opponent had such time, and he was too impatient to use it. :)

likesforests said...

I believe I'm not doing enough backwards thinking, "Where do I want to be a few moves from now." I'm almost always using forward thinking and that helps more with tactics than with positional considerations.

chesstiger said...

At first i thought that the solution of your second question was wrong since white can play 3. Qxc4 but after closer inspection i saw that black then can respond with Rxf2+ and then it forms rook and knight against queen which i think in the given position is beter for the black pieces, but i can be wrong.

likesforests said...

chesstiger, the problem with the desperado move 3.Qxc4?? is even worse than that. 3...Qf3+! 4.K(any) dxc4 and White loses his queen for a knight.

Aziridine said...

Anonymous makes very good points. I'll add more into the mix:

10.a3 and 11.b4 is ambitious and maybe not a mistake yet but the weakness of c4 makes itself felt later in the game. Probably you shouldn't have followed up with 13.Nd4 and traded the bishop that guards the light squares.

16.Nc2 looks terribly awkward. I wonder if 16...Neg4 17.h3 (17.g3 Ne4) 17...Nh2!? 18.Re1 Nf3+!? 19.gxf3 Qxh3 works. At a rough glance it looks scary for White (this is a lot of fun to calculuate for Black!):
20.Bxf6 Bh2+;
20.f4 Ng4;
20.Nf4 Bxf4 21.exf4 Nh5 and White has to deal with 22...Nxf4 and 22...Rc6;
20.Ng3 Bxg3 21.fxg3 Qxg3+ 22.Kh1 and does Black have something like 22...Ne4 here? Or 22.Kf1 Qh3+ 23. Kg1 Ne4? Maybe I'm just hallucinating (it's late here.)

Are you sure 19...Bxg3! 20.hxg3 Ng4 21.Nf3 is the end of the story? 21...Rc6 and if you meet ...Rh6 with Nh4 I'll have ...g5, won't I?

I think I'd even prefer to be Black in the endgame. 21...Ne4! is much stronger. I'm not sure what White can do about ...Nc3, e.g. 22.Rac1 Rfc8. The weaknesses left behind by White's queenside pawns loom large. Piece activity can also be an advantage in the endgame!

25.f3! takes advantage of Black's ...Ng4-Nf6. Is it time for Black to steer for a draw with 25...a5?

26...a6 is another waste of time (26...Kf8 right away). How does Black meet 28.Rd3 followed by 29.Nc3?

In the knight endgame White's pawn structure is no better than Black's. 33...Nd2 followed by 34...Nc4 would have emphasized this.

If it seems strange I'm asking so many questions it's because I don't know the answers. There were many interesting moments in this game.

likesforests said...

"Are you sure 19...Bxg3! 20.hxg3 Ng4 21.Nf3 is the end of the story? 21...Rc6 and if you meet ...Rh6 with Nh4 I'll have ...g5, won't I?"

19.Bxg3 hxg3 20.hxg3 Ng4 21.Nf3 Rc6 22.Qd4 Rh6 23.Nh4 g5 24.f3 gxh4 25.fxg4 Qg6 Nxe3! (desperado) 25.Qxe3 gxh4 26.g4 and Black is up a pawn. If he misses the desperado, White is fine.


(1) I didn't calculate it out this far in the game. I should try to calculate my moves to quiesence.

(2) There's a thin line between making a 'prophylactic move' and making an 'unnecessary defensive move'. In this case 19.h3 which meets 19...Ng4 with 20.Nf3 was justified not only because it saves the pawn but also because it makes the calculations simple which in turn saves vital clock time.

"Probably you shouldn't have followed up with 13.Nd4"

Fritz thinks 13.h3 Bxf3 14.Bxf3 is equal and 13.Nd4 is worse. The real question is, how would a human see this? I mean, with the knight gone the d5 pawn is going to mobilize soon. That's the dream for a player with an IQP and the point of ...Bg4.

How does Black meet 28.Rd3 followed by 29.Nc3?

Oh! It seems in this endgame I had some opportunities. 28.Rd3 Ke7 29.Nc3 Kd5 survives immediately but White can keep adding pressure with g2-g4, Kg2-Kf3, and h3-h4 + g4-g5.

"If it seems strange I'm asking so many questions it's because I don't know the answers."

Much appreciated. :)

Anonymous said...

Hey guys, this is a good discussion. I'm definitely learning something.

About the b2 pawn, my feeling is you want to keep it on b2 to support a knight on c3, which in turn attacks d5 (until there is a concrete gain.) I noticed that Black did not have to work too hard to defend the isolated pawn, and hence had more piece play.

tanc (happyhippo) said...

hi likesforests:

thank you for responding to my question on chessbase on greg's blog.

much appreciate them.

as with regards to your blog post and the comments i'm following..... all i can say is "wow".

that is some heavy duty analysis by some of the readers here. it has definitely opened my eyes!

great stuff!!! :)

Antonio Mendoza said...

Hey Likes, May I ask how you lost 40+ pounds? Did you go low carb, exercise like a maniac, combine deep breathing with visualization? I have five winter pounds I need to lose before bikini season. Thank you.

likesforests said...

Antonio, nothing so extreme, I kept a food journal. Believe it or not, people who keep a food journal tend to lose twice as much as people who do not. For the first six weeks I sent it to a PHd and nutritionist and we talked about what was working and what wasn't.

Aziridine said...

I wouldn't have seen the entire line after 19...Bxg3 either, but I would have appreciated that the threat of ...Rc6-h6 is difficult to meet and wouldn't have gone for that line as White.
If 13.h3 I would definitely play 13...Bh5 instead of 13...Bxf3. I don't see any reason for Black to cede the bishop pair and allow White to put the light-squared bishop on an active square. The position resulting from ...Bxf3 and ...d4 is at best dead equal - the pawn structure will be basically symmetrical. And Black really should be looking for an advantage, not equality.
By the way, I think the best move is simply 13.Bb2! creating the threat of 14.Nb5 and highlighting the flaw of 12...Qd7.
I don't have access to a program, so I'm still curious if my 16...Neg4 idea actually works :)

Ray Cheng said...


I hope you don't mind - I just wanted to brag on my first blindfold win.* As you know I've been practicing blind chess against LudoChess (a free java applet) for several weeks, about one game a day. I'm at the point where I can complete a game. Until now I've had one draw and all the rest were losses. [*There are three provisos: (1) I took notation, so I (in principle) had a memory aid; (2) I looked at a blank chess board while playing; (3) On move 24 I reviewed the position visually after a long break.]

It doesn't feel any easier, but I guess I'm making progress. I have no sense that this is improving my OTB chess (which is really the goal), but maybe that will come.



RC - LudoChess Jester
Blindfold Game, 25.02.2009

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 At this point I am out of book! I'm playing non-repertoire openings with this program in order to rely less on familiarity. 4.d3 Be7 5.0–0 0–0 6.c3 d6 7.a3 [Now that e5 is defended by a pawn, Black can play Na5. In that case, I plan to slide the bishop to a2, so it continues to nibble on f7.] 7...Na5 8.Ba2 Bg4 9.Nbd2 Nc6 10.h3 Bh5 11.Re1 [I have to fortify e4 in order to prepare d3-d4.] 11...a6 12.Nf1 h6 [The computer is out of ideas! But I have one: with Black having played h7-h6, I plan to post a knight on f5. Then g7-g6 would be met by Nxh6.] 13.Ng3 Bg6 14.Nh4 Bh7 15.Nhf5 Bg6 16.Be3 Kh8 17.f3 [Perhaps too overprotective of e4. Black cannot stop d3-d4 in any case. For example, 17.Qd2 (threatening Bxh6) 17...Ng8 18.d4] 17...Qd7 18.Bf2 [A wasted tempo. I should have played d3-d4 immediately.] 18...Bd8 Sure, it can keep its bishop pair, if it doesn't mind crappy squares. But of course I wouldn't give up the monster on f5 for a mere bishop. 19.Qd2 [Now Nxh6 is in the air. I check this each move.] 19...Nh7 20.d4 Bg5 21.Be3 exd4 22.cxd4 Bxe3+ 23.Qxe3 Nf6 24.Rad1 Rae8 [I had to go away for a while, and allowed myself a quick glance at the position before resuming blindfolded.] 25.Qf4 d5? [Ha! Now g7 is overworked, and I cash in my positional chips. Maybe instead Black should try 25...Nh7 to take some sting out of e4-e5.] 26.Nxh6! dxe4 27.fxe4 Nh5 [My heart stops for a moment.] 28.Nxh5 Bxh5 29.Nf5 [I've won a pawn, and I have a strong center. This move also sets a trap. Yet, my fancy screensaver prefers 29.Qh4! gxh6 30.Qxh5 Qd6 31.Bxf7+- White is up two pawns and has a safer king.] 29...Bxd1 [Falling for it. But even 29...Bg6 30.Nh4 Bh7 31.Nf3 is great for White.] 30.Qg5! Qxf5 [30...g6 31.Qf6+ Kg8 32.Qg7#; 30...f6 31.Qh4+ leads to mate - remember the bishop on a2 covers f8; 30...Rg8 31.Qh4+ with mate to follow.] 31.Qxf5 Nxd4 32.Qf2 [My silicon assistant points out 32.Qc5! Bb3 33.Bxb3 Nxb3 34.Qb4 trapping the knight] 32...Bb3 33.Bxb3 Nxb3 34.Qc2 Nd4 35.Qxc7 Rb8 [Both enemy rooks are passive, but mine will reach the 7th rank.] 36.Rd1 Ne6 37.Qe7 Kg8 38.Rd7 b6 39.b4 [Now the threat is Ra7.] 39...a5 40.Rb7 axb4 41.axb4 b5 42.Qd7 Rxb7 43.Qxb7 Nd4 44.Qd5 Ne6 45.Qxb5 Rd8 46.Qb7 Rd1+ 47.Kh2 Kf8 48.b5 Nd8 49.Qc8 Ke7 50.b6 Rh1+ 51.Kxh1 At this point the program freezes, which I take to mean resignation. 1–0

likesforests said...

Wow!! Being able to play that well blindfolded is quite an accomplishment, you definitely have bragging rights.

likesforests said...

"I have no sense that this is improving my OTB chess..."

Do you feel you can visualize accurately at a greater depth than before? Eg, you can see 12-ply ahead without missing an open lines or tactical possibility?

Aziridine said...

Hmm, after looking at it more closely, 16...Neg4 is just garbage. 16...Ne4 would've kept a big advantage for Black though.
19.h3 is possible but I think 19...Ne4 would be the strongest answer. 19.Nf4 might be safest. Anyway, interesting game.

Anonymous said...

"Do you feel you can visualize accurately at a greater depth than before?"

At this very early stage I don't really notice anything different, except that it is a bit easier to remember where the pieces are. Other than that it seems I'm only trying harder and taking more time. It is a particular struggle to conjure up any kind of global working visual representation of the position. I suspect that will improve with continued effort.


Will said...


I wonder whether a global impression is really possible since I am sure I read that the board is remembered in chunks.

If you ever decide to start blogging, please let me know I'd be really interested to read you thoughts on chess etc.


Anonymous said...


Thank you for your kind words. My opinion of chess? It sucks and it's screwing up my head! Anyway, here is another (mostly) blindfold win. I played sighted from about move 33 onward, since I was running out of time, and had to see how the game turned out. [N.B. I have not been sharing the dozens and dozens of losses.]


LudoChess Jester- RC
February 27, 2009
Irregular Opening

1.e4 d6 2.d4 Nf6 3.Nc3 e5 4.dxe5 dxe5 5.Qxd8+ Kxd8 6.Bg5 Be6 7.0–0–0+ Nbd7 8.Nf3 Kc8 9.Bxf6 gxf6 10.Bb5 c6 11.Ba4 Nc5 12.Bb3 Nxb3+ 13.axb3 Bc5 14.Rd2 b5 15.Nh4 Rg8 16.Rhd1 Kc7 17.h3 a5 18.Kb1 a4 19.bxa4 bxa4 20.f3 Ra7 21.g4 Rga8 22.Rd3 a3 23.b3 a2+ 24.Ka1 Bd4 25.Rc1 Ra6 26.Nf5 Bxf5 27.gxf5 Ra5 28.f4 Rc5 29.Rxd4 exd4 30.Na4 Rca5 31.Rd1 c5 32.Kxa2 Re8 33.Re1 c4 34.Ka3 Rxf5 35.exf5 Rxe1 36.bxc4 Re3+ 37.Kb4 Rxh3 38.Kc5 Ra3 39.Nb2 h5 40.Kxd4 h4 41.Kc5 h3 42.Nd3 Rxd3 43.cxd3 h2 44.Kd4 h1Q 0–1

Anonymous said...

Also -

"I wonder whether a global impression is really possible since I am sure I read that the board is remembered in chunks."

GM Jonathan Tisdall wrote that when he was an active player, it didn't matter in the slightest whether he had a chess set in front of him - his memory furnished an equivalent working representation of the position. That's what I am referring to. I don't perceive all of the possible captures, open vs. obstructed lines, etc. instantaneously. It's painstakingly slow, but getting a tiny bit easier with practice.


Will said...

I suspect however that GM Tisdall had a huge number of chunks in his memory from playing thousands of games since he was a child. I have often seen pictures of GM's starring into space while considering a move and think that this ability is linked to the patterns and chunks in their memory. It is an interesting testing ground for the idea of memorising vs blindfold playing to enhance visualisation ability if we had two guinea pigs so to speak.

Anonymous said...


You might enjoy the recent dissertation by Roy W. Roring. It contains a nice summary of theories of chess skill. A discussion of chunking begins on page 16.


Timothée Tournier said...

21.Na4 ? Bxg2! 22.Kxg2 Qxe3! 23.fxe3 Nxe3+ and Black is easily winning
if 23.Qxc4? then the intermediate move 23...Qf3+!.