The Chess Mind

Referring to Reuben Fine, of course. In the game Roussel Roozmon-Bluvshtein (presented in this post), we saw that sometimes two knights can hold an open position against the bishop pair, even at the cost of a pawn.

True enough, and sometimes the man bites the dog. Normally, though, things run the other way around, and it's more common to find the bishops generating problems the knights just can't handle. That's just what we find in this historically interesting game between former world champion Jose Capablanca and Reuben Fine. You won't find it in your databases, because it was played in a simul, but it's worth a look!

Capablanca, one of the all-time greats and a legend in the endgame, begins a two knights vs. two bishops ending with the knights and an extra pawn, and presumably heading for another routine win against the simulee. But Fine does not cooperate, and when Capa drifts for a couple of moves, his opponent completely takes over the initiative. Fine does an excellent job of revealing just how helpless knights can be against bishops when the center is open and they lack support squares.

The game can be replayed here; the notes are mine, except when Fine is explicitly mentioned (the game and his notes are from his Lessons From My Games (New York: David McKay, 1958)).

Here's the problem, presented here a few days ago.



White to move and draw.

This study, by 1999 FIDE World Championship runner-up Vladimir Akopian, is pretty challenging, but if you think you've figured it out, click here.

In a comment to my post Three Games and a Study, supergrobi refers to the Lucena, Philidor, and Karstedt positions as examples of basic rook endings. Now, I've heard of the first two, and I certainly agree that they are absolutely fundamental endings; indeed, they are always the first two rook endings I teach my students.

But what is the Karstedt position? I didn't recognize the name, so I fired up ChessBase 9, opened the Study Database and did a search for Karstedt. I found several studies I had seen before, significant positions (though not as important as the other two mentioned in the title) - but I wonder which of them supergrobi had in mind.

Therefore, endgame fans, all the candidates are attached! Each of the positions is valuable in its own right, and hopefully supergrobi will let us know which one he had in mind. (I'm guessing it's the fourth of the five.)

Click here to see them - happy learning!

The British Championship finished yesterday, and Scottish GM Jonathan Rowson retained his title with 8.5/11, half a point ahead of GM Stuart Conquest and IM Stewart Haslinger (both of England). (Adams, Short and Hodgson didn't play.)

A week or so ago I presented the miniature Richard Pert-Chris Ward (incidentally, despite the embarrassing loss, Pert rebounded to tie for 4th-5th with GM John Emms at 7.5/11, while Ward finished a point further behind in a tie for 8th-12th); today, two more substantial games.

The first, ironically, features another Pert loss, to talented young FM Stephen Gordon (who tied for 6th-7th with FM Andrew Greet with 7/11). The game features the sharp 8.Rb1 Gruenfeld line Gelfand calls his favorite (recall the Gelfand-Shirov game presented with my review of Gelfand's new book). Black normally has a passed a-pawn that can at times be quite dangerous, but Gordon allowed Black to retain his b-pawn as well in order to facilitate his own attacking chances. I don't know the theory of the 8.Rb1 Gruenfeld well enough to know how transportable Gordon's idea is to similar positions, or even if it was best in the game itself. Nevertheless, on at least this occasion, his dynamic approach worked to perfection.

The second game is an endgame win by Emms over Greet. White parlays a space advantage into the full point, but only after Black makes a mistaken judgment about which pieces to exchange and which not. I think there are some useful lessons to be learned there: the value of space, the relative strength of bishop and knight relative to each other and in conjunction with space advantages, and, finally, the game concludes with some instructive king and pawn ending technique.

They are fine games in their own very different ways, and they can be replayed here.

Our installment for this fine morning/late night includes some recent games that caught my eye for their instructional value, rather than for their competitive significance. (Not that the players and their most devoted fans would agree, of course.)

The first game, between Roussel Roozmon and Bluvshtein, shows a nifty defensive idea that conclusively saved Black in a two knights vs. two bishops ending. Generally the knights have a lousy time of it, but Bluvshtein alertly spotted a powerful drawing idea, one we can easily apply if we're paying attention.

Next, Kempinski is in trouble against van Wely in a rook ending, but utilizing an idea going back to the study composer Moravec, he finds an elegant path to the draw. (The Moravec study is presented afterwards.)

Finally, a sharp Chinese Dragon (that's redundant, I know) between Efimenko and Charbonneau reminds us that it's possible to sacrifice too many pieces. Charbonneau seems to unload every stock Dragon sac in the book, only to find that when the payoff comes, there aren't enough Black forces left to celebrate. Nevertheless, his basic attacking idea was correct and his position after the opening appears sound. So Dragon players, take heart!

Click here to replay the games and the Moravec study.

Up for a challenge? Here's an endgame study by 1999 FIDE World Championship runner-up Vladimir Akopian:



White to move and draw.

I suspect that almost every player who has looked at a beginner's book covering endgames has seen this position:



What's remarkable is that although White has no passed or even unopposed pawns, and the Black pawn structure has no weaknesses, White can nonetheless win with 1.b6! axb6 2.c6! bxc6 3.a6 (or 1...cxb6 2.a6! bxa6 3.c6). It seems like alchemy, but incredibly, it works!

Note that with Black to move in the starting position, only 1...b6! will save him - 1...a6? 2.c6! bxc6 3.b6 (or 1...c6? 2.a6! bxa6 3.b6) wins, reminiscent of the tactical trick in the previous post. After 1...b6!, none of White's tactical tricks are possible, so it comes down to the placement of the kings, which in this case ensures a win for Black.

A second note: while 1...a6 and 1...c6 both lose, with Black to move, if Black could play both moves, then he would be fine, even if we gave White an additional a, b, or c-pawn! (Unless one puts the pawn on a7 or c7, that is, but we're ignoring bughouse-ish possibilities here.) One application of this comes in the Exchange Variation of the Ruy Lopez, when after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.O-O f6 6.d4 exd4 7.Nxd4 c5 8.Nb3 Qxd1 9.Rxd1 etc. White's endgame fantasy is to trade off all the pieces and head for a pawn ending. White's 4-3 kingside majority is healthy and will result in a passed e-pawn, but Black's queenside majority can be stymied by placing pawns on a3, b2, and c3, and then only recapturing Black queenside pawns, never initiating the capture.

With all this by way of background, consider the following position, also, ironically, from another 1-minute game played the same night as the previous post's game:



My queenside pawn structure is healthier than Black's, but the most important feature of the position is the Black king's encroachment on my side of the board. If it were Black's move, we might see 1...Kh2 followed by 2...Bc5 3.Be1 (3.Bxc5 volunteers for a hopeless pawn ending after 3...bxc5, 4...Kh3 followed by ...Kg4xf4) Kg1 followed by ...Kf1. Maybe I can keep him from penetrating all the way to my queenside, but that sort of worry motivated my move in the game:

1.Bg3?

Maybe I'm losing anyway, but there's no question about it after his next move -

1...Bh4!

now I'm completely lost! After the forced

2.Bxh4 Kxh4

all my opponent needs to do is place me in zugzwang, so that my king has to allow ...Kg4, and then allow ...Kxf4. In part to delay (it is 1-minute chess!), and in part hoping for some sort of happy accident, I continued

3.b4



Here the hideous 3...a5?? loses to 4.c5! (again, remember the previous post), but 3...c5 and 3...b5 essentially terminate resistance. My opponent's choice didn't lose, but kept a spark of hope alive:

3...a6 4.a4



I trust everyone sees where this is going. 4...a5?? and 4...c5?? are horrible, due to 5.c5 and 5.a5, respectively, but Black's best move is 4...b5!, when, objectively speaking, resignation is required. Perhaps my opponent thought, in the constant time pressure that is 1-minute chess, that it was best to avoid pawn contact over there unless absolutely necessary, and decided to mark time with his king. Besides, the breakthrough strategy of the first diagram only works when White's pawns start a rank further up, right?

4...Kh3?? 5.b5 axb5 6.c5 bxc5 7.a5 b4 8.a6 b3 9.a7 b2 10.a8Q b1Q



We've both queened and Black has two extra pawns, but there's a little problem:

11.Qh8#

The moral of this story is that the trick from the first diagram can be effective even when the first move is 1.b5 rather than 1.b6, depending on the placement of the remaining material, especially the kings. Apply what you know!

Consider the following position:



The total number of pawns is even, but White is winning easily due to his outside passed pawn. White will start with 1.b5, using the b-pawn as a decoy. Black may garner the b-pawn, but White will help himself to as much of the Black kingside as he desires.

Next, a marginally more complicated example:



This works in the same way as the previous example, except that White first needs to convert his queenside majority into a passer. Obviously enough, we start with 1.b4 (as a general rule, first push the pawn on the unopposed file, and only push the pawn on the opposed file when its comrade needs assistance), and after 1...Kc6 2.a4 Kd6 3.b5 axb5 4.axb5 we reach the position after 1.b5 in the first example. It's a basic technique and indispensable, too.

But be careful - it doesn't always work so simply! Take this position from a recent 1-minute game:



My opponent (playing Black) has an edge, but the position is objectively drawn (disregarding the time situation, which wasn't at all a factor at this point). My pawn on f3 is a little weak but easily defensible for the foreseeable future. Black instead tries the thematic route, attempting with 1...a5 to create an outside passed pawn after 2.bxa5 bxa5 3.Kc3 Kc5, when Black is probably winning.

White can easily improve with 2.Kc3, when neither 2...a4 nor 2...axb4+ promise Black anything at all. Black's best is probably to stay put, when the game will either be drawn or (in light of the time control) degenerate into a mouse-racing contest.

Instead of 2.bxa5? or 2.Kc3, however, I chose a third option and a favorite trick from pure king and pawn endings: 2.c5+!. Sometimes it's even a winning idea, but thanks to his king's useful placement on the queenside, 2...Kc7 allows him to draw with ease. Unfortunately, he was probably surprised by my move and reacted spontaneously with the obvious but bad 2...bxc5?, when 3.bxa5 brought about a complete reversal: now White has the outside passer and all the winning chances.



I wouldn't say that Black is lost yet, but the game ended in the quick, routine way one would expect from 1-minute chess:

3...c4?! 4.Be2 Kc5 5.Kc3 Kb5 6.a6 Kxa6 7.Bxc4+ Bxc4 8.Kxc4 Kb6 9.Kd5 Kc7 10.Ke6 1-0

A lightweight example, admittedly, but the tactical trick exemplified by 2.c5+ is one to remember and include in your endgame toolbox.

On Monday, I offered this peculiar position



and challenged the reader to find a breakthrough, if any. It's subtle, and even your chess software probably won't get it (mine doesn't even come close), but it's there! (Congrats to sbb1cpa for solving it.)

Click here for the replayable solution.

Black to play; can he break through?