For me ,when solution to Saturday <POTD> as usual a long ,timid line this shot was not expected.

When you are getting accustomed to something, it is not easy to see the real picture. I'd come to Bb8 only after 17.Qxa8 Qxa8 18.Rb8+ Qxb8 19.Bxb8 and have not too much.

some of us were lucky to grow up with English descriptive and so it's a doddle. :o)

Definitely worth learning because there are many classic e.g. Batsford titles which are out of print but can be bought on the cheap from 2nd hand stores.

I think Ray Keene was one of the people who made a campaign for, what he called the more logical, algebraic in Britain in the 1970s which then became standard in all magazines and books from the 1980s. Of course, it is now a FIDE rule that games must be scored in algebraic. (Someone who famously scribbled down the game score in descriptive notation was Bobby Fischer.)

Because white delays Nc3. That means that black can't play Bb4 to pin that knight.

The key move is white's 3. Nf3

click for larger view

White leaves the Nb1 at home for the time being to take black out of his opening preparation.

Now, 3...Bb4+ would be a Bogo-indian and 3...b6 would be a queens indian, but black chose to strike back in the centre with d5 instead.

There are lots of transpositional possibilities which mean that it is difficult to be precise about the naming of openings. This game could also be called a Queens Gambit, a Tarrasch or, as you rightly say, a Catalan. CG have chosen to call it an anti nimzo, which is a little unusual but reasonably accurate.

Material is even. White has firm control on the b-file, while Black exerts pressure on the d-file.

There seems to be nothing in:

17.Qxa8 Qxa8
18.Rb8+ Qxb8
19.Bxb8 (19. ...a5, 19. ...Ra4, or 19. ...Rd2)

I also can't see any advantages for White in:

17.Rfc1 Bd6
18.Bxd6 Rxd6
19.Rc7 Rd1+
20.Rxd1 Qxd1+
21.Kg2 Rf8
22.Qxa7 Qxe2

So I think the best move is:

17.Bb8 [removing the guard of the a8 rook]
17. ...Rxb8
18.Qxb8 [White has won the "Exchange"]

So B-KN5 is bishop to king's knight 5, which we would know as Bg5.

Captures are <name of capturing piece> x <name of captured piece>.

So PxP is pawn takes pawn, and BxKt or BxN is bishop takes knight.

Things get a little complicated if more than one piece can do something. Take this example:

click for larger view

In this position, white played 3. N(b)-d2. Because both knights can get to d2, in descriptive this would be written as 3. QN-Q2. In other words, Queen's knight to queen 2.

This can get confusing later in the game when you have to remember which knight is the king's knight and which is the queen's knight, and the same with the rooks.

In this position ...

click for larger view

If white played 4. cd (also writted as cxd5), the descriptive would be 4. BPxP. And 3. dc would be 4. QPxP.

And to get really silly, in this position...

click for larger view

5. cb would have to be written as QBPxNP. In other words, Queens bishop pawn x knight pawn.

Because there are several pawn captures, we can't simply say 4. PxP. That would be ambiguous.

And because both bishop pawns can capture a pawn, we can't say 4. BPxP.

And because the queens bishop pawn can capture two pawns, we can't just say 4. QBPxP. Hence 4. QBPxNP.

I think that's why descriptive died out! But if you can understand all that you can read just about any chess book.

Hiding in church street is a new chess and games shop, where I parted with some earth pounds in return for an endgame DVD in the vain hope that I could buy some ratings points.

Playing on the TV was a recording of game 2 of Spassky-Fischer from 1972. We paused for a while as I explained how Fischer didn't show up and lost the game by default. I could tell that his heart wasn't really in it, but for a brief second or two we shared a precious father-son moment.

From there, via a discussion of Von Bardeleben to a Nero where we shared a cheescake. He read New Scientist and told me (with much more excitement and interest) that they hadn't actually found a particle that could travel faster than the speed of light. It was a calculation error. Einstein is still right.

But many white players don't like to face the nimzo-indian. So they play a move order that makes it impossible. That's why it's called an "anti nimzo indian" - a system designed to prevent the nimzo from happening.

There's a similar name in the Ruy Lopez. Frank Marshall invented the Marshall gambit where black offers up a pawn in order to get an attack. He famously saved up this opening for years until he got a chance to spring it on capablanca. Capa saw the attack, decided he was honour-bound to walk into it, and managed to defend and win.

------------

OK, I feel quite stupid now. When you see Bb8 you
instantly get it. Too bad.

But it's by no means universal. Some other books use a different method where an ambiguous piece is described by its current square.

For example, in front of me right now is my much loved copy of Fischer's "My 60 memorable games". I have an edition published in 1989 in descriptive notation. It uses the KN/QN type of notation.

In the very first game in the book ...

Fischer vs J Sherwin, 1957

... we get to this position, with black (Sherwin) to move ...

click for larger view

In algebraic notation, his move was 15...Nce7. In my copy of the book, it is given as 15...QN-K2.

But, the other way to describe that move in descriptive would have been 15...N(B3)-K2. In other words, the knight that is sitting on the bishop's third square (c6) to e7.

Some books even mix the two styles. I pulled Irving Chernev's "Logical Chess" off the shelf. This tends to use KN and QN in the opening, but N(B3) in the middlegame and ending.

Esconced in cove d8 idea win a8 edge us bet in spread it among i st b8 ave washout racing free def inceed it jam in utter edifice b8,

Provincial queen or rook is chucked for one in looks it assumption in a hold it 17...qxb8 18.qxb8+ rxb8 19.rxb8 firm bedrocks for white in which,

too build in land it apost in easy it after bishop found to it nice in

b8 low single mind briefly considered in c7 or c1 vested it exchange in

dry it often in c7 equal pond i gauge in boot it f4b8 in fecd it's top

in flurry it ok in got him by the ghosting brevity of anymore in here

White to play (17.?) "Very Difficult"

I think the most difficult thing about this problem is that my head got in the way! I was trying to make it more difficult than it was, looking at 17. Bc7, which wins after 17...Qe8 18. Bb8 (trapping the rook and winning the exchange), but after 17...Qd5 18. Be5!? Rd2, White gains nothing. Then it dawned on me that Bf4-c7 isn't necessary; in fact, it blocks the queen's attack on Black's Be7!

17. Bb8!