Main
Date: 02 Oct 2008 12:12:42
From: John Salerno
Subject: Pinning the knight to the king/queen in the opening
I found this Q&A interesting. This is something I've wondered about when I
play, because it seems most people I play (on Yahoo) always attack my bishop
when I pin the knight (and often they preemptively move their pawn forward
before I even move my bishop!), and usually I just retreat the bishop. They
then attack again and as a result severly weaken their pawn cover for the
king. After this second attack I usually move my bishop down again, just as:

1.Bg5 h6
2.Bh4 g5
3.Bg3

I'm not convinced that move 3 is that great, but at this point I figure I've
at least forced them to weaken their pawn cover and I still haven't lost a
piece. For my part, when my own knight gets pinned, I never attack the
bishop because it seems like a waste of time. It doesn't seem like that big
of a deal in the opening.

Anyway, here's the discussion (from ChessCafe). I wonder what others think
about this. Should you just trade pieces immediately, or keep retreating the
bishop? Or not move the bishop to g5 to begin with?

-------------
Question There's a common theme in many openings, where a bishop pins a
knight to its king or queen, for example, in the Nimzo-Indian Defense. I've
heard it said that it is considered more favorable in contemporary opening
theory if the bishop takes the knight if the rook's pawn comes up and
threatens said bishop. For instance, in the Nimzo-Indian, 1 d4 Nf6 2 c4 e6 3
Nc3 Bb4 4 a3 Bxc3. I'm aware of other responses, like 4...Ba5 5 b4 etc., so
really my question is, why is taking the knight considered more favorable?
Justin Lett (USA)

Answer Sometimes, even when we're trying to be general, we tap into the
specific and wind up going wrong, very wrong. While I think I understand the
point you're raising, you may have gotten too casual in its presentation.
The suggestion you offer as an alternative, 4.Ba4, actually runs into
trouble after 5 b4 Bb6 6 c5, which is a version of the Noah's Ark Trap. So
in the line you've offered, taking the knight is better than retreating the
bishop, since the retreat loses a piece. But it's clear what you're after,
some kind of general truth. And I'm going to give it to you, though I doubt
it's the answer you've been seeking.

Let me first take the liberty of rephrasing your question, since I think you're
really asking something else. I think what you're really asking is, in
comparable situations where the bishop can safely retreat to queen-rook-four
(let's keep it on the queenside for our discussion, since kingside
transactions bring other considerations with them), why would players fancy
to take the knight instead of opting to withdraw the bishop to rook-four?

Well, in some cases, they might not prefer to take the knight. They might
find it desirable instead to keep the pressure on the knight and lure the
opponent into making a potentially weakening pawn move (let's say b2-b4, in
some comparable situations, where the bishop is not lost by retreating it).
In other instances, a player might prefer to keep the tempo, taking the
knight, knowing that a move must be expended to take the knight back. So
that kind of action may revolve around keeping or fighting for the
initiative. And why is it better to take the knight after the bishop has
been attacked by a rook-pawn (as opposed to taking the knight before the
bishop has been so menaced)? If it's better, it might be so because the side
with a similarly positioned bishop probably wants to encourage such a pawn
move (say, a2-a3). Such a pawn move wastes time, weakens queen-knight-three
(b3), and places the pawn on a square that then denies the other side's
queen bishop access (a3). Whereas, if the bishop takes the knight first (say
Bb4xc3), before it's attacked (say a2-a3), the rook-three square remains
open. If the b-pawn has taken back on c3 (b2xc3), the queen-bishop would
then have the option of deploying to queen-rook-three (a3 in the skeleton
you're concerned with). But this is not how chess should be played. I'm not
sure how chess should be played, but I know this isn't it. It's good to be
aware of such general things, but, in the play of an actual game, it's best
to really understand what's happening in the very specific position before
you, I think. Okay, I know.






 
Date: 02 Oct 2008 17:38:25
From: Carlo Wood
Subject: Re: Pinning the knight to the king/queen in the opening
I allow me to give a truely general, and therefore almost mathematic
answer. First of all, chess is fun because it isn't calculable till the
end; and that devides ones thinking process, or the methods used to make
a decision about the next moves into two categories: 1. combinations
(also called tactics) which ARE calculations; one follows a tree of
possibilities until it is proven that a (sub)goal can be reached or not,
2. strategy. Naturally your question, it being general, falls in the last
category. However, we have to state clearly that it ONLY falls into the
last category when any tactical results are over the horizon of our
calculable abilities. Thus, you look at both possibilities (take the
knight and not take the knight, and neither can be proven better).

Therefore, in order to answer your question I have to define what
'strategy' is.

Strategy means: optimizing your chances. You do that what makes it most
likely that a favourable opportunity (a combination) will be thrown into
your lap later on.

Ok, that still doesn't make it clear what to do; so how to optimize your
chances in chess? Mathematically I'd say: you maximize the weighted sum
(over all your pieces, subtracting the same for the result of all your
opponents pieces) of the possibilities that your pieces have. I'll get
back to what we're weighting with in a second. First, lets look at why
'possibilities' are so important:

Well, suppose you can think EXACTLY 5 moves ahead - and so can your
opponent. And suppose further more that you have twice as many
possibilities as your opponent to choose from, say your opponent has
exactly 10 possible moves to choose from and you have 20 possible moves
to choose from. Then the number of total positions after 5 moves is: 20 *
10 * 20 * 10 * 20 * 10 * 20 * 10 * 20 * 10 = 200^5 = 320000000000
positions. However, you can 'steer' into what direction you're going: if
you pick your moves carefully - then the only possibilities that your
opponent will be able to choose from is 10^5 = 100000 positions. On the
other hand you have more control: every move you have twice as many
possibilities and you will be able to steer towards 20^5 = 3200000
positions (the other 320000000000 - 100000 - 3200000 positions are not
favoured by either of you). So, if you end up "lucky" and suddenly--
while looking ahead 5 moves-- you see a combination (that you couldn't
see a move ago) then that will be one of those 3200000 positions that
just came out of the 'fog' beyond the horizon of your tactical
capabilities. On the other hand, your opponent will have to do with only
100000 positions! So, ever move he has 32 times LESS chance on a
combination! Naturally, you're going to win that game. But, ok -- having
TWICE as many possibilities is a lot of course. In reality it is much,
much more subtle but the idea is the same: if you have options that your
opponent doesn't have, then your chances for a combination are larger.

Ok, so now we can reformulate strategy: strategy means to increase the
possibilities of your pieces. However, they need to weighted: not all
possible moves are as interesting of course! They need to weighted with
two major ingredients: 1) The ammount of Threat those possibilities
impose, 2) The long-term-ness of those threats (how easy is it to
neutralize your threat). Starting with the first: giving check or taking
a piece is a great threat, because they severly limit your opponents
possibilities. Attaching something is a lesser threat, but still rather
large: again it limits your opponents possibilities (he will have to
defend his piece, or move it away) and of course, threating to attack a
square that your opponent can move to before you attach it, and not
anymore after you attack it, is also a threat because it limits, once
more, your opponents possibilities. The second ingredient means the
following: suppose you have a knight that is standing at the front;
taking all kind of squares away from your opponent but more importantly,
threating with each of it's possible moves to take "random" other squares
away, taking something that has to be taken back or even giving check.
Then that knight imposes a great threat: it has many possibilities and
each is potential going to be valuable for some successful combination.
However, as we said already before: both, you and your opponent can think
ahead five moves... So, IF you get a combination thrown into your lap it
will be one that is five moves deep (in this ideal example/model), hence,
your knight will ONLY be part of a combination of five moves, if at all.
Logic therefore tells us that if that knight can easily be chased away by
your opponent, by attacking it with a pawn say, then the CHANCE that it
WILL become a part of a combination is in fact very small! Hence, it has
a lot of threating potential, but that threat isn't 'long-term' and
therefore not adding much to the weighted sum that we seek to maximize.

Now back to your pinned-knight / bishop question.

There are several things to consider in regard to what will optimize your
chances. First of all, as said before, you have to look those 5 moves
ahead. Is there not ALREADY a combination on the board? If you miss it,
then strategy is of little use. Secondly, you have to consider:
Will my chances be increased by getting rid of that knight, and also of
my bishop? Getting rid of the knight might weaken a vitial square that
the knight is defending (note that if such a square exists, it will be of
a DIFFERENT color then the knight is standing on, and hence of a
different color than the bishop: the bishop is not capable of attacking
that square itself; it can ONLY sacrifice itself upon a knight to have
any influence on that square. The disadvantage of that however is that
one loses the bishop pair with such an action which has considerable
influence on the chances in a possible resulting end-game: for example,
bishops are better in an open game, and knights are better if the (pawn
in the) center is frozen and immobile. Those, however are long term
considerations; one might gamble on the fact that getting rid of the
knight will allow you to win a pawn within 30 moves, and before that such
end-game things became relevant.

Knowing the above we can make the following observations:
Consider the moves: Bg5 (pinning the knight on f6).
Why would black attack it almost immediately with h6?
Well, because it is clear that white has more possibilities/threats:
black can not use the knight (it is pinned), and has NO options
therefore. While white can, at any time chose to exchange the knight (or
not); white threatens to eliminate the knight at any moment. What is
valuable here is the THREAT, not the exchange itself, usually (unless the
center is totally stuck, thus-- or, like in many variations of Sicilian,
black has no pawn on the c-file, a pawn on d6 and pushed his pawn to e5.
In that case the whole opening evolves around the square d5: black wants
to push that pawn forwards and white wants to stop it. Hence, if black
plays h6, white would take the knight on f6 because the knight defends d5
and the bishop cannot attack that square). Thus, normally white would
WAIT with exchanging the knight until that exchange is part of a
combination that ACTUALLY wins something (ie, a pawn). So, if at some
moment we suddenly approach this moment and the threat turns into a
combination then it is TOO LATE for black to do something about it! He
might play h6, seeing he NEEDS his knight - but white will take the
knight. Therefore black often plays h6 as soon as possible getting RID of
whites threat either way: if white takes then the possibilities are
leveled again (both pieces have no options, they are next to the board)
and if white retreats to h4 THEN black has to option to, at any moment,
play g5! And now white can NOT respond with taking the knight. So, in
this case black KEEPS the option to use his knight. If now the moment
approaches that black really needs his knight and Bxf6 threatens to
become part of a real combination, then black plays g5. In this case the
threat of the bishop (to exchange itself on the knight) was 1 ply deep,
instead of 3 ply deep.

Thus, in the most general case, white pins the knight to take away
possibilities of his opponent: can't move the knight anymore, and create
possibilities for himself: can exchange at will (but shouldn't unless
there are other considerations that favour the knight over that bishop in
that position). And black will play quickly h6 (or Be7) in order to allow
himself to react within one ply again, in case he really needs his
knight, but will in general wait with g5 until that is really necessary
(assuming it is a weaking of his pawn structure), or until the position
asks for such a move because the pawn advance is simple good (ie, a
kingside attack after white castled short).

For clarity, let me summarize that there is NO general rule that tells
you to take the knight or not in all cases. It depends; and in order to
understand when it is good and when not, you probably will have to read a
few specific opening books about such cases (that actually explain
things, not just say 'that is the move').

Carlo



  
Date: 02 Oct 2008 14:09:28
From: John Salerno
Subject: Re: Pinning the knight to the king/queen in the opening
"Carlo Wood" <[email protected] > wrote in message
news:[email protected]...

> Therefore, in order to answer your question I have to define what
> 'strategy' is.

Thanks for all that! And by the way, I am about to start reading Seirawan's
book on strategy tonight. I read the intro last night and it seems very
interesting. In contrast with tactics (which book I already read and I find
tactics very interesting), strategy seems to be more about this long-term
planning and positional play, something I've wondered about for a while. I'm
excited to start reading it, and maybe that will also help me understand the
things you are saying about this particular question as well.




 
Date: 02 Oct 2008 10:33:34
From:
Subject: Re: Pinning the knight to the king/queen in the opening
On Oct 2, 12:12=A0pm, "John Salerno" <[email protected] > wrote:
> I found this Q&A interesting. This is something I've wondered about when =
I
> play, because it seems most people I play (on Yahoo) always attack my bis=
hop
> when I pin the knight (and often they preemptively move their pawn forwar=
d
> before I even move my bishop!), and usually I just retreat the bishop. Th=
ey
> then attack again and as a result severly weaken their pawn cover for the
> king. After this second attack I usually move my bishop down again, just =
as:
>
> 1.Bg5 h6
> 2.Bh4 g5
> 3.Bg3
>
> I'm not convinced that move 3 is that great, but at this point I figure I=
've
> at least forced them to weaken their pawn cover and I still haven't lost =
a
> piece. For my part, when my own knight gets pinned, I never attack the
> bishop because it seems like a waste of time. It doesn't seem like that b=
ig
> of a deal in the opening.
>
> Anyway, here's the discussion (from ChessCafe). I wonder what others thin=
k
> about this. Should you just trade pieces immediately, or keep retreating =
the
> bishop? Or not move the bishop to g5 to begin with?

John, this is a question that cannot be answered well in general
terms, but only in the context of a specific position. And even in a
given position, there may be no single correct answer. For example, in
the Ruy Lopez, after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6, both 4.Ba4 and 4.Bxc6
are perfectly good moves. The latter creates a pawn structure slightly
favorable to White, while the former envisions eventually retreating
the bishop to c2, from which post it is highly effective and usually
safe from further harassment by pawns or knights.
And whether or not to nudge the bishop immediately is a matter of
taste. Instead of 3...a6, Black has perfectly playable alternatives in
3...Nf6, 3...Bc5, 3...d6, or even 3...Nge2 or 3...g6. In the Nimzo-
Indian, after 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4, White can immediately force
the B-for-N exchange with 4.a3 (Saemisch variation), but he also has
many other alternatives: 4.Qc2, 4.Qb3, 4.e3, 4.Nf3, 4.Bg5, etc.
When it's the QB pinning a knight on f6 or f3, more caution is
usually called for, since unless one plans to castle queenside,
pushing h7-h6 and g7-g5 can be weakening and dangerous. For example,
in the Queen's Gambit, after 1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Bg5, an
immediate 4...h6?! can be answered by 5.Bxf6 and Black must already
submit to pawn structure disruption by 5...gxf6, or loss of a pawn
after 5...Qxf6 6.cxd5 exd5 7.Nxd5. Therefore Black usually answers
4.Bg5 with 4...Be7 or 4...Nbd7, and later refrains from chasing the
bishop by h7-h6 and g7-g5, since then his king can find safety neither
on kingside nor queenside.
If there's any general guide to deciding between retreating your
bishop or capturing the knight, it might be something like this:
Bishops are most effective from a distance, and on points where they
are not subject to harassment by pawns or knights. On such squares as
b5, a4, b3, g5, h4, and g3 (or, for Black, b4, a5, b6, g4, h5, g6)
they are subject to one or the other, by pawn pushes and ...Nc6-a5
or ...Nf6-h5. Aside from fianchetto development (b2-b3 and Bc1-b2, or
g2-g3 and Bf1-g2), good, usually safe squares are a2, c2, h2, and
sometimes (though less often) f2. Thus, if a bishop on b5 has, by
retreating, the chance to reach a2 or c2 safely, or one on g5 to reach
h2, then retreat is likely a good option. If by taking the knight some
compensation is obtained (e.g. pawn structure), then capturing may be
a good option. If neither of these goals can be reached, then moving
the bishop to b5 or g5 may have been a bad idea in the first
place.

> -------------
> Question There's a common theme in many openings, where a bishop pins a
> knight to its king or queen, for example, in the Nimzo-Indian Defense. I'=
ve
> heard it said that it is considered more favorable in contemporary openin=
g
> theory if the bishop takes the knight if the rook's pawn comes up and
> threatens said bishop. For instance, in the Nimzo-Indian, 1 d4 Nf6 2 c4 e=
6 3
> Nc3 Bb4 4 a3 Bxc3. I'm aware of other responses, like 4...Ba5 5 b4 etc., =
so
> really my question is, why is taking the knight considered more favorable=
?
> Justin Lett (USA)
>
> Answer Sometimes, even when we're trying to be general, we tap into the
> specific and wind up going wrong, very wrong. While I think I understand =
the
> point you're raising, you may have gotten too casual in its presentation.
> The suggestion you offer as an alternative, 4.Ba4, actually runs into
> trouble after 5 b4 Bb6 6 c5, which is a version of the Noah's Ark Trap. S=
o
> in the line you've offered, taking the knight is better than retreating t=
he
> bishop, since the retreat loses a piece. But it's clear what you're after=
,
> some kind of general truth. And I'm going to give it to you, though I dou=
bt
> it's the answer you've been seeking.
>
> Let me first take the liberty of rephrasing your question, since I think =
you're
> really asking something else. I think what you're really asking is, in
> comparable situations where the bishop can safely retreat to queen-rook-f=
our
> (let's keep it on the queenside for our discussion, since kingside
> transactions bring other considerations with them), why would players fan=
cy
> to take the knight instead of opting to withdraw the bishop to rook-four?
>
> Well, in some cases, they might not prefer to take the knight. They might
> find it desirable instead to keep the pressure on the knight and lure the
> opponent into making a potentially weakening pawn move (let's say b2-b4, =
in
> some comparable situations, where the bishop is not lost by retreating it=
).
> In other instances, a player might prefer to keep the tempo, taking the
> knight, knowing that a move must be expended to take the knight back. So
> that kind of action may revolve around keeping or fighting for the
> initiative. And why is it better to take the knight after the bishop has
> been attacked by a rook-pawn (as opposed to taking the knight before the
> bishop has been so menaced)? If it's better, it might be so because the s=
ide
> with a similarly positioned bishop probably wants to encourage such a paw=
n
> move (say, a2-a3). Such a pawn move wastes time, weakens queen-knight-thr=
ee
> (b3), and places the pawn on a square that then denies the other side's
> queen bishop access (a3). Whereas, if the bishop takes the knight first (=
say
> Bb4xc3), before it's attacked (say a2-a3), the rook-three square remains
> open. If the b-pawn has taken back on c3 (b2xc3), the queen-bishop would
> then have the option of deploying to queen-rook-three (a3 in the skeleton
> you're concerned with). But this is not how chess should be played. I'm n=
ot
> sure how chess should be played, but I know this isn't it. It's good to b=
e
> aware of such general things, but, in the play of an actual game, it's be=
st
> to really understand what's happening in the very specific position befor=
e
> you, I think. Okay, I know.



  
Date: 02 Oct 2008 16:01:45
From: Chess One
Subject: Re: Pinning the knight to the king/queen in the opening

<[email protected] > wrote in message
news:fd7285e5-0648-437f-af62-e2dd3eb9cbd2@m36g2000hse.googlegroups.com...
On Oct 2, 12:12 pm, "John Salerno" <[email protected] > wrote:
> I found this Q&A interesting. This is something I've wondered about when I
> play, because it seems most people I play (on Yahoo) always attack my
> bishop
> when I pin the knight (and often they preemptively move their pawn forward
> before I even move my bishop!), and usually I just retreat the bishop.
> They
> then attack again and as a result severly weaken their pawn cover for the
> king. After this second attack I usually move my bishop down again, just
> as:
>
> 1.Bg5 h6
> 2.Bh4 g5
> 3.Bg3
>
> I'm not convinced that move 3 is that great, but at this point I figure
> I've
> at least forced them to weaken their pawn cover and I still haven't lost a
> piece. For my part, when my own knight gets pinned, I never attack the
> bishop because it seems like a waste of time. It doesn't seem like that
> big
> of a deal in the opening.

Your move 3 [or move 4 in a game after you open 1. d4 constitutes the Torre
attack] and there are several points to make about it. If the Kt captures
[also by balck playing after 1. ... Nf6 2. ...Ne4 3. ...g5] then the
recapture by the h-pawn develops the white rook at h1 on a half-open file.

A point about attacking the bishop by moving a rook pawn to rook-three is
that it 'asks the question' of the bishop, to exchange, or to retreat to the
rook file [maintaining the pin, but limiting the scope of the bishop].

> Anyway, here's the discussion (from ChessCafe). I wonder what others think
> about this. Should you just trade pieces immediately, or keep retreating
> the
> bishop? Or not move the bishop to g5 to begin with?

**The bishop move is fine and the retreat is also fine. In respect of what
follows - a /very/ different situation and no analogy at all, is that in
this instance the bishop pin is relative, whereas with Bb5 it is absolute.
Bg5 allows Black to break the pin [even if this loses the Queen] but more
importantly, by tactical finesse, offers to reverse the pin. The challenge
to the bishop by the Rook pawn decides if there is to be an excahnbe B for
Kt or a more limited future for the Bishop. This is not analogous to Bb5
where [in the Ruy] black may not move the knight to break the pin without
advancing the b pawn.

John, this is a question that cannot be answered well in general
terms, but only in the context of a specific position. And even in a
given position, there may be no single correct answer. For example, in
the Ruy Lopez, after 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6, both 4.Ba4 and 4.Bxc6
are perfectly good moves. The latter creates a pawn structure slightly
favorable to White, while the former envisions eventually retreating
the bishop to c2, from which post it is highly effective and usually
safe from further harassment by pawns or knights.
And whether or not to nudge the bishop immediately is a matter of
taste. Instead of 3...a6, Black has perfectly playable alternatives in
3...Nf6, 3...Bc5, 3...d6, or even 3...Nge2 or 3...g6. In the Nimzo-
Indian, after 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4, White can immediately force
the B-for-N exchange with 4.a3 (Saemisch variation), but he also has
many other alternatives: 4.Qc2, 4.Qb3, 4.e3, 4.Nf3, 4.Bg5, etc.
When it's the QB pinning a knight on f6 or f3, more caution is
usually called for, since unless one plans to castle queenside,
pushing h7-h6 and g7-g5 can be weakening and dangerous. For example,
in the Queen's Gambit, after 1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Bg5, an
immediate 4...h6?! can be answered by 5.Bxf6 and Black must already
submit to pawn structure disruption by 5...gxf6, or loss of a pawn
after 5...Qxf6 6.cxd5 exd5 7.Nxd5. Therefore Black usually answers
4.Bg5 with 4...Be7 or 4...Nbd7, and later refrains from chasing the
bishop by h7-h6 and g7-g5, since then his king can find safety neither
on kingside nor queenside.
If there's any general guide to deciding between retreating your
bishop or capturing the knight, it might be something like this:
Bishops are most effective from a distance, and on points where they
are not subject to harassment by pawns or knights. On such squares as
b5, a4, b3, g5, h4, and g3 (or, for Black, b4, a5, b6, g4, h5, g6)
they are subject to one or the other, by pawn pushes and ...Nc6-a5
or ...Nf6-h5. Aside from fianchetto development (b2-b3 and Bc1-b2, or
g2-g3 and Bf1-g2), good, usually safe squares are a2, c2, h2, and
sometimes (though less often) f2. Thus, if a bishop on b5 has, by
retreating, the chance to reach a2 or c2 safely, or one on g5 to reach
h2, then retreat is likely a good option. If by taking the knight some
compensation is obtained (e.g. pawn structure), then capturing may be
a good option. If neither of these goals can be reached, then moving
the bishop to b5 or g5 may have been a bad idea in the first
place.

> -------------
> Question There's a common theme in many openings, where a bishop pins a
> knight to its king or queen, for example, in the Nimzo-Indian Defense.
> I've
> heard it said that it is considered more favorable in contemporary opening
> theory if the bishop takes the knight if the rook's pawn comes up and
> threatens said bishop. For instance, in the Nimzo-Indian, 1 d4 Nf6 2 c4 e6
> 3
> Nc3 Bb4 4 a3 Bxc3. I'm aware of other responses, like 4...Ba5 5 b4 etc.,
> so
> really my question is, why is taking the knight considered more favorable?
> Justin Lett (USA)

It is not considered more favorable. The bishop retreat to a5 is higly
analysed, and occurs in the French defence [which your example could
transpose to] and is played at the very highest levels. Ex W Ch Khalifman is
probably the best analyst of 'the Armenian Variation' of the French.

But one cannot speak of more or less favorable without understanding what
mode of play is favored! And what skill levels are involved. Generally one
should avoid highly tactical openings below 1600 rating. Otherwise one is
likely to become befuddled.

If you are a real bruiser of a player, and everything is tactics to you,
then play tactical openings - but you must be able to look quite deeply to
the result of your tactics, some 6 to 8 moves to be able to achieve success
this way. That is not likely for a sub-1600 player.

Therefore, the general recommendation is to avoid tactical openings and
simply concentrate on developing your pieces without creating weaknesses in
your position. Otherwise you likely attempt to do what you are not yet able
to achieve, long-string tactical sequences.

> Answer Sometimes, even when we're trying to be general, we tap into the
> specific and wind up going wrong, very wrong. While I think I understand
> the
> point you're raising, you may have gotten too casual in its presentation.
> The suggestion you offer as an alternative, 4.Ba4, actually runs into
> trouble after 5 b4 Bb6 6 c5, which is a version of the Noah's Ark Trap.

Sorry - your correspondent is not entirely correct here - since either
White's a or c pawn may advance to provide the b luft, and these three black
pawn moves only encourage a retreat of the bishop, at some expence of
black's own development. This is the sort of position that a strong White
player may prefer, a black q-side already committed, which can be moped up
in a middle game.

> So
> in the line you've offered, taking the knight is better than retreating
> the
> bishop, since the retreat loses a piece.

Not so, c5 then c4 is necessary, this allows white time to retreat his
bishop, by a3 or even aggressively retreat it by a4! immediately challenging
the black pawn chain when any ba suffers bxc4 in an already superior game
for white.

So - not so abstract, and not so tactical?

Cordially, Phil Innes


> But it's clear what you're after,
> some kind of general truth. And I'm going to give it to you, though I
> doubt
> it's the answer you've been seeking.
>
> Let me first take the liberty of rephrasing your question, since I think
> you're
> really asking something else. I think what you're really asking is, in
> comparable situations where the bishop can safely retreat to
> queen-rook-four
> (let's keep it on the queenside for our discussion, since kingside
> transactions bring other considerations with them), why would players
> fancy
> to take the knight instead of opting to withdraw the bishop to rook-four?
>
> Well, in some cases, they might not prefer to take the knight. They might
> find it desirable instead to keep the pressure on the knight and lure the
> opponent into making a potentially weakening pawn move (let's say b2-b4,
> in
> some comparable situations, where the bishop is not lost by retreating
> it).
> In other instances, a player might prefer to keep the tempo, taking the
> knight, knowing that a move must be expended to take the knight back. So
> that kind of action may revolve around keeping or fighting for the
> initiative. And why is it better to take the knight after the bishop has
> been attacked by a rook-pawn (as opposed to taking the knight before the
> bishop has been so menaced)? If it's better, it might be so because the
> side
> with a similarly positioned bishop probably wants to encourage such a pawn
> move (say, a2-a3). Such a pawn move wastes time, weakens
> queen-knight-three
> (b3), and places the pawn on a square that then denies the other side's
> queen bishop access (a3). Whereas, if the bishop takes the knight first
> (say
> Bb4xc3), before it's attacked (say a2-a3), the rook-three square remains
> open. If the b-pawn has taken back on c3 (b2xc3), the queen-bishop would
> then have the option of deploying to queen-rook-three (a3 in the skeleton
> you're concerned with). But this is not how chess should be played. I'm
> not
> sure how chess should be played, but I know this isn't it. It's good to be
> aware of such general things, but, in the play of an actual game, it's
> best
> to really understand what's happening in the very specific position before
> you, I think. Okay, I know.