Calculation - Any tips from strong players?

Hello everyone,

try to solve this tactic problem: http://www.chess.com/tactics/?id=66161

I was wondering, which white's move strong players calculate first? Is that completely random or is there a correct way to choose one? There are really many candidates moves. Tactics is  not so difficult, but it can take long if one starts with wrong moves.

Hi duniel,

I'm not a very strong player but maybe my thoughts are helpful to you anyway, so I post them for you.

Here's the puzzle you're refering to:

 The task is mate in three. As a general advise for our moves we should always look for the moves which are restricting the oponent's king mostly. As a second consideration, I'm looking if Black can interrupt my series with checks by himself. Since both the queen and rook are able to attack White's king I know that I have to look for possible checks by myself and can't allow a break for Black. This reduces the possibilities to two pieces, the queen and the rook. For two reasons I firstly looking out for queen moves: 

a) The queen is generally more powerful and able to deprive the king of more fields. 

b) In this special situation White's bishop is hanging and only the queen can protect it at the same time while giving check.

So our first candidate is Qb5+. Looks good at the first glance, leaving the king only one square to escape. Unfortunately, after 1.Qb5+ Kd4 there is no way for White to continue his series of checks without sacrifising the queen. Such sacrifice would not help since the position is to open for a mate with rook and bishop only. Since we already established that we can't interrupt or series of checks 1.Qb5 is a dead end.

There's another queen check to consider, Qc7+. But since it let's the same problematic square d4 open for the King to escape and without better oportunities to continue we can disregard this move as well.

This leaves us with the rook, sacrifising the bishop: 1.Ra5+ Kxc4. NOw the task is easier. actually, the king did us a favour capturing the bishop because this opens the 4th rank for our pieces.

Since it would make no sence to move the rook again for a check because the king could simply retreat to the same field it's time for the queen. Only one square makes sence in this position, 2.Qb5+ Kd4.

The third move is almost playing by itself, the queen is in a good position and restrains the king heavily so we use the rook. 3.Ra4# and done for this variation, but our task is not finished yet.

Instead of capturing our bishop Black's king can reject the sacrifise try to run: 1.Ra5 + Kd4. We know the same mating attack wouldn't work because our own bishop is blocking the 4th rank. But with the rook guarding the 5th rank we've got another way to hunt the king: Our queen, rook and bishop are all three eying d5 whilst Black is protecting it only two times. That suggests we have an opportunity for a good trade. Easy góing: 2.Rd5+ Rxd5 Qxd5#.

Was this reply helpful?

Cheers,

Spielkalb

duniel,

Spielkalb explained it very well. I believe that when he writes, "field(s)" he is referring to the imaginary fields of force that the pieces and pawns create on squares on the board whether occupied or not by any other pawn or piece. The complex of squares that the enemy K is restricted from forms the mating net. As you get stronger in your mind it will become second nature to see the mating net.

Also, after 2.Rd5+ Black has 3 options:

a. 2...Rxd5 as SpiSelkalb writes

b. 2...KxB(c4) 3.Qb5#

c. 2...Ke4. 3.RxR(d6) (discovered check with White Q at b7 check)

3...Ke5. 4.Qd5#

It is not a forced mate in 3.

Restriction the King and possible counter check. 

@transpo, thanks for your compliment and the additional variations which I overlooked. Your right in your interpretation what I meant by "fields". In German we use a related term, hence the linguistic confusion. 

Is there anything to add how you narrow down the number of possible moves? 

Spielkalb wrote:

@transpo, thanks for your compliment and the additional variations which I overlooked. Your right in your interpretation what I meant by "fields". In German we use a related term, hence the linguistic confusion. 

Is there anything to add how you narrow down the number of possible moves? 

When you become a strong player you only need one diagram. The one in your mind. Then it becomes very easy to do calculations by moving the pieces on the diagram in your head. Also, you learn to pay very close attention to what squares the enemy K can and cannot go to from the square that it is on.

Got that one right, first look at mating nets, look at all the checks and forced lines with checks. If you can't see anything forced, you start looking for ways to win pieces (which is also often done by checks or check threats, skewers forks etc)

GM Melik just did a video on calculation btw

I think you're quite right regarding tactical short-term archievements, just a little remark. Allow me to add that some tactics are aimed not at material but at positional advantages to support a long-term strategy. E.g., forcing a piece trade to produce a doubled pawn, weakening dark or light squares in the opponents set up etc. Does this make sense?

Thanks for the information on the video btw, have to take a look at it. 

@ Spielkalb, I understand, I was refering specifically to chess.com's tactics trainer, I think you won't find those there. think there's a criterium that something must be won by force.

Your suggestions are good though, and I would start looking for that after my 'checklist'

Ah, I thought you'd be talking about real games. Sorry, my bad. I haven't used chess.com's tactics trainer very much, but one from another side on which goals are pretty clear: Either mate in x moves or win an material advantage of at least 2 pawns in the next couple of moves despite of positional long-term archievements for example.

It's amazing how difficult mate in 2s can be, especially if the first move isn't too forcing. I did the m2 chapter of Yusupov's book and really struggled on some.

In both real games and problems, sometimes candidates is NOT the first thing to look for.  If you look at Jacob Aagaard's book on Calculation (it's called "Grandmaster Preparation: Calculation", the first of a 5-book series he's working on, 2 of which are out now, 3 still to be published), there are 8 different forms of calcualtion.

Candidates is only one of the 8.  In the case of this problem, Prophilaxis is important.  What threats does Black have?  If I don't check him, am I in trouble?  The answer is yes.  Black threatens 1...Rd1+ 2.Kg2 Qg5+ 3.Kh2 Bd6+ and White's dead, so Quiet moves don't work.  White must make what are called "Loud Moves", a term borrowed from "Forcing Chess Moves" (Great book, by the way)

Therefore, I'd immediately write off 1.Qb5+ because of 1...Kd4, and White has no "reasonable" checks, or "Loud Moves".

So next I look at Ra5+.  If he takes the Bishop, then at worst, you win the Queen, so Ra5+ immediately is a good "Time Trouble" move.  If you are in time trouble, don't try to find the most flashy mate, take the first route that wins easily.  Winning the Queen for a Bishop is an easy win.

Therefore, no need to calculate further with 1...Kxc4 (a common mistake by beginners, spending another 5 to 10 minutes calculating this line to the end when you already know it at worst wins the Queen and stops the mate).  If he plays it, then if you have time, calculate that further, don't waste time on it now, and if time is short, take the Queen.  Instead, put your energy on 1...Kd4.  A quick glance says Qa7 and Rd5 are the only checks.  Since Rd5 blocks off the Rd1 threat, I'd look there first, and low and behold, Black can't take the Rook because Queen recapture is mate.  Going 2...Ke4 gets you the Rook with Discovered Check, so again, quit calculating there until he actually plays Ke4, and lastly, taking the Bishop leads to 3.Qb5#

So thru the use of Prophylaxis, instead of candidates, the move was found pretty quickly.  If he takes the Bishop on move 1, only then continue to calculate as you know at worst you get the Queen with a safe White King.

Good stuff thriller.

I have a feeling that will be similar to what Kotov says in his book, where he criticises looking at one line, rejecting it and going back to it.  Efficient calculation is important.

I really don't want to be mean, but why are 1500's and 1300's talking as if they were strong players? I mean...when I saw this thread it didn't even cross my mind to give advice to a fellow 2000 player, because I don't consider myself to be a strong player. Yet 1500's and 1300's do. How about you guys go back into "learning" mode rather than "teaching others" mode?

Now on topic: all I could say is that you would look at the most forcing moves first. And for calculating, after you have your candidates: Kotov's method is excellent.

well, yeah, even though 2000 in turn-based isn't 2000.

@ThrillerFan:

Thanks for your helpful explanation which provides a much deeper insight than my attempt. Learned a lot of it!

@georgemarian: I didn't check the rating of the opening poster but simply  saw the question which rose my interest and I wanted to be answered. As for my experience the best way of learning is to try to express your thoughts in the most accurate way, which I tried. If that avoked the impression of a "teaching mode", please accept my apologise. I know that I'm not a strong player which I've pointed out in the very beginning of my first post.

I really like how you disqualify other people as strong players and then portray yourself as one.

If we were to nuance that we are not strong players (which is something most people know ofcourse) 90% of the forum would be disclaimers.

How bout you stop telling us what to do on an open forum.

Thanks everyone for sharing ideas, especially ThirllerFan, it was helpful. However, Spielkab, in Tactics trainer you do not know that you are supposed to look for mate in 3.

The way I solved was actually the one I used. I checked Qc7+ and Qb5+ becouse queen is the strongest piece and it did not involve saccing material. Then I checked Qd5+, but did not find anyting. Winnign Ra5+ was the last to consider and I found winning line almost immediately.

This is why I find it interesting. Out of 1m 35s it took me to solve the problem I spent 90% calculating wrong first moves, this is why I was looking for ideas which move to condider first.

Actually, I am 2000+ FIDE but this does not seem too relevant to me.

Have you tried Kotov's 'Think Like a Grandmaster'? The whole first chapter is on the analysis of variations.