how do you know whether or not attacks are unsound

Sort:
Karrysparov
I understand an attack that immediately can be refuted is bad, any other examples? And do you have tips to create better attacks
BoardMonkey

You have to calculate.

Karrysparov
Please elaborate on why I should calculate, are you stating that I should calculate to see whether or not I have an advantage after the attack? And if so, what do you consider an advantage
KevinOSh

Learning how to evaluate a position correctly at the end of a line is an art which usually takes years to get good at. So except for the most obvious cases there are no simple answers for how to know whether an attack or any move is good or bad.

There is a famous book called the Art of Attack.

Also there are many lessons on chess.com for example

https://www.chess.com/lessons/attacking-king-hunts

https://www.chess.com/lessons/how-to-checkmate

https://www.chess.com/lessons/champion-tactics-with-gm-wolff-double-attacks

An advantage can either be material or positional. There are static and dynamic features in a position, and one player could have one or more static advantages while the other has one or more dynamic advantages. You could be up a whole Queen but the other player could still have the advantage if your king is very unsafe.

BoardMonkey

Think Like a Grandmaster, Alexander Kotov

https://www.amazon.com/gp/aw/d/B00PPH2Q8W/ref=tmm_kin_swatch_0?ie=UTF8&qid=&sr=

Why and how to calculate. Good luck. It's an advanced subject about which I know little. Except something about choosing between candidate moves.

paper_llama

Four points come to mind.

1
You usually need at least 3 attackers or 2 more attackers than there are defenders. You can count either by dividing the board in half, or by thinking in terms of quadrants.

2
A successful attack often needs to remove the king's pawn cover. There are two ways. First is playing a pawn break, and second is sacrificing a piece. It may sound obvious, but it's useful to note since in most positions only one of those will be an option. In other words when a pawn storm isn't an option you automatically know to be thinking in terms of sacrificing on squares like f6, g7, h6.

3
Other than opening files and diagonals to the king, a successful attack needs attackers that can combine on a single square (meaning they attack the same square). So sometimes it's useful to judge an attack like this, you see one of your attackers is a dark squared bishop, so you look for ways your knight or rook can influence a dark square.

4
Finally, an attack is more likely to be successful when your opponent doesn't have counterplay. Often this means it's hard for them to open a file in the center or on the queenside. The most common way of dealing with an attack is generating play somewhere else. If your opponent can create a few threats or trade off a few pieces then it will diminish your attacks. So in a game you can make a judgement like this... ask yourself how many moves it will take your opponent to open a line in the center, or how many moves you need to make a threat near their king. If your opponent's play is much faster, then a direct attack is probably not worth it at that moment.

paper_llama

And it's also good to look at many attacking games as examples, that way you get a feel for some common patterns.

For example some attacks chase the king out to the e file... so if you already control the e or d file with a rook, you may become much more confident in that type of attack.

Some attacks are thwarted by the existence of a knight or bishop on f8 (when black 0-0), and some attacks are more effective when a rook is on f8 (the rook tends to box the king in without helping defend).

So other than general knowledge, it's good to learn some specifics from actual attacking games

 

MonstrousReprobate
zolidmill wrote:

Four points come to mind.

...

Your comments are fantastic. Thanks for this info. I'm working on drawing up diagrams of decision-making-algorithms for me to follow when I play (obviously can't make something perfectly all-encompassing but I think creating such a system will help me improve a lot) and this is exactly the kind of strategic stuff I need to collect insights on. I appreciate you taking the time to answer OP so fully!

paper_llama
MonstrousReprobate wrote:
zolidmill wrote:

Four points come to mind.

...

Your comments are fantastic. Thanks for this info. I'm working on drawing up diagrams of decision-making-algorithms for me to follow when I play (obviously can't make something perfectly all-encompassing but I think creating such a system will help me improve a lot) and this is exactly the kind of strategic stuff I need to collect insights on. I appreciate you taking the time to answer OP so fully!

No problem. I'm glad if someone finds it useful happy.png