Middle game calculation

I have a question about calculation and safety check.

When I analyse a move or a variation, what is the best way to be sure this move is safe :

A) Visualize the move(s) as played and check for opponent's threats;

B) Identify move's effects (square or pieces newly attacked, defended and undefended) and verify threats with logical deduction.

I find that B works better for me but I have the feeling that A would be better on long term improvement and I wonder if B cannot become a crutch or act as an anchor to progress.

What do you think about that ?

I think what helps me the most with outright blundering is simply doing tactics puzzles. You get to the point where you just automatically see stuff you were missing before.

IMO

Option B is fine. Personally I love using logical shortcuts as much as possible. But having said that, there's no real way to avoid the requirement of calculating short forcing sequences. Luckily it's (usually) not much work. In most positions your opponent can't check your king, can't attack your queen, and you don't have a lot of pieces hanging.

So I guess I'm saying it's a bit of A and B... I guess my main point is don't go down rabbit holes of calculation trying to find threats. Keep it logical and short.

And yeah, solving tactic puzzles helps build good habits because they always involve forcing sequences.

Tactics are what you use to see if your move is safe, but it doesn’t necessary mean the move is playable. There are long term positional features that need to be considered as well. The harder part is the calculating a position that isn’t on the board yet and sometimes can be 6 or so moves later. I’ve read of GMs saying “long variation, wrong variation” because chances are you are going to miss something. Zwischenzug can strick at any time during these. Even 3 move calculations that find the best moves can be quite difficult in a lot of hectic middlegames.

COUNTING is one of the most useful calculation shortcuts. It's absolutely not a replacement for calculating, but - if you need to figure out if you can push a pawn to a protected square - just count the pieces attacking, the pieces defending, and - in case of a tie - whoever is OCCUPYING The SQUARE wins. 

You can also use this method of figuring out whether your king (or opponent's king) is in critical danger: Count the attacking pieces (including pawns), count the defending pieces (excluding pawns - but count the king), ties usually favor the defender. Again, you're not going to get out of calculating certain specific sequences, but let's say you're trying to decide if it's safe to move a piece away from your king while your opponent is putting pressure on your king: Count! They're not going to break through if you have 3 defenders (including the king) against 2 attackers. Low level players will simply become frustrated, trade pieces, or sacrifice wildly, and that's what you want.

One of the most important calculation habits you should start cultivating now: Remember which pieces are gone. It seems rudimentary, but - when you're calculating - it's quite easy to forget which files have opened up (for example) or which backward pawn might be hanging at the end of the sequence. You have to learn "tok, tok, take the rook" or else you can't play chess at all, but it's a very good idea to start building your consciousness of what's changing in the position BESIDES the material count.

Oh, that's a neat trick.

When I was new I didn't like the counting thing because you had to remember what to do in case of a tie, but the memory trick "whoever occupies the square wins" (or at least, doesn't lose material) is pretty cool.

Yep.  Agree on the zwischenzug catching me.  And also my opponents as well.   Those are hard to see in advance