Calculation OTB

Hi,

I find difficult to calculate long lines OTB, more than online (I can better visualise online).
That's even more the case in long games since in blitz i use patterns recognition much more quickly...
Hope it makes sense

Do you have some suggestions/exercices to help me in this area? Just tactics (which can't hurt either way)?

Thanks a lot in advance!

Like anything else, calculation just takes practice. That is actually one reason I'd recommend checking out this forum of mine https://www.chess.com/forum/view/game-analysis/calculation-training Hopefully, some people here will find it useful. 

Forums aren't the only way to improve though. Here are a few other ideas I've heard to improve your calculation:

1) play non-chess-related memory games and it should indirectly help your calculation.

2) pick a random spot in a GM/strong player game and then try to find the "correct" ideas/moves or simply evaluate the position yourself to see who is winning and why. The idea is that a random place may or may not be tactical or spectacular; this will train you to calculate many candidate lines. Some interesting, some boring, some positional etc. 

3) Play the chess.com "guessing game" (I think it is called). Watch a stronger player in live chess (usually titled players) and "guess" the move before they play it. You compete with other viewers to guess more moves correct. It is fun for blitz, but if you are watching a rapid game: it can also improve your calculation ability. 

@JesuisCrescendo

The precise calculation has little to do with pattern recognition and more with acknowledging the tactical details in a position, as these become the guides when choosing candidate moves to fulfill an idea and when comparing the cost in tempos.

Pattern recognition works as a "hint" or guideline if you prefer. Take the "Greek gift" sacrifice (Basics: Qd1, Nf3, Bd3 against Kg8, Rf8, pawns on h7, g7 and f7): The main idea is 1.Bd3xh7+ Kxh7 2.Nf3–g5+ Kh7–g8 3.Qd1–h5, threatening 4.Qh5–h7++. But from here White has to calculate accurately if the Black King can escape after moving his Rook or if he can but conceding material or the consequences of 2...Kh7–g6 instead of 2.Kh7–g8.

We can also consider the following game played earlier today and GM Bojkov's notes after White's 22nd movement:

Both players missed the idea of trapping White's Queen in the center because both were playing in "default mode" (as there are few examples –patterns– of a Queen trapped in the center with that many –seemingly good– escape routes). And, certainly, the Queen escapes but at the cost of several concessions, topping them the weakening of e3, increasing the "situational value" of Black's dark squares Bishop and Knight (after Nc6–e7–f5, with coordinated activity over e3).

Would Ding or Radjabov missed that idea in a classic game? Not likely. The reason is that any strong player knows that advancing the pawns forfeits chances to repel invaders economically, and White's Queen on c5 certainly qualifies as an invader. Thus, Qf7–c7 to support b7–b6 would have been in both players' maps.

So, to clarify, fast chess is mostly about who's better at guessing –supported by calculation at key moments. OTB is mostly about going deep to extract the most from a position –supported by precise calculation every time.

thanks for those comments, quite interesting!