Hi,
Pattern recognition is really only one of many skills that you need to find success in a game. Pattern recognition only allows you to quickly identify potential ideas, plans and tactical shots in the position. From there you'll definitely need your calculation skills to back up what your intuition is telling you. Every position is different, so in the end, you'll have to do the calculation to find out exactly what works and what doesn't
Would harder tactics problems help increase calculation skills? I had been expecting that around 1900 on puzzles would require calculation. But mostly this hasn't been the case.
No! Calculation skills and tactical skills are entirely different skill sets. Unfortunately, calculation skills are much harder to train, and can really only be done with either the help of someone who knows what he/she is doing or lots of hard work.
The reason this is the case is because calculation involves both looking for tactical and positional motifs and combining them into a coherent move/plan.
For the more positional side, I recommend this video over here in order to come up with a best move and to help you learn how to calculate! https://www.youtube.com/watch?v=AmFDHhhPhsI
I would have thought harder puzzles would help develop your calculation and visualisation skills, unless I mistunderstand calculation. You need the puzzles to be hard enough that your fast pattern recognition cant help you so you have to find candidate moves based on checks, captures, threats, then calculate the puzzles strongest replies, etc. and visualise the position as it changes. To solve hard puzzles , you'll be evaluating multiple lines in this way, probably 4 to 6 moves deep. If you are doing 5 to 10 puzzles every day like this your calculation skills should strengthen over a few months. Calculating out a mating attack when the king has multiple escape squares and defensive resource is hard work. I'm surprise you've got to 1900 without this.
@MarkGrubb Unfortunately, puzzles and actual game application are completely different. When doing a puzzle, you know there's a right move and thus, your mind begins to spot for motifs and ideas you've learned when doing puzzles. In a real game, you don't know when a tactic exists and they usually don't if your opponent is playing decently enough. I've seen some people with 2000+ ratings in tactic trainer but still be around 1000 rating in live chess as they can't get to the positions with tactics in them.
Instead of just practicing tactics (still important to do them!), it's important to garner a positional understanding of the game as well.
...
Follow my Club in the following link https://linktr.ee/royalchess32 and participate for a prize of $ 10 to the first place
Siga mi Club en el siguiente enlace https://linktr.ee/royalchess32 y participe por un premio de $ 10 al primer lugar
I am trying to get to 1400. I have a problem in that my puzzle rating is around 1900 and yet I normally solve most of the puzzles by only seeing one move ahead, rather than by calculating. This appears weird and paradoxical. It seems I am somehow using pattern recognition to solve them rather than calculation. Which makes me wonder just how high I can go OTB by just using pattern recognition rather than calculation.
I know 2 guys that made it to USCF Expert based on nothing but tactics, and nutty openings like the halloween gambit, and stuff like that. They do well until they have to play anyone Master and up. Then they get destroyed. So yea..you can become a very good player based off of nothing but tactics.
No, im playing my friend and he plays randomly. His name is ooglyboogly251
If you have a puzzles rating of +500 (compared to your chess rating), you are in a fine spot: most problems will definitely require calculation from here onwards. More and more calculations. I my case, I am about +900 right now, and I feel I hit a wall 
1160 isn't a bad score. I mean....any four digit number is all right. I guess??
I am trying to get to 1400. I have a problem in that my puzzle rating is around 1900 and yet I normally solve most of the puzzles by only seeing one move ahead, rather than by calculating. This appears weird and paradoxical. It seems I am somehow using pattern recognition to solve them rather than calculation. Which makes me wonder just how high I can go OTB by just using pattern recognition rather than calculation.