Puzzle Improvement but no game improvement

When you look at a puzzle, you know that there is a tactic. If you look long and hard enough, you'll likely find it.

In a game, there may or may not be a tactic, and if you spend 15 minutes looking for it you're gonna be in time trouble.

Puzzles are fun but don't expect doing them to result in huge improvement. Having said that, they will improve tactical awareness, and are a much better use of your time than playing at a speed that is too fast to think properly.

Puzzles just make you aware about what you could in certain positions. The main thing you might be missing is your middlegame

At the 700 level (where the OP is), every single game is decided by tactical blunders. Players don't have any meaningful strategical and positional skills, and those wouldn't be important anyway.

A more likely explanation is that the OP isn't really getting better at tactics, his puzzle rating is mostly increasing because it was too low previously. But keep it up, and something good will happen sooner rather than later.

Smothered mate is an example where tactics training really shows a giant difference. If it ever happens in a game, an experienced puzzle solver will spot it in a second. On the other hand, someone who never saw it will never spot it. No chance. They won't even look at it.

Also, solving puzzles has another big upside. Not only will you spot the tricks that you know, but you will probably have a much better chance to spot the tricks that you don't know. Solving puzzles improves calculation skills, and they will help in every situation.

I have bad puzzle score. And im 1500 (ish). Positional play is the future!

I think so...
I made comments 5 days ago that strategy is important (#3 and #4), but after that I seriously thought if they were right advice for OP. I still think those comments are important, but at 700 level I think it was vague and unrealistic. (So I think I have to admit that my advice wasn't helpful. Rather, I'm worried if that kind of advice makes chess too difficult.) #3 #4 are probably most valid for 1200-1500 ratings.

(to OP)

In this post, it seems like answering to another question is first: "How to deal with tactical situations."
 While "think about every possibilities" is a common advice, the problem is that there are so many active minor pieces in middle games, making it difficult to find which move your opponent will make. Your opponent can move his knight to unexpected square, make an unexpected fork, or make a discovered attack with check. It's hard and ambiguous to expect what could happen. Should we calculate every possibilities?
 In fact, even strong players don't think about all possibilities. In the same position, strong players can find good moves faster. In Blitz, they are even able to find good moves in a much shorter time. How is this possible? What is the difference between beginners and strong players? Is it just that they calculate better?
 No, it's possible because they're used to restricting the number of good moves that their opponent can make. They're not thinking about every possibilities, they are only thinking about a limited number of valid possibilities. (So even experts sometimes experience facing unexpected moves)
 So now the important question arises: "How can you narrow down the number of moves that your opponent can make in a tactical situation?"

Actually, you just need to find a move that could be a threat when your opponent plays it next time. Other moves are things that you don't have to think about right now.
In most tactical situations, the biggest threat is minor pieces. queen and rooks can also be a threat, but less so than knight and bishop. So the first thing to consider is your opponent's activated minor piece.

To do this, you need to get used to the movement of pieces.

Possible moves for knight is octagon shape with the knight at the center. When you see your opponent's knight on the board, you can reduce a lot of 'unexpected knight movement moments' if you are constantly conscious of the octagon.
 But you don't have to think about all eight of these moves. Eight situations are still too many. Instead, think of tactical moves that you should pay attention to right now. In the position where the opening has ended and the middle game has started, you usually only have to pay attention to four movements when you're watching your opponent's knight. Four movements towards your territory.

For example, let's talk about this position. What movements should we consider about knight on c6 here?

The answer is these three movements. Ne7 is possible, but it's not a movement that needs to be dealt with right now. Even if your opponent does that, you can think about it then. So what kind of movements can knight on f6 make this time?

The answer is these two things, you don't have to consider Nd5 or Ng8.

When you're looking at your opponent's bishop, draw the arrow from bishop diagonally, especially considering the main diagonal that bishop occupies.

This is another situation, now your opponent's bishop's way is blocked by his knight. So is this the right way to draw the arrow?

No, you have to draw like this. It is his piece that blocks his bishop's way. Therefore, opponent can move the knight and attack with the bishop at any time. (Discovered attack.)

So what should be drawn in your head if you see a position like this?

It should be drawn like this. And imagine each position after your opponent's piece moved. Think about how to respond to each situation.
 After looking at all minor piece movements, you can find that there are no clear chance of tactics here. Then you can now consider that your opponent might have another move, like O-O or a6.

After looking at the board like this, follow this thinking process:
1) Is there any threatening movement that needs to be prevented in advance?
2) If yes, can I create a bigger threat right now?
3) If no, how can I stop my opponent's threat?
4) If there is no threat right now, what can I do? (strategy)
 If you train this process of thinking repeatedly, you'll be able to do this automatically.
I hope my comments could help you this time.

My puzzle rating boosted from 1100 to 1900 ...But when I was 900 in rapid my puzzle rating was 1150..Now I am 1240 on rapid and 1900 on puzzles ..I can assure that doing right amount of puzzles everyday can surely boost your rating...