Just like anything else that you want to get better at; you get what you put in, no more, no less. You need to study and play the game with the intention of playing your best, learning from your mistakes, and avoiding being predictable.
Read 'The Mental Game of Poker' by Jared Tendler if you want a crash course on how to handle yourself and have a winning attitude. It doesn't cover any poker strategy, but one of the main messages that it has is that the minute you think you know it all or have nothing to improve on in your game, you've already lost. Keep a journal, track your winnings and loses, read books and forums, watch streams, etc.
I haven't come across "The Mental Game of Poker" (so I can't say much on it) but I agree with everything else said. I've always had a natural gift for psychology (in general, not just in poker), so I can't tell a story of how I got this ability. However, I would practice this skill like with everything else. You don't need poker to become a master of psychology. Have you ever sat in a public area (the mall, a party, a park, or wherever else) and just observed people and the surroundings? I know I can't be the only one. I've learned a lot of psychology through doing this and sometimes I feel a bit like Sherlock Holmes. Just be careful to not think like Watson by making funny assumptions in a manner too linear.
For example, what if we observe someone with some animal hair on their sweater? Watson might say something like: "Ah, from this we can determine that they own a cat. Further inspection reveals that this animal hair doesn't look like that of a dog." That is a good start to get the gears in ones' head turning, but we can't even say with much certainty that this is true. What if the owner of the sweater just returned from an animal shelter? They may not even own a cat. This sort of observation and assuming guesses can be a bit of fun, but we must realize that there is a high probability that we are wrong when we are assuming. This can be very destructive to our
poker game if we don't consider these other possibilities.
Let us say that we are in a poker hand: heads up: lots of chips in the pot. Our opponent then scratches their forehead. Is this a tell? What would you think? They must be uncomfortable and Watson would be quick to point out that we should shove all-in because we have evidence that they are weak. Not so fast Watson, what if our opponent merely had an itch? This might be a tell, but then again, it might not. This is why the best poker players combine lots of psychological elements before they make an assumption. They evaluate the situation, the timing, the person and our history with them and many other things. In the exact concept of tell-spotting, it is always best to have multiple
tells on someone to reinforce your certainty on the situation. Plus, there is less of a chance to fall for a reverse tell if you are looking at many things, not just one observation to go on.