Pay attention and know who the players are at your table. For example, one player, TAF, is always tight, and another is fishy when calling bets with poor cards because he thinks everyone has a better hand, etc.
Knowing these things will help you decide when to bluff and how much of a bet to make based on what you think the range of hands they have maybe. mamasboyct.com is a popular site.
Don’t let losing streaks affect you or put too much excitement into winning; rather, treat every decision as an expected value equation (EV = Probability x Value).
This includes not getting overly upset if someone calls your big bluff and wins without folding once they see that their hand isn’t the best.
You can use this to inform your decisions in future scenarios by paying close attention to how often they will fold when their cards don’t improve and betting less or not at all on that expectation.
Be as objective as possible without sacrificing creativity, being reactionary, etc. In other words, don’t just be a robot and always do what you would have done if it was the first time you had seen any given a hand; rather, take advantage of situations where you can outthink your opponent(s).
For example, the most skilled players are those who can tell whether they are getting action from others because they want everyone involved or because someone wants to isolate another player. This type of knowledge is tricky to acquire but understood by some of the best.
If you see a player is very consistent in his style, then bluffing becomes easier because he won’t fold to your bets unless he’s seen them multiple times already (e.g., someone who checks the flop every time and calls all turn cards except twos or less).
Try calling with marginal hands when they bet preflop to trap him into thinking you have something and might give up later in the hand if he has a better-made hand than you do — know that your call could become costly if someone else also raises so there may be a value in folding instead.