1. Keep your players happy. Having the entire starting squad on "superb" morale before a game pretty much guarantees a win. If your tactics don't suck.
2. Have a tactic that doesn't suck. It's hard to make this work in the beginning (putting everything on often and sliders up to max on attacking? bad idea.), but trial and error is basically all there is to it in FM.
3. Classic tactics are the way to go. Lets you understand and vary play far better than assigning roles which just make you confused.
4. Get a feel for it, especially the media part. Team talks, press conferences and private chats are key, as in point 1 here. I go through these instinctively as i've learned most of the "right answers". It's important to let the game know you're right, eg. it's a good thing if you predict messi to win the ballon d'or instead of, i dunno, jose baxter. In team talks, don't ask the team to go "for the fans!" when playing wolves away - that setting is for big games, specifically at home. If playing a lower-division team in a cup tie, say you expect to win in the conference but don't get cocky.
5. I find the best way to make tactics is to imagine something on paper and get a general idea before you sit down and make it in the program. F.ex. "i want attacking full backs and play down the left flank with crosses from the byline aiming for a powerful libero striker". Go and see how this idea can best be implemented with the tools given in the game.
Handling the game is like learning to use photoshop. Tons of possibilities, tons of way to play. In the end, you're just using a program.
This just off the top of my head. It may seem stocky and boring at first, but it's ridiculously immersive once one starts to understand the big picture. For me, the big wow part is that we don't really know how the mechanics work. In most games, you're told things like: Do task X, benefit factor Y. Or "this boss is an ice creature, i should use fire attacks". Or "this building gives +25% production", and so on. This isn't really the case in FM. Sure, you can try for "fast defenders cause the opposition's forwards are strong and quick", but it says that before every game, and then you must consider if it is a good idea to break up the level of affinity the defenders already have for the way they play - you could do this if you had an idea of the numbers (moving the defender's mentality slider 3 points down makes him 25% better!) but you don't. And there is a ridiculous amount of factors involved in this game. The end result: it's eerily realistic, thusly immersive.
I feel like I need to give it more time.