The one part that always gets me is building stuff that isnt profitable. Its definitely more important at higher levels and something I need to get my head round. Its been useful because I'm analyzing this gamne more and spent a whuile last night cataloging my errors as I saw them. I'll' play some turns later and see what improvements I can make.
I am definately NOT and expert, but I am looking at my first win on Emperor soon. It will be long and bloody, but I think that a win is in the bag. This will be a bit repetative with vmxa, but somethings that I have started to do that I never paid much attention to are:
1. I alwasy assumed that having a marketplace in every city was the best course of action, regardless of cost. Now I pay much more attention to how much gold is being added to the treasury and how big the town will grow before making the commitment. A lot of the time, size doesn't even matter. Make those unhappy workers into taxmen or something else of use - at least in corrupt cities.
2. I find that buying techs are cheaper than researching it myself. This was a big step for me. No need for those libraries and universities. This might not always be true, but that is why you need to analyze the cost/benefit. If you decide to add a library/university (for the purpose of research, not culture), check how many beakers you will receive. Low beaker towns means no library etc (and probably no other improvements if it is that corrupt).
3. Trade is much more important. You can generally ignore the AI in Monarch and below if you want. You are generally strong enough to take the lead with only a few good purchases. This is no longer true. In the game I am currently playing, I had the opportunity to take control of the trading. I waved at it as it passed by. Now that I see the errors of my way, I am playing catch up - which is why this win is going to be long and bloody. Lesson learned. Check trading each round, no matter how tedious it may seem. It is even more tedious later - and less fun - when you are trying to catch up from behind.
4. Maximize gpt. This seems like a given anyway, but I found that I am squeezing every gold I can find now. I never understood why you would sell a building you already own, but that upkeep can be the difference between having another cavalry or artillary unit and having a library. If no one is reading the books, sell it off for a cannon. Exactly how you reap those shields from a libary I don't know - maybe the cannon fires books. Guns for books anyone?
Edit - 5. When a player like vmxa suggests that you should read an article, take their advice. Although reading won't teach you everything, it will be very helpful. Thanks vmxa and everyone else that takes the time to help new and developing players!