My recommendation: lots of save-scumming. I have autosaves set for every 2 turns, and I manually save several times per era (the new UI that shows the game mini-map on the load screen is a godsend for tracking the progress of my crap-tons of save files!). If something bad happens (like I lose a valuable unit, or an AI surprise war decs me, or another civ beats me to a wonder, or I realize that I mis-timed the adoption of a policy, etc.) I go back a few turns and see what I could have done to avoid that outcome and get a better outcome.
Sometimes, I'll reload a save 5 or 6 times and fail to get the outcome that I want. Sometimes that means I go a few turns further back. But sometimes it also means that I have to accept that I just played poorly and I can't avoid that bad outcome without sacrificing hours of play-time and progress. This helps to teach me when I've over-extended myself, and what moves are worth the risk or the cost. In future games, I try not to repeat the same mistakes again.
Also, this might sound obvious, but: you need to make sure that you're paying attention to the game's rules. There's a lot of nuances in unit movement, zone of control, line of sight, city sieges, and so forth. I've been playing the game since release, and have been writing strategy guides for it (
http://www.megabearsfan.net/category/Strategy.aspx/?tag=/Sid+Meier's+Civilization) and participate in a bi-weekly podcast about the game (
http://polycast.civfanatics.com/), but I'm still learning things.
Just one example: putting a city into the "siege" state (by surrounding it with your zone of control, so that it can't heal between turns) makes a HUGE difference when attacking a city. (
https://steamcommunity.com/app/289070/discussions/0/312265526374151235/) If you're failing to that because you don't know about the siege rules (or you forget to encircle the city before you start attacking it), then you're going to have a harder time. That siege rule is an easy rule to overlook, and I'm sure many players either don't know about it, or forget about it when they try conquering other civs.
On a related note, don't take old assumptions for granted. Throughout my playtime with Civ V and with Civ VI vanilla, I would never settle a city on top of any resource. In previous iterations of the game, doing so would destroy the underlying resource and reduce the tile's yield. I wasn't sure if that would happen in Civ V and VI, so I just never risked it. Then I started watching YouTube let's plays and saw that the general consensus on optimal strategy was to try to settle on a plains hill tile with a luxury on it. You work that tile for free for the entire game and get the full yield. But it was just something that I had never tried because I had assumed that settling on the resource would clobber it.