Wouldn't really say I'm hugely successful or anything, but here's what I've learned from 10 years of making games:
Join game jams, and always do something you've uncomfortable with as your entry. It's good to try out new ideas to widen your skillset and views in general, and jam games are either forgotten instantly (if they're bad) by everyone including their maker, or they're a good prototype for making a "real" game later. It's the dev equivalent of throwing stuff onto the wall and seeing what sticks, basically. You also get experience at prioritizing in a stressful situation, which you WILL go through both as a hobby developer and an industry developer... might as well try to grind for some stat points in that right away, right?
Be careful about "publishers". I signed a contract with one once, and ended up having to do all the media-attention work myself in the end anyway. So I basically signed a contract that gave me less money and some exclusivity clauses to be aware of that did absolutely nothing towards getting me that big breakthrough I was dreaming about.
You'll need to work for your attention. I started off searching for the '#indiedev' and '#gamedev' hashtags on Twitter each day, following 10-20 people that seemed legit, and hoping they'd follow back (and then unfollow them later if they didn't). It's not really the best or most ethical way of getting networking done, I realized later (right now I'm posting witty comments and actually talking to people there instead) but when you're completely new it doesn't hurt to get your numbers up in any way you can. But actually networking with people gives you more attention in the long run, it's just a slower process. Probably doesn't hurt to use both a little bit of pure-advertisement spamminess on the side, just don't overdo it to the point where people get fed up on you or it'll actually hurt you in the long run.
Write to-do lists on actual paper. Sounds a bit silly, but I find it has plenty of uses. It's incredibly satisfying to crumble up a fully done to-do list in your hand and crushing it with your raw power and stuff, for a start, and it gives you a reason to look away from the screen regularly (which help reduce eye strain). It also means you can jot down ideas even when you're doing something else, like cooking or sleeping, which helps you from forgetting spark-of-the-moment ideas. It's also faster to doodle up diagrams and sketches on paper than using a CAD program, and doing those things more often on your todo lists can make some things much easier to do (e.g. coming up with level design ideas), increasing your overall productivity.
Ideas are cheap, execution is expensive. Always look for ways to cut corners. A typical example of this is the AI design maxim: "An AI isn't smart, it just appears smart in the situations the players see it in". Trying to make things perfect not only is impossible to truly succeed at, because you can always find ways to make something incrementally better, it takes a lot less time and effort than making the minimal viable product and just polish the rough edges a bit. And most players won't even notice the extra effort, either. So if it ain't broke, don't fix it. The last thing you need is unnecessary work nobody will care about in the long run. Also, on a related note, beware of feature creep. You'll never get games complete if you keep deciding to add new stuff to them when they're almost complete - several of my biggest project failed in the end because I kept adding new stuff instead of finishing them, and I either ended up making the engine so convoluted the game broke down into buggy messes, or I got bored at the projects and wanted to try new stuff. Don't do that.