I hear this sentiment a lot (and notice it in myself, though I have a lot less experience so it hasn't been as long running), and it worries me.
Are you sure it's better? Are you keenly aware that there are no, say, cycles in your coding approach?
I know this year I look back at last year's code and go, "oh boy, why'd I do it that way?" And I feel like it's better, but I'm always a little afraid I'm just chasing novelty. Not sure how to objectively know if it's better.
Don't forget also that familiarity makes things seem more clear. Code you wrote yesterday needs to be particularly atrocious for you to find it hard to understand. Code you wrote a year or two ago might as well have been written by a stranger (really memorable hacks aside).
Perhaps look at code from 2 years ago and 3 years ago, and if possible blind yourself as to which is older, and rate each one.
I'm pretty sure I remember my code. I agonize over it, put all my effort into it. I remember code I wrote in the '80. I remember the code I wrote 7 years ago, and refactored last year. Code is my life.
I've had that sneaking suspicion that maybe I'm not really getting any better, yes. But I have independent corroboration - I've worked with another guy for most of that time, and we've got one anothers' code to look at. What one of us learns, the other learns too. He's not cycling, or just writing bad code. And he says the same about me.
Are you sure it's better? Are you keenly aware that there are no, say, cycles in your coding approach?
I know this year I look back at last year's code and go, "oh boy, why'd I do it that way?" And I feel like it's better, but I'm always a little afraid I'm just chasing novelty. Not sure how to objectively know if it's better.