The code I invent myself is almost always a big mess. When I'm working seriously, I research the answer to the problem. This produces slower results, better c

The code I invent myself is almost always a big mess. When I'm working seriously, I research the answer to the problem. This produces slower results, better code. The other trade off is I'm not working from direct memory or skill, so co-workers have less respect for me, even if "my" solution is better than theirs. This makes me functional years ahead of my experience, but belies a true understanding of the concepts. As I methodically catch up with myself over time, this is less true.