I’ve generally had more good outcomes than bad, as long as I don’t take on the emotional burden myself.

Some people don’t actually want advice. In those cases, the issue isn’t technical, it’s interpersonal. In my experience, engineers who refuse to hear advice tend to struggle the most for obvious reasons.

Where I’ve gone wrong is taking on the emotional weight of other people’s projects. When I do that, the balance shifts toward more bad outcomes than good ones.

That seems like a good approach.