self-worth, and probability

I dont have much to say about this topic, but I was walking today and that phrase was stuck in my head "regression to the mean". I've read a few books on probability in the past year and I have a few take aways

1 - uniform random sampling is hard

2 - most times we only end up with conditional probabilities instead of truly representative

3 - repeating experiments increasing probability that you get closer to true mean

I guess these things are pretty obvious. But today I was thinking of self-worth (or self-confidence or self-esteem) and regression to the mean. For myself, example, I think I started with a level of self-worth that was a bit lower than my true worth, as I got older, or as I did more experiments, I began re-evaluating my worth hopefully closer to the true mean. Sometimes I overestimate, and sometimes I underestimate. If I was using a numeric scale, I wonder if I would get a normal distribution to my true mean. Am I supposed to get a normal distribution, I should be remembering something about central limit theorem right about now, but I cant recall it properly. Actually I do remember central limit theorem, just not sure if my sampling of my own self-worth would be considered as sum of independent variables. Anyways, I need to go back and read some more probability because obviously I suck at it and probability and statistics is a cool knowledge base to have.

On the smart side, I've been hearing this debate claiming that the reason many experiments are being disproved these days is because of regression to mean or because time is the changing factor? I would like to believe that it's regression to mean, because if it's time then it makes it much harder to prove anything because time is always changing. Wow, this post made no sense. BOOOOOOOOOOO.