Two issues here: biased sample
Of course, but it's a pretty big one nonetheless.
and confirmation bias due to a subjective evaluator employing subjective methods.
That's a tough pill to swallow. Especially after being called, directly and through "the grapevine", everything from "overly forward" to "creepy". Of course that has only happened a few times, but it wasn't too hard for me to figure out:
Givens:
1a) Women A through G exhibit physical response Q and
1b) give verbal confirmation X.
2) Women H through Z exhibited the same physical response.
Conclusion: They probably felt the same way even though they didn't confirm it. It seems reasonable to me.
And the few instances of women NOT exhibiting physical response Q coincided with friendliness and a desire to be around me more. But in all cases, the latter category include women in whom I have no romantic interest and who are well removed from the possibility of that developing (relatives, already dating or married to someone else, much older or younger than I am, lesbian, et cetera).
The only way I can reconcile my experience with your rejection of my conclusion is to say that there's something in particular about the women in the area in which I live (the Pittsburgh region). But I wouldn't want to pass a judgment like that. Or that I just have fantastically bad luck - but I don't believe in luck.
Whoosh. That's over my head.
Caetaris paribus means "all other things remaining equal". In other words, unless I can change the parameters somehow, I can expect the past trend to continue into the future. But I don't know which parameters can be changed, or how to change them if they can (except for moving out of the area, but I'm going to be stuck here for a long time).