Ask HN: What factors affect the result of an interview?

I've gone through a couple of tech interviews (over phone & onsite) for internships, and I got rejected by all of the companies I interviewed with. I know I did poorly in some of the interviews, but I think I did well in some of them problem-solving wise. Because none of the companies gave me detailed feedback, it's really hard for me to improve my performance against a black box. Maybe HN can provide some help?

I'm kind of socially awkward, so I tends to just focus on technique questions and ignore other things. Does that negatively impact my chances? Should I have small talks with the interviewer? Like talking about my impression on the city and workplace and stuff? Or maybe discuss some of the recent news in tech circle? In one of the interviews, an interviewer (a really nice person) showed me around the office. And I wasn't able to make any comments. Was that bad?

How detailed should I be when I talk about the project I current work on? Is sounding unenthusiastic bad? (Working on same project for years can be tiring) Does this really matter?

And what about the way I solve technical problems? Since I had a bit of ACM/ICPC training, sometimes I was able to see the optimal solution straight away. Should I go right to that? Or should I lie and start with a inefficient solution and start from there?

How bad is it to take a hint from the interviewer when you stuck? Is it really true that the most important thing is the way I think, not whether I can solve the problem?

And what about when the interviewer tells you to ask them a question? At first I always tell them I had nothing to ask. After several rejections I started to make up questions to ask, but the result didn't change. So maybe this doesn't matter at all?

Sometimes it really fells like I'm the interviewee in that Monty Python sketch.