Teenagers are most likely wired for later nights because they have reached initial fertility. They will sneak off in the night to fool around, especially if they haven't had a child yet. At the risk of sounding patronizing, this is how babies are made. Adolescence is the age when women would first become pregnant, and the ancient's concept of marriage was probably not the same as ours. (There is evidence of monogamistic tendencies in humans for a long time but that isn't synonymous with marriage.)
Teenagers are biologically wired to want lots of sex and they tend to do it at night. I would put a lot of money on the fact that that hasn't changed in millennia.
Obviously, this is a critical aspect of the process that allows for the continuation of the human race. Humans didn't have the consistent ability to wait to reproduce until 30 as is common now. And probably not the desire too; I can't see career-building interfering.
The initial party must have ended pretty quickly. Poor ancient teeenagers.
(Edit: considering this speculative is reasonable and factually correct. But speculating there's been behavior change is currently more speculative. What reason do we have to think teens had daytime sex in ancient times and our behavior has changed? What about for mature adults? Never heard of an existing tribal culture that mostly has sex during the day, and that's how we get much of our anthropological evidence.
Humans tend to have sex at night, and the most fertile, desirious and energetic would be wired to spend more time at the time people have sex doing it: that would be selected for because of our primary biological directive. Is there a more logical evolutionary reason?)