Actually, there is evidence to support my model: the fact that classic front page is very similar to the normal front page, except with a different sorting order.
This evidence can be interpreted in (at least) two ways, (a) that most voting occurs based on the front page, or (b) that the front page topic themes, and by extrapolation user tastes, have remained consistent over time.
I (based on my own voting, and now yours too, but also the number of votes front-page articles get vs stuff that doesn't make it to the front page, evident from resubmission successes etc.) would guess that hypothesis (a) is more likely than hypothesis (b); but in any case, the fact that there's an alternative explanation for the similarity in the two front pages means that PG's hypothesis of (b) is on shaky ground.
This evidence can be interpreted in (at least) two ways, (a) that most voting occurs based on the front page, or (b) that the front page topic themes, and by extrapolation user tastes, have remained consistent over time.
I (based on my own voting, and now yours too, but also the number of votes front-page articles get vs stuff that doesn't make it to the front page, evident from resubmission successes etc.) would guess that hypothesis (a) is more likely than hypothesis (b); but in any case, the fact that there's an alternative explanation for the similarity in the two front pages means that PG's hypothesis of (b) is on shaky ground.