You can watch it in action sometimes, with amusing results. Some very active posts do, in fact, vanish from the front page with surprising speed, whereas less-active posts can literally linger for days.

While your claim is plausible there is a lot of confirmation bias, among other biases, when it comes to our idea of what what see on a socially algo'd front page.

I'm not saying you are wrong but I'm saying my confidence starts quite low with this kind of idea. (e.g. the cause could be something different like manual moderation; the effect may just be confirmation bias and not exist; etc.)

Qualifier: I haven't looked into the algo at all, for all I know it could be openly published by YC/HN. I'm mainly responding because I feel like it is important to TRY to be self-aware of our own biases. Emphasis on the TRYing and asymptotically approaching 100% confidence (Qualifier to the Qualifier: it is also important to be at peace with decisions based on <100% confidence lol)