Honestly, I suspect HN is starting to fuzz votes the way Reddit does. Some of the voting I've seen on my own comments lately makes zero sense, including weeks-old comments of perfectly neutral tone. It isn't worth worrying about.
Don't discount better and better "I can't believe it's not human" aping behaviour of personal bot armies of HN accounts - there's clear and obvious spam green accounts, and a steady drip of bland comment, weird voting behaviour, biege textual tone accounts that evolve through green and into background noise.
The admins potter about classifying coarse behaviour and looking to reject spam, voting blocs, overly weird AI comments, etc. The creators make their bot accounts less obvious with random votes, etc.
Comments made by recent accounts appear with green names until a grace period passes.
HN prefers "legit accounts" (subjective) - good faith comments from real people, reasonable uses of alias and spun up fresh accounts for regulars to say things without being part of their main history, etc. New commenters welcome.
There are obvious and less obvious dark patterns of bad faith account creation.
The admins do a pretty decent job, sweeping cobwebs without hitting real genuine people is an artform ...
Returning to the main GP point - there's a lot of low key background churn activity that can result in "inexplicatable" votes, some from bots and some from the general case of "people are strange".
If you see my message quickly enough (I can't remember how long the green lasts, so not sure when the deadline is for this example), here's the most recent green I saw to give you an example:
jonathan7977, https://news.ycombinator.com/item?id=46264260 (linked directly to their comment but it shows green on the main thread also - https://news.ycombinator.com/item?id=46263317)
