My impression was that any given upvote or downvote was applied by a sapient being, after some degree of conscious consideration of the post’s content. I have yet to see a post including ‘leetspeak’ or egregiously bad grammar with positive karma, yet I have seen several such with negative karma.
Accordingly, I am curious as to why sixteen (or more?) people approved of the post in question.
Right, but don’t those aggregate upvotes and downvotes follow some sort of random distribution from our shared perspective of incomplete information? That’s a sincere question; I am not at all confident in my ability to do statistics or even my basic understanding of statistics.
It was funny! Humor is not discouraged here when it is legitimately funny. Something of sufficiently high humor value, especially if it contains tangential insight, can be modded high.
The individual upvotes are not independent of each other. A post can be thought of as having an “upvotability” related to its quality, snappiness, originality, conformity to group ideas, adherence to conventions etc that can be expressed as the probability that a random (registered) viewer will upvote it. The total upvotes of a post are determined by the number of viewers, the upvotability and chance (people are also influenced by the current karma of the post and their opinion of the poster, but I think we mostly agree that one ought to try to compensate for that).
Calling the fact that a post received a lot of upvotes “poisson noise” implies that you consider random chance a better explanation than a high upvotability and that your posterior for the upvotability of the post is (close to) unchanged (at the very least it implies that your posterior is closer to your prior than to an estimate based on (net- ) upvotes and estimated number of viewers [and an estimate of downvotes]). Does this match your beliefs? Otherwise guessing the password sounds like an accurate description of your behavior.
EDIT: You might also think of upvotability itself as poisson distributed, but that’s a non-explanation equivalent to answering a question about what geological processes cause the Himalaya to be the highest mountain range with “one mountain range has to be highest and that just happens to be the Himalaya”.
Poisson noise.
It wasn’t a single data point. Posting jokes and funny but trivial comments is consistently among the most karma-profitable behaviour. See e.g. this.
and come on!
That sounds suspiciously like a password. Are you saying that upvotes and downvotes are randomly distributed?
Upvotes and downvotes aren’t randomly distributed?
My impression was that any given upvote or downvote was applied by a sapient being, after some degree of conscious consideration of the post’s content. I have yet to see a post including ‘leetspeak’ or egregiously bad grammar with positive karma, yet I have seen several such with negative karma.
Accordingly, I am curious as to why sixteen (or more?) people approved of the post in question.
Right, but don’t those aggregate upvotes and downvotes follow some sort of random distribution from our shared perspective of incomplete information? That’s a sincere question; I am not at all confident in my ability to do statistics or even my basic understanding of statistics.
It was funny! Humor is not discouraged here when it is legitimately funny. Something of sufficiently high humor value, especially if it contains tangential insight, can be modded high.
The individual upvotes are not independent of each other. A post can be thought of as having an “upvotability” related to its quality, snappiness, originality, conformity to group ideas, adherence to conventions etc that can be expressed as the probability that a random (registered) viewer will upvote it. The total upvotes of a post are determined by the number of viewers, the upvotability and chance (people are also influenced by the current karma of the post and their opinion of the poster, but I think we mostly agree that one ought to try to compensate for that).
Calling the fact that a post received a lot of upvotes “poisson noise” implies that you consider random chance a better explanation than a high upvotability and that your posterior for the upvotability of the post is (close to) unchanged (at the very least it implies that your posterior is closer to your prior than to an estimate based on (net- ) upvotes and estimated number of viewers [and an estimate of downvotes]). Does this match your beliefs? Otherwise guessing the password sounds like an accurate description of your behavior.
EDIT: You might also think of upvotability itself as poisson distributed, but that’s a non-explanation equivalent to answering a question about what geological processes cause the Himalaya to be the highest mountain range with “one mountain range has to be highest and that just happens to be the Himalaya”.