Reactions to this post have been pretty lukewarm. I’ve been thinking about information sharing and feedback and may write up those thoughts at some time, and this seems as good a time as any to see how much info I can glean from the voting feedback system.
Based on the only 2 metrics I see (vote-count and total score), it has been mixed but not enough to call it controversial. Based on my total karma change, it has been slightly more popular with more established users, but there’s probably not statistical significance there. Information that wouldn’t be available to me if I had managed to be less emotionally invested in how helpful I was is the history of up and down movements—but since I did check, I know two additional facts − 1. that the score has bounced around 0, perhaps suggesting corrective voting, but also very plausibly random, and 2. that most votes have been strong votes, giving it a few more controversial points, though I don’t have a baseline to compare to. Finally, while technically not part of the voting system, I see a “Front Page” tag, indicating at least 1 moderator thought it was at least somewhat useful/relevant.
I should separate out the conclusions that I can draw from just the given info and common background knowledge vs those that are conjecture or require additional knowledge.
Just the voting info: Currently: 5 votes, 0 karma subtract the auto-self-strong-upvote to get 4 votes, −2 karma. −2,-1,-1,+2 or −2,-2,+1,+1 are the two possible paths I see. Either helpful to 1⁄4 or slightly helpful to 2⁄4 and quite unhelpful to the other 2. No read count, so can’t distinguish between low readership or low reaction rate. General background knowledge + low reactions → boring or confusing. Moderator approval + low score → niche or hard to understand.
That’s about the limits of what I can glean from the voting info, so the rest of the post will be conjectures pulling from outside knowledge, gut feel, and the human ability to see patterns whether or not they exist.
Conjecture/outside info: 1st place hypothesis: My best current guess is this: I didn’t target a specific audience, which is effectively being guilty of my own #5 Expecting short inferential differences. I didn’t write a long post building all the pre-requisite ideas to bring anyone who hadn’t been pondering this up to speed, and it may be possible that to people who have pondered this long enough, my conclusion is old hat, so it would only be helpful to a small group of readers in the middle. I reasoned that by not spending text explaining all the pre-requisite thoughts, I was saving the reader time and that by being short, the post didn’t have to be THAT helpful to justify the time it cost a reader. Maybe I was wrong and it’s just SO unhelpful that it isn’t worth it’s relatively short length. Or maybe the voting system skews feedback. It could be that readers getting the “well that was a waste” feeling isn’t very sensitive to exactly how much time was wasted. It could also be that since everyone only gets 2 points to vote, a post that hits a small target audience will get a bad score even if it was immensely helpful to a small portion of readers.
2nd place hypothesis: Seeing several highly rated LW posts, they are “punchier” than mine. A highly rated LW post might look like “How to be persecuted 101″, which lays out several commonalities between historically persecuted groups, many of which seem counterintuitively to be in those groups’ control, and then ends with “and if you REALLY want to be persecuted, don’t forget X.”. Reading it gives a feel of “I learned something most people don’t know while reading a witty post”. Here, I didn’t flesh out a problem scenario to let the reader imagine why they might care to have this knowledge, so it feels more like: “Hey, want a free wrench?” “Um, why would I want a wrench?” “I’ve found wrenches useful—you never know” “Um, thanks? I guess?”
3rd place hypothesis: Brains and minds are core topics on LW. Maybe most LW readers are very familiar with the idea that brains <> minds, so it gives a “what else is new” feeling. And when the only answers are: “Well here’s 5 examples of the fact you already knew, plus a phrase I like as a reminder” it feels pretty low value.
Honorable mention hypotheses: I notice that my “questions” have historically done better than my “posts”, so maybe posts are expected to contribute more value. Maybe I didn’t spend enough time editing it. I notice a lack of comments, so maybe readers want posts they can engage with more than ones that just tell them a thing and move on and I didn’t make this very discussion-open. Maybe my thoughts just aren’t as great as they seem in my head and I should just let smarter people contribute. Maybe the use of the word “magic” rubbed people wrong, given the community’s general stance against magic. Maybe my style was just too weird, or I came across as “I know better than you”. It’s likely that none of my hypotheses are correct, or that my idea of there being THE REASON is wrong and in fact each reader had their own separate reason for their reaction. Either way, making more conjectures at this point would be clearly stretching.
Reactions to this post have been pretty lukewarm. I’ve been thinking about information sharing and feedback and may write up those thoughts at some time, and this seems as good a time as any to see how much info I can glean from the voting feedback system.
Based on the only 2 metrics I see (vote-count and total score), it has been mixed but not enough to call it controversial. Based on my total karma change, it has been slightly more popular with more established users, but there’s probably not statistical significance there. Information that wouldn’t be available to me if I had managed to be less emotionally invested in how helpful I was is the history of up and down movements—but since I did check, I know two additional facts − 1. that the score has bounced around 0, perhaps suggesting corrective voting, but also very plausibly random, and 2. that most votes have been strong votes, giving it a few more controversial points, though I don’t have a baseline to compare to. Finally, while technically not part of the voting system, I see a “Front Page” tag, indicating at least 1 moderator thought it was at least somewhat useful/relevant.
I should separate out the conclusions that I can draw from just the given info and common background knowledge vs those that are conjecture or require additional knowledge.
Just the voting info:
Currently: 5 votes, 0 karma subtract the auto-self-strong-upvote to get 4 votes, −2 karma. −2,-1,-1,+2 or −2,-2,+1,+1 are the two possible paths I see. Either helpful to 1⁄4 or slightly helpful to 2⁄4 and quite unhelpful to the other 2. No read count, so can’t distinguish between low readership or low reaction rate. General background knowledge + low reactions → boring or confusing. Moderator approval + low score → niche or hard to understand.
That’s about the limits of what I can glean from the voting info, so the rest of the post will be conjectures pulling from outside knowledge, gut feel, and the human ability to see patterns whether or not they exist.
Conjecture/outside info:
1st place hypothesis: My best current guess is this: I didn’t target a specific audience, which is effectively being guilty of my own #5 Expecting short inferential differences. I didn’t write a long post building all the pre-requisite ideas to bring anyone who hadn’t been pondering this up to speed, and it may be possible that to people who have pondered this long enough, my conclusion is old hat, so it would only be helpful to a small group of readers in the middle. I reasoned that by not spending text explaining all the pre-requisite thoughts, I was saving the reader time and that by being short, the post didn’t have to be THAT helpful to justify the time it cost a reader. Maybe I was wrong and it’s just SO unhelpful that it isn’t worth it’s relatively short length. Or maybe the voting system skews feedback. It could be that readers getting the “well that was a waste” feeling isn’t very sensitive to exactly how much time was wasted. It could also be that since everyone only gets 2 points to vote, a post that hits a small target audience will get a bad score even if it was immensely helpful to a small portion of readers.
2nd place hypothesis: Seeing several highly rated LW posts, they are “punchier” than mine. A highly rated LW post might look like “How to be persecuted 101″, which lays out several commonalities between historically persecuted groups, many of which seem counterintuitively to be in those groups’ control, and then ends with “and if you REALLY want to be persecuted, don’t forget X.”. Reading it gives a feel of “I learned something most people don’t know while reading a witty post”. Here, I didn’t flesh out a problem scenario to let the reader imagine why they might care to have this knowledge, so it feels more like: “Hey, want a free wrench?” “Um, why would I want a wrench?” “I’ve found wrenches useful—you never know” “Um, thanks? I guess?”
3rd place hypothesis: Brains and minds are core topics on LW. Maybe most LW readers are very familiar with the idea that brains <> minds, so it gives a “what else is new” feeling. And when the only answers are: “Well here’s 5 examples of the fact you already knew, plus a phrase I like as a reminder” it feels pretty low value.
Honorable mention hypotheses: I notice that my “questions” have historically done better than my “posts”, so maybe posts are expected to contribute more value. Maybe I didn’t spend enough time editing it. I notice a lack of comments, so maybe readers want posts they can engage with more than ones that just tell them a thing and move on and I didn’t make this very discussion-open. Maybe my thoughts just aren’t as great as they seem in my head and I should just let smarter people contribute. Maybe the use of the word “magic” rubbed people wrong, given the community’s general stance against magic. Maybe my style was just too weird, or I came across as “I know better than you”. It’s likely that none of my hypotheses are correct, or that my idea of there being THE REASON is wrong and in fact each reader had their own separate reason for their reaction. Either way, making more conjectures at this point would be clearly stretching.