More on writing to not be disliked · 11:22pm Apr 22nd, 2013
You may have read my earlier blog post, Writing to not be disliked, where I said that because most down-votes seem to come from people who have some random bug up their ass rather than people who have a valid reason for disliking a story, the best thing to do to get good ratings is to write something safe and inoffensive.
I've written a few long Amazon reviews. I noticed my Amazon reviewer rank was pretty low: 14,630,784. I wondered how low that was. I went to the reviews page for Twilight and sorted the thousands of reviews by "usefulness". Then I went to the last page, which was full of stupid reviews (some complaining about shipping, one by a person who hadn't read the book but said that all long books are bad), mostly misspelled, many in all caps. Then I checked the Amazon reviewer ranks on the bottom 20 reviews.
The lowest I found was 14,637,000. That puts me in the bottom 0.0005% of Amazon reviewers, according to the ratings my reviews have gotten.
"Not useful" is the Amazon equivalent of a downvote. I looked to see where I'd gotten all my Amazon downvotes. All my positive reviews combined had only one downvote. All but one of my downvotes were from mixed or negative reviews. One was for a graphics card that came with adware on the installation disk, instructions written for some other installation disk, and that crashed my computer whenever I tried to run a game or watch a video: 0 up, 4 down. That confused me. But most were for books or videos that I didn't like, especially the artsty ones with hipster rep.
So the day has already arrived in which balanced, thoughtful reviews are likely to be buried at the very bottom of the heap, among the dreck, by majority vote.
Recruit downvotes for things that are popular. Recruit upvotes for things that are unpopular. Prey on the things that are near you, shoving them down so you can
feast on their delectable fluidsbetter yourself slightly by comparison.1027368
That seems, somewhat unfair though. Some things are popular because they are good after all.
Really makes me question how much weight I should give to customer reviews.
1027380
Hah! Good? Do you believe in objective measures of quality, or do you believe in the Evil League of Evil!?
"Bad Horse, what is best in life?"
"To crush your rivals, see their stories downvoted before you, and hear the lamentations of their watchers."
1027380
The example given in the blog post is reviews for Twilight. It seems unlikely that vote ratios are being done fairly. In any case, I'm not recommending dishonest voting. Hunt for unfair popularity (such as Twilight). Hunt for unfair obscurity (such as Bad Horse reviews on Amazon). Don't fear to push dreck into the dirt when you encounter it along the way (such as the reviews that Bad Horse reviews are unfairly grouped near on Amazon).
I've noticed the same. People with helpful things like "Hey, this A/V switcher for your TV is a cheap peace of crap and here is a valid list of reasons" get all the "not helpful" votes. The review that says "OMG Best thingy ever!" get upvoted.
This is the graphics card maker sockpuppeting you in revenge, I'd say.
1027419
On the rare occasion when I'm browsing product websites, I usually make a point of downvoting meaningless hosannas. Yes, great, it saved your life. Wonderful. I'm happy for you
, mister sockpuppet. How did it do that? Moreover, how did it do that better than its competitors would have?It's all a matter of how you write reviews. Let me demonstrate:
Ermagerd this was the best thing ever!!!!!1!!1! I mean, it like, totally, you know, ermagerd. On the first try! It was so awesomely awesome, I had sex with it! Twice! I traded a small child and a cute blue-eyed kitten for it! I would have married, it, but it's not legal in this state. Then it stole my car and took all my money. Still, I'd buy it again.
You can go on like this nearly forever.
That top spot is practically yours.
--admiral biscuit.
I can count the number of conversations I have had on this site in regards to positive things I have said on one hand.
I can't even count the number of conversations I have had regarding critical or negative things I have said.
I would rather be interesting than be liked.
1027385
HA! A fair point sir!
1027387
Ah, I thought you were also referring to stories on FiMfiction.
1027508
The same hunting style can improve vote quality on FiMFiction too. It's not exclusive to product reviews. The problem just isn't as desperate here. I have come to advocate topkicking. Hunt the top stories for things that don't live up their hype. Stories that maintain a high rating in spite of some downvotes are more impressive than stories that skate to the top without anyone objecting to them. Worse, when people are afraid to downvote, bad stories rise to the top on the power of fanatic appeal. There are still terrible stories on the site that have some niche appeal (or eldritch momentum) guaranteeing them an abundance of upvotes.
I think the phenomenon can be traced to the fact that people tend to read reviews[1] not to decide what to read/buy/watch so much as to get validation for having read/bought/watched it. You'll find this tendency reaches its platonic ideal in video games.
[1] Or, rather, those who read a lot of reviews and bother to downvote/upvote them.
1027384
xkcd knows how much weight you should give them.
I had no idea that the reviewers were ranked, and anyway I've done only four so I wasn't expecting much in the way of placement.
#2,650,954. About where I figured I'd stand here, actually.
1027524
I think somebody did a study to that effect years ago about car purchases--people would find positive reviews about their car after they bought it.
My philosophy is spiderman's.
1027368 1027499 1027524
Remember folks, Dante was hated, Voltaire was hated, Ayn Rand had her seething haters. If you write in order to not be disliked, you won't write anything to be loved either.
1027771
Past tense with Rand's haters?
My personal experience matches up, both offsite and here. The one downvote on my story so far was from someone who disliked my characterization of Luna.
1028115 Well "was" is past tense as well. And to be fair, Rand still has her fans today. I loved how she could write antagonists who were both pathetic and threatening at the same time. I was quite simply engaged by the story. And this is from the guy who usually hates mystery stories with a seething passion.
I find it better to always speak my mind. There's always something to criticize, and sugarcoating is an emergency option. But I drop a purely positive comment here and there, when I'm merciful.
Quality and popularity rarely go hand in hand. Fans of something will see detractors as "attacking" the object of their affection and will either dismiss or retaliate against any such attempts at a negative review.
For the video card, my gut tells me that was an attempt by the manufacturer to quash a bad review.
Regarding your Amazon experience specifically: what is the date of your most recent review?
I picked a random political book, to get a source of polarized opinion, and randomly clicked through to find a gold standard of bad Amazon reviewers: someone with NOT A SINGLE UPVOTE, and a handful of downvotes, out of six reviews. Their ranking? 14,562,620. (Not an aberration: here's another at 0% in the 14.5M range.)
Given that you are ranked below authors who are 0% helpful, there is something really fucking with the numbers here.
Researching Amazon's system suggests that they're giving reviewers a sort of web-2.0 metric: they say "The more recently a review is written, the greater its impact on rank". That plus your sub-zero rating implies they're applying negative points for age. Both of the 0-percenters above wrote reviews earlier this year.
Another data point: my most recent review is from 2005; I've got 190 upvotes out of 220 total; and my reviewer rank is 10,337,855.
--
Edited to add: This reviewer is on the Top 10,000 list, and has a score of 73% helpful (807 upvotes out of 1,100). I'm at 86% helpful and in the bottom third.
Top 10,000, 60% helpful.
1029848 Those top reviewers have hundreds of ratings. Something favors lots of reviews over highly-rated reviews.
My most-recent reviews were 2 that I did a week or two ago.
1030102
Reviewer rank 594,381. Only 21 reviews total, 100 ratings total, 56% helpful, last review July 2012.
I am having a great deal of trouble coming up with any coherent theory which accounts for all of these data points and still puts your rank at 14M+, unless you're rocking something nutty like +5 helpful out of 50 ratings (that might actually put you at a lower confidence interval than +0/6 because of the higher number of data points). I really want to know how you're glitching the system now.
1030102
1030199
Maybe Amazon has some kind of customer-tracking metric which prejudices the system in favor of reviewers who successfully sell product. Good reviews sell products. Are there "good" bad reviews at all? Do "good" bad reviews include links to other products? The objective quality of the review may be irrelevant to the reviewer rank. If Bad Horse turns customers away from the site, then Amazon doesn't want him there.
1034470 A clever idea. Maybe if a customer clicks on "This review was useful", and doesn't buy the product, they count it against you.
1034670
The even more depraved corollary is that if someone clicks on "This review wasn't useful" and buys the product, they may count it in your favor.