JackbyDev comments on Which one is boosted and which one is real

Bikini Bottom Twitter

4,903 readers

286 active

0 users here now

Ahoy, me buckos! Welcome to Bikini Bottom Twitter! Your digital reef for the latest salty gossip and treasure tales! And while you're at it, be sure to drop by the Krusty Krab for a delicious Krabby Patty so I can get yer mon- err I mean, 'cause they're the best treat under the sea!

Rule 1 - This is Bikini Bottom Twitter, all posts should be Spongebob related in "(Old-School) Twitter-like" form

Rule 2 - Political posts, as long as it follows rule 1, will be permitted, so long as you behave yourselves.

Bikini Bottom Municipal Code §33-07: Anti-Tankie Ordinance Residents are prohibited from circulating tankie ideology or other authoritarian propaganda on Bikini Bottom Twitter. Offenders will be permanently banned from BPT by the BBPD faster than Plankton is ejected from The Krusty Krab.

Rule 3 - Please no reposts within the last couple days, at least

Rule 4 - All posts should be at least above a "Squirdward-krusty-krab-shift" level of effort

Rule 5 - Be chill, be a Patrick not a squidward.

founded 3 years ago

moderators

[–] JackbyDev@programming.dev 2 points 1 year ago (2 children)

I think about this a lot. Let's assume for a second all reviews are legitimate (I know they're not, but not detection isn't what I'm talking about). I hate when you sort by ratings you get 5.0 (1 review) at the top. Why isn't there some semi objective agreed upon way to do this? It doesn't need to be perfect. Search engines aren't perfect but we use them all the time. Something like 4.9 (10 reviews) should be above 5.0 (1 review).

permalink

fedilink

source

hideshow 2 child comments

[–] funkless_eck@sh.itjust.works 2 points 1 year ago (1 child)

In NPS (net promoter score) tracking - you know the survey "would you recommend this to your friends and colleagues"

it's worded and scored 0-10 so that you can assume everyone that scores 7+ is happy and anyone below isn't.

It's a psychological thing that wouldn't work if it was thumb up or thumb down.

In 5 star systems 4 and 5 should be considered promoters and 3 and below should be considered detractors, but again you should bifurcate and dichotomize the output so you see how many scored 4>= and how many <=3

permalink

fedilink

source

parent

hideshow 1 child comment

[–] JackbyDev@programming.dev 1 point 1 year ago

Thanks for the information, I'll look further into it. But even in those scenarios, say you get a score of like 90% recommending it with 100 people, but another with 89% recommending it but 1,000,000 people. I'd say the million people with 89% score should be higher. That's more what I'm talking about. It's like the more people that rank highly the more confident the result is. (Which is why for this discussion I said assume bots don't exist lol.)

permalink

fedilink

source

parent