Question for any xitter experts.

[BoomerSperg]

How does this happen? My redditor friend made a comment that gets 8x the views as some random comment our street shitting queen made that he was replying to?



How does any of this fawkin' work!

 

[CuntFucker .]

I'm glad that cunt isn't allowed to see her children. That might give them a chance.

 

[Jims_Maroon_Pants]

Why is she screaming for anal in the last part of the sentence?

 

[guest]

Twitter impressions are notoriously unreliable. I don't know if this has changed since Elon took over, but probably not. For twitter you only really care about interaction. Likes, comments, retweets.

As of years ago, when I was researching this, the impressions don't really filter out any kind of hot linking, seo bot scraping, etc. They also do some amount of fuzzing to impression metrics. This is similar to how youtube does it. If you give exact numbers people can reverse engineer things and game the SEO system much easier.

 

[NoBacon]

Twitter impressions are notoriously unreliable. I don't know if this has changed since Elon took over, but probably not. For twitter you only really care about interaction. Likes, comments, retweets.

As of years ago, when I was researching this, the impressions don't really filter out any kind of hot linking, seo bot scraping, etc. They also do some amount of fuzzing to impression metrics. This is similar to how youtube does it. If you give exact numbers people can reverse engineer things and game the SEO system much easier.

Some dude created a fresh account, private, no followers, random string username. He was getting like 30 views per tweet when he was the only person who knew it existed. Even nitter and other places scraping API data doesn’t account for it, because I’d imagine they can’t access private accounts.

 

[FurBurger]

View attachment 141148

Fawk yeah! PCMCIA modem and flip phones; you, Sir, have all the toys!

 

[Meownaw]

View attachment 141148

I assume the red hand is used to rape the ladies on the bus.

 

[NoBacon]

If they are not purposely obfuscating the numbers by adding a random multiplier or something. Then its likely just picking up some internal api calls or something. WIthout knowing their architecture, this could be an accident or by design. Given my experience so far, its likely just technical debt that's too expensive to remove from a cost-benefit perspective.

We do these cost benefit analyses all the time. There's a bug with the impression counter, its counting internal api calls. Then your boss, if he's good, will ask something like. "Ok well whats the risk and impact? How many users does it affect etc." If the answer is , "a small number of users will notice if they have very low engagement. The percent is negligible for any accounts that rely on impressions for ads, or anything important.

Unless its a quick fix, then there's no business case to spend the engineering time on fixing it.

My bug list to fix has cases that are 4 years old and pending because there’s a $0 dollar cost saving to fix it, just ignore the emails from the users it bothers.

This is the way. It’s only really boomers who are angry the data elements don’t tab in order and are confused there isn’t a full rework to make it so every time they press tab it descends in the same order it’s presented. It’s not like there’s weekly crisis aversion to manage that keeps the entire operation actually working, the boomers tabs need to be tackled because it upsets Linda.

It’s never being looked at, you fat boomer retards. I’ll give you a case number and laugh at how you put it in a sticky note and keep bringing it up.

 

[cardwell]

Some of my tweets have zero impressions, even after they are retweeted