vortex
143 posts

vortex
@vortex670683731
apenas um fur que gosta de 3D 🇧🇷 18y
Katılım Nisan 2025
66 Takip Edilen7 Takipçiler

@Bernard03852922 @VascodaGama Até dezembro desse ano ela vai ficar na minha cabeça.
Português

@vortex670683731 @VascodaGama Essa frase tem que ser proibida
Português

Adversário da 5ª Fase da Copa do Brasil definido! 🏆☑️
O Vasco da Gama vai enfrentar o Paysandu, do Pará. Os jogos são de ida e volta, e serão disputados entre as semanas de 22 de abril e 13 de maio. Os mandos serão definidos mediante um novo sorteio.
#CopaDoBrasil2026
#VascoDaGama

Português
vortex retweetledi

não ironicamente isso é um indicativo de que vcs se gostam! se a pessoa sente facilidade em dormir com vc perto é pq ela se sente segura e etc.
𝘇𝗲́ ⁷@jkozao
meu namorado sempre que encosta em mim ele dorme feito bebe e eu acho isso muito fofo pq ele fala q eu sou a calmaria dele
Português

"lésbica gosta de xota , gay gosta de rola "
lésbica gosta de mulher e nb
Gay de homem e nb
"Ah mas a genitália " você só é transfobico
demetrio fodao@krlhdc
Transfobia rolando solta na tml contra os nb de NOVO
Português
vortex retweetledi

Big announcement!
I have created an algorithm which estimates the number of followers that individual accounts gain from their viral Twitter posts. Over the past year, I've been tracking many accounts with suspicious engagement metrics, and I'm finally able to use this data to highlight the suspicious vitality of accounts like @esjesjesj.
I've tracked the hourly followers gained by these accounts over time in addition to the like counts of all of their Tweets, and by cross-referencing this data, I can estimate the proportion of followers gained from any individual tweet. Obviously this is not 100% accurate for many reasons, but it does provide a plausible guess as to how many followers can be attributed to each individual tweet.
To be more specific, I created a model for the distribution of likes tweets gain over their lifespans, which factors in the virality of the tweet and the time of day the Tweet was sent in order to create a plausible representation of how many likes each tweet gained during each hour that it was active. I can cross-reference this data with all the other tweets which were active during the lifespan of any individual tweet in order to estimate the total follower growth that can be attributed to it.
Obviously this system isn't perfect. It relies on estimates for engagement longevity. Measuring the actual hourly like counts would cost me tens of thousands of dollars a month, so that isn't an option. Another thing I'd like to do is measure the engagement of the reply sections to expose how suspiciously barren they are, but once again that would cost a lot of money in API usage.
The results from this are fascinating. Firstly, as you might have expected, accounts like @esjesjesj exhibit unbelievably suspicious follower growth in proportion to the virality of the tweets. The only account I've tracked which exhibits more suspicious virality is @DreamLeaf5. The only other accounts which come remotely close to such a low score are ones which already have such a critical mass of followers that it makes sense for them to have a lower follower growth rate because so many of the people who would be interested in their content are already following them (such as @ShitpostGate). It would make sense for meme accounts in general to experience less follower growth than average, but they still greatly outperform accounts like @esjesjesj and @DreamLeaf5.
However, accounts like @esjesjesj are the complete opposite of this. You would expect them to experience immense follower growth because they are a medium-sized account which constantly experiences viral tweets. In reality, if this engagement was authentic, his account would be experiencing more growth because the hundreds of thousands of people who like his tweets on a daily basis would actually start following him.
For a specific example, @esjesjesj's most viral tweet with over 1 million likes resulted in an estimated follower growth of just 400 followers. Yes, you heard me correctly, one of the most viral tweets ever posted on this platform only netted him 400 new followers.
So what am I going to do with this data? From now on, I am going to be focusing on posting content which I have measurable data to back up my claims on. There are limitations of this new system which prevents me from tracking tweets/accounts retroactively. Essentially, I need to be tracking an account ahead of time in order to calculate an estimate for how many followers were gained by a post. If you have any suggestions for accounts with suspicious metrics which I should watch, let me know and I will add them to my tracking system. Starting out, I will focus on posting the most egregious examples of posts with suspicious engagement from the handful of accounts I have been tracking.
One final disclaimer that these are rough estimates using very limited data.
Please share this post to support me and justify all the hard work and money that went in to create this!

English

@danibecouto1898 Antigamente eu te achava chato pra krlh e tava até bloqueado, mas hj em dia eu curto cada post seu kskkk mt foda rambo
Português

@forcajovem70 É PAPO DE IR NO CT AMANHÃ E FAZE DISK O KRLH A 4 TMB
Português

"volume negativo" então pega aqui ó

Rez é o Shadow@RezDavi
Eu achando meu cabelo lindão nessa foto e a galera só notando no volume 😔💔
Português

@cavacogamer FALTA SÓ A CARALHA DA PLACA DE VIDEO PFVR NAO LULA
Português










