According to the telegram poll, 1.184 million people voted for Tikhanovskaya. That's over 5x as many as according to the official numbers.
The sample size doesn't matter here, were talking absolute number. It seems pretty unlikely 5x more people (absolute number, not fraction) voted for her in a poll than in the real elections.
(I have not checked the source for the numbers, I've just assumed the above poster used the right ones)
[1] https://en.wikipedia.org/wiki/2020_Belarusian_presidential_e...
Bear in mind that Wikipedia's policy is use "official" sources when in doubt. According to Wikipedia, Assange is a criminal. Snowden is a traitor.
Not really.
They use credible sources as far as they can find them, their numbers on Belarus look like what the international press is reporting.
Wikipedia doesn't 'think' that Assange or Snowden are anything. They have long articles detailing their history, and possibly what some others might think.
Did all the eligible voters who phone-polled actually vote?
Did they poll the way they actually voted?
Some people don't have two phones?
Can we trust that telegram doesn't have a flaw in the means by which it is measured? (i.e. register again with the same SIM, or something like that?)
It's nice that there is a phone number as validation, but this doesn't 'prove' anything.
Edit: I should add, apparently people are registering photos of their votes, which don't jibe with the tally, which is probably a much better indication of problems with voting [1]
Whom exactly benefits from that? It would have to be a nation-state, and they weren't exactly doing much about the situation before the "election".
What is more likely: that the existing dictator and his supporters made defrauded the election to keep power or that some unknown entity defrauded an Telegram poll for not much gain.
Unless Telegram poll's system was completely broken, but you'd expect other large polls to already have revealed that.
Propaganda and narrative is the whole name of the game here.
It's Telegram poll, there are any number of ways it could be messed up, including very easily someone sympathetic at Telegram (though I doubt this), it has no material credibility.
It's a data point, not 'proof'.
... but doesn't really matter how big the poll sample was if the sample was not representative - the results don't have a lot of meaning.
And it's pretty fair to say it was not representative.
In fact, you don't even need remotely that size of a sample if the sample is representative.
Polling is not about sample size, it's about the quality of the sample.