Poll of the Day > Twitter has a right-wing promotion bias just over 150% in Canada...

(+) Yeah! (+)

Yup!

While there is a slight ring-wing bias across the board, it's a massive 150% in Canada...

Is this just to arbitrarily keep the 50/50 split or what

Yellow posted...

Is this just to arbitrarily keep the 50/50 split or what

I have no idea, it was just something in their transparency report

Any idea how it even happened?

Yellow posted...

Any idea how it even happened?

My sinister inner edge-lord makes me think it's a hard push to keep the conservatives in power in Canada... but honestly I have no idea

It's in every country (analyzed) except germany.

It's either a glitch or an oversight probably. Kind of hard to keep that conspiracy under wraps with a team of (mostly) liberal programmers.

I don't know, though. I guess no one does. I might be able to decipher what's going on, but they haven't released the data yet, and I don't know how much I want to learn the inner mechanisms of the "algo".

One possibility is that almost news outlets (outside of Toronto) are owned by right-wing biased Chatham Asset Management.

As such, NEWS on Twitter will have a right wing bias... and if Canadian SOMEHOW use twitter mostly for news... that would be where the bias comes from

found the source:

https://theconversation.com/twitters-algorithm-favours-the-political-right-a-recent-study-finds-175154

It's 167% in Canada :D

What exactly is "right-wing promotion"? Because Twitter became the new place for Tumblrites after the porn ban, but I guess their shitposting isn't really promotional, even when it's prominent.

Okay so looking at it, it's just saying right wing politicians are shown to people more. This makes sense when the left want to engage with right wing politician's tweets directly, but the right will more just talk about it in right wing spaces in a circlejerk.

Kyuubi4269 posted...

What exactly is "right-wing promotion"? Because Twitter became the new place for Tumblrites after the porn ban, but I guess their shitposting isn't really promotional, even when it's prominent.

Okay so looking at it, it's just saying right wing politicians are shown to people more. This makes sense when the left want to engage with right wing politician's tweets directly, but the right will more just talk about it in right wing spaces in a circlejerk.

twitter has a system where they amplify tweets so they show up higher on peoples feeds; this means what the study calls right wing content is boosted 167%... so if you are completely politically neutral and have no celebrity interests or any internet footprint whatsoever and go on twitter, you are 167% more likely to see Conservative party content if your location is detected as Canada.

These weightings are important, things like emergency broadcasts/missing childrens amber alerts get many thousands of times boosting so that people are more likely to see them.

Left wing parties in canada do enjoy a 43% amplification in tweet importance as well

Lokarin posted...

if you are completely politically neutral and have no celebrity interests or any internet footprint whatsoever and go on twitter, you are 167% more likely to see Conservative party content if your location is detected as Canada.

...which makes sense if in aggregate the average Canadian is 167% more likely to engage with a right wing post than a left wing one. I'm assuming Canada is fairly left and that right wing people are less likely to use Twitter to begin with, so rage posting and retweeting to amplify the mobbing will boost its visibility.

Ultimately Twitter is an advertising platform, so the algorithm isn't tuned to what people will enjoy the most, it's the one that gets most traffic, and when large groups mob a post, that post has much more advertising value than one people like a lot but don't reply or retweet much.

Kyuubi4269 posted...

...which makes sense if in aggregate the average Canadian is 167% more likely to engage with a right wing post than a left wing one. I'm assuming Canada is fairly left and that right wing people are less likely to use Twitter to begin with, so rage posting and retweeting to amplify the mobbing will boost its visibility.

Ultimately Twitter is an advertising platform, so the algorithm isn't tuned to what people will enjoy the most, it's the one that gets most traffic, and when large groups mob a post, that post has much more advertising value than one people like a lot but don't reply or retweet much.

either that or conservatives are more likely to click adds

Like I said, I don't know why the bias is there or its effects... only that it exists cuz of the article

Kyuubi4269 posted...

...which makes sense if in aggregate the average Canadian is 167% more likely to engage with a right wing post than a left wing one. I'm assuming Canada is fairly left and that right wing people are less likely to use Twitter to begin with, so rage posting and retweeting to amplify the mobbing will boost its visibility.

Ultimately Twitter is an advertising platform, so the algorithm isn't tuned to what people will enjoy the most, it's the one that gets most traffic, and when large groups mob a post, that post has much more advertising value than one people like a lot but don't reply or retweet much.

This. Rightwingers are crazy and evil, so their posts will generate more outrage, which drives more traffic, which generates more ad revenue. Social media doesn't care that they're accelerating the downfall of society, so long as they get their cut.

papercup posted...

This. Rightwingers are crazy and evil, so their posts will generate more outrage, which drives more traffic, which generates more ad revenue. Social media doesn't care that they're accelerating the downfall of society, so long as they get their cut.

Which also means the "social media is silencing right-wing viewpoints!" thing is objectively nonsense. Pretty much every platform has some degree of right-leaning bias in the outcome of its algorithms, which is entirely the opposite of what the free speech pearl clutchers believe and claim.