Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TD1]

The casual use of force from almost the beginning is disgusting.

[voliol]

Does that video show anything "new" or is it purely more of the same flavor of horrible, this time from another (camera) angle?

[Dostoevsky]

I know this is like expressing surprise that the sun came up this morning, but goodness those youtube comments. Sigh.

[Reelya]

The number of times he begs not to be shot at the start, i think the cop was thinking "heh well i won't *shoot* you". I can't watch more than that much.

[WealthyRadish]

I know this is like expressing surprise that the sun came up this morning, but goodness those youtube comments. Sigh.

It's better if you sort by "newest", so it isn't just a wall of bots and comments bots upvoted.

[Reelya]

I know this is like expressing surprise that the sun came up this morning, but goodness those youtube comments. Sigh.

It's better if you sort by "newest", so it isn't just a wall of bots and comments bots upvoted.

I don't know if they do this, but they should: one account gets 1 point worth of upvotes per video, so if you upvote 50 comments, that counts as 0.02 upvotes each, for the purposes of sorting. Also, scale your vote's effect by the percentage of each video you actually watched.

[Bumber]

Does that video show anything "new" or is it purely more of the same flavor of horrible, this time from another (camera) angle?

It's footage from before they pinned him to the ground.

I don't know if they do this, but they should: one account gets 1 point worth of upvotes per video, so if you upvote 50 comments, that counts as 0.02 upvotes each, for the purposes of sorting. Also, scale your vote's effect by the percentage of each video you actually watched.

Nothing more bots couldn't fix!

[bloop_bleep]

I know this is like expressing surprise that the sun came up this morning, but goodness those youtube comments. Sigh.

It's better if you sort by "newest", so it isn't just a wall of bots and comments bots upvoted.

I don't know if they do this, but they should: one account gets 1 point worth of upvotes per video, so if you upvote 50 comments, that counts as 0.02 upvotes each, for the purposes of sorting. Also, scale your vote's effect by the percentage of each video you actually watched.

So then a random guy who watched and liked a couple youtube videos gets 50x the vote influence of someone who watches and likes lots of videos? Sounds like a problem.

[Iduno]

https://twitter.com/LAPDHQ/status/1291180758963556354

[Reelya]

Does that video show anything "new" or is it purely more of the same flavor of horrible, this time from another (camera) angle?

It's footage from before they pinned him to the ground.

I don't know if they do this, but they should: one account gets 1 point worth of upvotes per video, so if you upvote 50 comments, that counts as 0.02 upvotes each, for the purposes of sorting. Also, scale your vote's effect by the percentage of each video you actually watched.

Nothing more bots couldn't fix!

The main thing it would help is that number of bot upvotes is O(n) rather than O(n²). Also make it so each account has a hidden cool down on upvotes

[wierd]

Linear rather than exponential would be a vast improvement.

[da_nang]

Leaked bodycam footage from George Floyd incident.

Systemic police problem confirmed: Incorrect and insufficient training, as well as bad police policies, leading to tragedy.

I expect at most manslaughter and a reenactment of the LA riots.

[Bumber]

The main thing it would help is that number of bot upvotes is O(n) rather than O(n²). Also make it so each account has a hidden cool down on upvotes

Linear rather than exponential would be a vast improvement.

Not at the expense of legitimate human upvoters.

[Reelya]

The main thing it would help is that number of bot upvotes is O(n) rather than O(n²). Also make it so each account has a hidden cool down on upvotes

Linear rather than exponential would be a vast improvement.

Not at the expense of legitimate human upvoters.

It wouldn't really be.

Spoiler: Getting Off topic (click to show/hide)

What it would be saying is that each account had an equal weight on the upvotes for any particular video. So if you came on and read the comments and found one you agree with and upvoted that, you'd have the same weight as someone who went through the comments and upvoted 50 of them. So someone who's more selective with their upvotes, we'll weight them higher than someone who's spamming them left right and center.

However, another thing I'd do to counter bias is to say that upvotes for content further down is weighted higher than upvotes for the top comments. So if you read through 3 pages of comments and upvoted exactly one of them, that upvote would be considered more noteworthy than people just spamming their upvote on the first page worth of comments. It'd still show the actual amount of upvotes given, however the algorithm would arrange them based on these criteria.

Cluster analysis can also fight the bots. For example you could detect that the same accounts were all upvoting specific comments on different videos. Then you could lower the relevance of those votes too. After all, even if those accounts are all legit, they clearly only represent the views of a very specific identifiable subset of people on youtube, and their block of upvotes/downvotes may not reflect the community as a whole

[Doomblade187]

Here's what looks like a good video primary source for Portland protest news: https://mobile.twitter.com/MrOlmos