Gadgets News Technology

Periscope is using viewer juries to fight trolls

Written by Jon Fingas

Chris Ratcliffe/Bloomberg via Getty Images

With most social internet services, getting rid of trolls is usually a matter of reporting a post or blocking the offender. But how do you do that in a fast-moving livestream service like Periscope? By asking viewers for help, that’s how. Periscope has introduced a moderation system that creates “flash juries” whenever a comment is up for dispute. If someone flags a message as abuse or spam, a few random viewers are asked to vote on whether or not it’s a problem. If the majority believes it is, the offender faces a minute-long ban on comments; a repeat offense mutes the person for the rest of the broadcast.

You should see moderation in effect starting today (May 31st) through app updates.

The system isn’t mandatory. Viewers can opt out of voting if they’d rather not participate in a mini trial, and broadcasters can turn moderation off if they’re comfortable with the occasional outburst. And Periscope is quick to note that this isn’t the sum total of its anti-abuse efforts. You can still kick people out of broadcasts, limit viewers to those you know and report ongoing problems. The new approach primarily tackles Periscope’s trickiest abuse problem: those hit-and-run comments meant only to cause some temporary grief and ruin an otherwise happy stream.

About the author

Jon Fingas

4 Comments

  • I implemented something similar on my shitty little hand-made message board well over a decade ago.
    Well actually I had a better system. Telling a troll they can’t post for a minute, how do you think that makes them feel? Well my guess was they’ll storm off and annoy somebody else, wait out the minute typing some abusive essay on their censorship, or change platform and (in this case) take to twitter proper.
    The first rule was always, “don’t feed them”.
    Just let them carry on posting without a clue the vote has taken place – but simply make their posts only visible on their phone. From their side they’ve not been judged or silenced, simply nobody appears to be responding to them.
    You can tweak this further if you want, on one side if some people like the post, let them still see it and reply. On the converse set up a google image search to the worst thing you can think of, and embed zero-sized linked-images in their personal pages.
    From there it builds.. clever troll will be paranoid and start setting up alternate accounts.. so clever admin starts using IP address rules tied to triggered cookie-flushes to avoid harming people on shared networks and browser fingerprinting..
    Oh it was fun in the early 2000s…
    I’d have thought a modern phone app has much more scope for mischief. You could fire off a “troll-alert” to app users geographically near them (every time they load the app). Maybe enhance it with media from their twitter-feed to provide easy identification. Maybe sound off a klaxon?
    I’m reasonably sure this is why I’m the lawyers blocked my purchase by Facebook and my billions… fine, ban them for a minute. W007

Leave a Comment