Pages Menu
TwitterRssFacebook
Categories Menu

Posted by on Aug 1, 2013 in Miscellany, Philosophy, Politics | 28 comments

Block Bot, #ReportAbuse and a Twitter ‘panic mode’

One thing I didn’t make clear, and should have, in my previous post on the #ReportAbuse campaign on Twitter is that even though I had concerns around the effectiveness of a “report abuse” button, I did agree that Twitter should do something about the trolls. It’s not clear to me why having a particularly thick skin, or staying away from discussing topics that invite abuse, or crudely, simply possessing the property of not-being-a-woman should pretty much guarantee experiencing a friendlier time on Twitter.

A concern around ‘false positives; in terms of being identified as trolls (i.e. people being labelled as abusers too readily, or on too subjective a set of criteria, or for entirely personal reasons) is that reputational harm could be fairly costly, even if you aren’t aware of what it has cost you. For example, in the skeptic/atheist/humanist community, you might be passed over as a potential speaker at a conference, or contributor to a book or journal, because someone points out that you’re a dodgy character, as evidenced by your inclusion on some list or other.

However, I’ve got no issue with people maintaining such lists, so long as they are clear about what the lists are for, don’t add people to such lists capriciously or arbitrarily, and also, allow for some mechanism to be removed from the list.

One such list is James Billingham’s (or oolOn’s, to use the name some will be more familiar with) Block Bot, that has recently enjoyed more widespread attention than usual thanks to the BBC clip embedded below. The Block Bot is clear on what the list is for, but to my mind errs in including a publicly-readable list of offenders. I have no problem with a self-selected community deciding to universally pay no attention to a list of Tweeters, but when these lists are available publicly, passers-by might conclude that there is broad consensus that they are abusers, rather than this being the decision of a specific community. I’d also suggest that the explanatory notes on what the levels of abuse mean should be linked on the same page as the list of abusers, to make the context clearer.

And then, one can – and should – raise concerns about how people get added and taken off the list. It looks like a single report from an authorised blocker is sufficient to get you added to the list, even as a Level 1 (i.e. you’re classed as maximally offensive) abuser. That seems too easy, allowing for a label to be applied in a moment of hasty or capricious judgement, rather than as a result of careful consideration. And then, being removed from the list seems to require posting on the Atheism+ forums – and that seems a high bar, seeing as the sorts of people who might get listed on the Block Bot overlap substantially with people who won’t believe they’ll get a fair hearing there. It also smacks of a guilty-until-proved-innocent model.

There are some folk on the list, especially at Level 3, that I don’t think belong there at all – but this wouldn’t be a concern except for the fact that, as I say above, this list isn’t viewable on an opt-in basis, and also might be construed as indicating worse ‘crimes’ than it does, given the lazy way in which many of us make judgements. You’re told that there’s a list of abusers, and you go and see person X listed as an abuser – especially when levels 1, 2, and 3 aren’t clarified alongside the list, you might leave with a false impression as to the severity of the abuse perpetrated by certain individuals on it.

This is a central point in Damion’s post on the BBC NewsNight story, where he rightly criticises the BBC journalist for not making the distinctions mentioned above, of their being different levels of abuse, perceived or otherwise. The (undoubtedly many) people who visit the BlockBot site subsequent to that programme airing will be primed to make exactly the mistake mentioned above, namely thinking people are worse abusers than they are. Especially in light of the exceedingly vile forms of abuse that are on people’s minds right now (think Criado-Perez or Lindy West), it doesn’t seem fair that someone who has tweeted cynically about Atheism+ stands a chance of being perceived as aligned with the sort of troll who would tweet rape threats.

I’ve gone on too long, so to conclude: in case you didn’t spot it, Flay has made a very interesting suggestion with regard to controlling abuse on Twitter, and I encourage you to read his post detailing what he calls “panic mode” for Twitter. In summary, enabling panic mode would only allow mentions from people you follow to appear in your feed, allowing for a respite, and also simultaneously flagging your mentions for monitoring by Twitter, so as to highlight people who might be violating Twitter’s terms of service.

As I said at the top, something does need to be done, and Twitter has now introduced a button to report abuse on their iOS app, with the same functionality to follow for other platforms. But hopefully, they’ll keep thinking about what mechanism might be best for controlling abuse, and the ‘panic mode’ idea seems worthy of further consideration.

P.S. Two notes after checking some details with oolOn via Twitter

  1. A misconception that’s sometimes arisen around this Block Bot is that people who are added to it are reported for spam, and could therefore stand a greater chance of being blocked suspended by Twitter. This is false – they are hidden from your timeline (though you can still follow them on an individual basis if you like), but would still need to manually be reported as spammers or blocked if you choose to.
  2. The source code for the bot is freely available, in case you wanted to bake your own list of people to hide.

[EDIT] re. point 1 above and the strikethrough, oolOn just said:

[EDIT 2] In paragraph 4 above, I point out that the Block Bot is most useful to a specific community, rather than as a general tool (in the latter case, it could mislead, and itself lead to abuse). Tim Farley expands on this and other issues with the Bot in this very worthwhile post.

  • Not sure about that misconception at the end. I’ve heard people say that they can no longer follow people they want to, because they’re actually blocked in the usual sense.

    Also: https://dev.twitter.com/docs/api/1.1/post/blocks/create seems to be what does the blocking (I checked the source, blockem.php). That’s just an ordinary block, as far as I can tell. I think on Twitter oolon was telling you that if you are already following someone it doesn’t block them for you- i.e. it passes over them. It still blocks the others. The code does this by subracting the list of people you follow from the list of people you’re about to block. It then proceeds to block each person on the new list:

    $connection->post(‘blocks/create’, array(‘user_id’ => $x_users_to_block[$x]));

    Still, my beef isn’t really with the bot itself (people can use it if they really wish to let others choose who they block) – it’s more that the BBC program implied that the bot was in some way relevant to the rape threats that have been thrown around recently, and described it as a ‘shared list of abusers’. It was poor journalism, all the more so since after I complained politely to the journalist in question I was myself blocked, in spite of the fact that on that very Newsnight program the BBC presenter was criticising Twitter staff for blocking complainants on Twitter.

    I agree that something needs to be done. It isn’t ok to expect people to just take it, and it certainly isn’t ok to tell them to ‘get off Twitter’ as I’ve seen from some. Panic mode might be a good idea – better than what there is now, at least.

    • Hopefully oolOn can clarify, but users_to_block might mean “users to hide” – my impression is that the people you’re talking about, who say they are blocked in the usual sense, have been reported by individuals, co-incidentally (albeit, related in that the list might have brought them to mind) to the BlockBot list.

      Agree entirely with your 3 & 4th paragraphs.

      • Oh, users_to_block was just the variable name oolon gave that list. I was talking about the post(‘blocks/create’) part. That does what is referenced in the documentation linked above, i.e. an ordinary Twitter block (it’s a Twitter API call, not an arbitrary name oolon made up – like it is with ‘users_to_block’).

        Still, oolon knows more about this than I do (I’m not a programmer) so I defer to his expertise!

        • ool0n

          Indeed the users to block are those not in the current blocks and not people the person blocking is following. So blocks once and never anyone you choose to follow.

      • Richard Sanderson

        Don’t expect ool0n to clarify anything. He’s been asked several times to clarify what “abuse” landed EllenBeth Wachs on the block bot. He can’t answer.

        We know the reason why – he’s a liar and a troll.

        • Calling someone a “liar and a troll” doesn’t conform to discussion policy here. You might of course not have read that, but please desist in future.

          • Richard Sanderson

            PZ Myers described him as a troll. Don’t worry about it.

            Oh, and I am most certainly right.

          • MosesZD

            But he’s both. It’s like calling me ‘white.’ Yeah, I’m white and oolOn meets all the defintions of ‘troll’ and ‘liar.’

            His block bot is deliberately engineered to report blockee’s as ‘spammers’ to have their accounts disabled by Twitter for ‘spamming.’ It’s right there in the code. No ifs. No ands. No buts.

            And yet oolOn claims otherwise and hides this fact from others.

          • Tim Farley went through the code, as is detailed in the link above. And only level 1’s are reported. I haven’t seen oolOn say anything to the contrary. Provide the quotes, show the lie, in other words. As for troll, that’s a subjective and clearly derogatory term. Saying “but he’s both” for troll is entirely circular, as there is no possibility of “evidence” except on subjective criteria. Whether you’re right or wrong, it’s abusive.

          • It’s all being discussed in the comments there, in case you hadn’t read them: http://skeptools.wordpress.com/2013/08/01/block-bot-twitter-report-abuse-ool0n-atheism-plus/#comment-1186 . If you – like the person he’s engaging with there – think that he’s lying, fine. But please restrict the claim to that, ie. person x is lying about y, and here’s why, rather than ‘x is a liar’.

    • An Ardent Skeptic

      The BBC reporting tarnished the reputations of everyone on the Block Bot list as not only an abuser but as slimy. Paul Mason stated that he was only going to block the “super slimy”, the Level 1ers on the Block Bot list. To my mind, that casts everyone on any level of the list as slimy but with varying degrees of sliminess. The definition of “slimy” has very negative connotations about someone’s character and should never have been used in regards to Oolon’s Block Bot list.

      Paul Mason has since given an equivocating response about the various levels of the Block Bot list and what those levels mean. He has not apologized for having tarnished everyone on that list by his use of the words “abusers” and “super slimy” in the original piece.

      I also find the BBC’s equivocating response, about not having mentioned names or given the address for the Block Bot website, completely unacceptable. Based on the banner that was shown during the news report for the Block Bot website it is simple to find the website. And, it was possible to read some of the names on the list despite the fuzzing out that was used. We are supposed to accept these excuses for bad reporting while also accepting the excuse for Paul’s use of “super slimy” because it appears in the description of the Level 1ers on the Block Bot site which flashes very briefly on the screen during the broadcast? We weren’t suppose to see the banner for the site (which is in big bold letters) or the names on the list, but we were supposed to be able to read all the small print about what the different levels mean when it appeared on screen for a couple of seconds?

      This is the type of reporting I expect from the ‘Daily Mail’ not the BBC.

      • I’m level 2 yet found i was blocked by Mason when i tried to tweet him for the 1st time ever. Kinda proves the lie to the claim that he was only blocking “The worst of the worst”, whatever THAT means? [My Twitter handle is @OffensivAtheist btw]

  • Richard Sanderson

    The most sinister aspect if that Oolon controls WHO gets added. As we can see, that includes people he has disagreements with, dissent against the FTB narrative, etc. People who have not being abusive end up being added. His spambot is a like a large fishing net that kills dolphins.

    I’ve repeatedly asked Oolon to provide evidence of the abuse of just two individuals (EllenBeth Wachs and Barbara Drescher). He has failed to respond. That is just two of the people on the list, and there are many more on there in the same position.

    The question we have to ask is: would Oolon add one of his chums or FTB buddies to the bot if they fell foul of the vague but wide-ranging “rulez”. Evidence suggests the answer is “no”, given Oolon’s snarky dismissal of a claim one of his buddis made a sexist tweet.

    You are added to the list for ideological and political reasons. It is a way of demeaning and depowering those who dissent – a tactic we are all too familiar with in totalitarian regimes.

    • Sure, oolOn (and, he suggests in the FAQ or elsewhere on the site) and others who have authority controls who gets added. As I suggest above, if this is done in an arbitrary, unprincipled or capricious (etc.) way, then I’d feel aggrieved if I was one of those that were listed.

      However – we can still ask “what’s at stake?”. If you’re right about “the FTB narrative” and the “totalitarian regime” – and I think these generalisations far too broad – then the sorts of people who would ordinarily use the block bot would themselves have blocked a bunch of the people on the list in any case, wouldn’t they? So the block bot simply provides an efficiency – you’ve got an ideological disagreement either with those who use the bot, or with those who block person x or person y.

      If the harm to those who are unfairly blocked (leaving aside the huge issue of how to define that) amounts to only losing out on having a certain number of people (those who use the bot) see their tweets, that’s too trivial a thing to get upset about. However, if – as I ask above – it means losing out on opportunities of various sorts, or having reputations harmed, that’s another. And for all the complaining about injustice here, I don’t see much evidence that this has to do with anything more than hurt feelings.

      To summarise, there are two issues: a) you can think that oolOn or whomever operate on dodgy ideology in making their lists; b) you can ask if that matters, except in terms of the quality of their judgements. On (b), I’m not convinced that it does, and the fact that this got BBC screen-time shouldn’t be allowed to cloud the issue too much – that will be forgotten, very quickly, as things tend to be much of the time in this attention economy.

      • Richard Sanderson

        So the block bot simply provides an efficiency – you’ve got an
        ideological disagreement either with those who use the bot, or with
        those who block person x or person y.

        It’s much more than that. As we can see, it is there so various influential bullies can point to this list and say “look, this is a list of abusers – these are the bad guys”. The Bot is part of their “threat narrative”, and is used to hide their own victim blaming, harassment, and bullying.

        • I haven’t seen that happen myself, but would agree that that would be an inappropriate use of the Bot’s list, for the same reason as the BBC erred (in over-generalising).

          • Richard Sanderson

            Ophelia Benson and Rebecca Watson are going it right now. They are conflating EVERYBODY on the bot list as abusers. Ophelia is even moaning that Newsnight had to clarify their mistake.

          • Ophelia Benson

            No I’m not. I’m not making any assertions about the people on the bot list at all, not least because I haven’t seen it. I don’t use the block bot. I’ve never ‘point[ed] to this list and sa[id] “look, this is a list of abusers – these are the bad guys”.’ Never.

            Nor did I “moan” that “Newsnight had to clarify their mistake” – I linked to the video of the discussion Gavin Esler and Paul Mason had, and said they’d had it.

          • Richard Sanderson

            Ophelia – your snarky and careless response at your site reveals that you just don’t make any distinction between those abusive trolls, and the people who Oolon (and his friends) doesn’t like. If you actually search for Oolon’s reasons for adding people to the list (or even ask him), you’ll find he added someone simply for being “boring”.

            As we found out from the NN report, many not completely familiar with the spambot DO MAKE THE ASSOCIATION (even supposedly trained journalists – although he was perhaps misled by your close friend) of everybody on the list being an abusive troll, or worse. Many are on that list for simply disagreeing.

      • I have gelato

        I really don’t care if anyone from A+ blocks me on twitter, my only concern is that I and others have been added to the He-man woman hating rank. And no, I don’t think that’s trivial.

  • Ted Dahlberg

    Since I occasionally comment at the Slymepit (mostly silly stuff of no consequence), I felt it was best to check. And I am indeed listed as a level 3 undesirable. I find that faintly disturbing, especially since I have practically never interacted with anyone on Twitter that I don’t already know, let alone anyone who would consider blocking me. I invite anyone to have a look at my tweets and try to find anything offensive (it will take all of two minutes to read all of them, I’m that prolific). I’m @TedDahlberg

    • Level 3 is an absurd hodgepodge of people that the 5 (I think it is) people who can add to the bot find “annoying”. So yes, I’m not surprised that a lot of people are surprised to be there. Being on that list (at that level) wouldn’t matter much, I reckon, without the conflation between the lists, of everyone being trolls and abusers – like happened in the BBC interview.

      • ool0n

        How was there conflation? Did you see Zvans post on it, clearly three levels shown on screen. Clearly about level1 in the program. Anyone going to the site would see there are three levels and realise the nasties are in level1, if they can read. I fail to see it, apart from in the minds of those blocked. No real person would see anyone on there and think they are one of the ppl threatening rape.

        • I saw Zvan’s post, yes, but think you’re over-estimating the degree to which people pay attention. The clip allow you to see 3 levels, briefly, but doesn’t contextualise or explain them, and only shows them once. Anyone who knows the ins and outs of this story would not be confused, no, but the idle passer-by, happening to catch that insert, could easily have been. As I say above, the fact that they might be so confused has little impact, in my view – but denying the possibility that it could have been misleading is far too charitable to the average viewer.

          • ool0n

            Yes and the idle passer-by isn’t going to have a clue who is on the bot list! How would they know unless they look on the website and see more than one level and the descriptions? Immediately they know that the BBC missed out the story on L2/3.

            So I don’t see the mechanism for people seeing person X is on the list and managing to conflate the levels.

          • I’d say quite easily, through hearing mention of someone being “an abuser” on a different website entirely, making the connection or being prompted to make the connection) to the BBC report, and not bothering to investigate it further.

            The picking at these details in this exchange could create the impression that this is significant, so again, I regard this as a pedantic point, likely to affect a vanishingly small number of people, rather than something to be aggrieved about. But any impression that those who manage the Bot aren’t aware of potential repercussions simply feeds into paranoia about its potential to generate/support incomplete or misleading impressions.

  • Tim Farley

    Thank you for the kind words about my post and for linking to it. Much appreciated.

  • Tim Farley

    FYI I have addressed some misunderstandings about my blog post on Virtual Skeptics this week. The video, cued directly the relevant section, can be seen at the bottom of the original blog post here: http://bit.ly/19gLw6q