Keeping Safe on Twitter, Blackstone’s Ratio, and Ways to Support Victims

A lot’s happened since my “Innocent until proven guilty in a court” isn’t the end of the story went viral on Twitter. New and credible allegations of abuse in the games industry have emerged, and the pushback from those who believe all the allegations are lies has been typically fierce.

And yes, there has been tragedy – but this isn’t the moment to lay low. The attacks on victim of abuse may get worse before they get better, so I hope to share some useful lessons from my recent experience of receiving hundreds of abusive tweets (which isn’t at all comparable to the far worse attacks others have endured), including:

  • Ways to keeps yourself safe on Twitter
  • Responses to common “gotcha” arguments
  • Ways to support victims

First, keep yourself safe

If you find yourself the subject of hundreds of abusive tweets but you want to keep using Twitter, your first step is to activate advanced mute filters:

In particular, checking Who don’t follow you will hide 99% of trolls. I didn’t check that one because I don’t follow many nice people whom I still want to hear from (yes, I’m a bad person) but you should.

In extreme cases, you may want to lock your account temporarily.

Don’t engage, just block. If you get an abusive or bad faith tweet, you shouldn’t hesitate to block the senders. That’ll clean up your notifications quickly.

Blocklists don’t work. Twitter allows you to block accounts en masse by importing blocklists. Unfortunately, all of the blocklists I found weren’t being actively updated, and they’re full of accounts that you might still want to hear from, e.g. @nytimes. I get why people might want to block @nytimes but that wasn’t what I was looking for, and it made me distrust the entire list. Other blocklists suffer from even worse selection biases. So I can’t recommend using them.

The truth is, even a well-resourced volunteer blocklist would struggle to keep up with the ever-changing number of abusive Twitter accounts. The only proper solution is via Twitter itself. Unfortunately, they seem to have little interest in cracking down on abuse or even developing tools for users to manage abusive tweets themselves. There is no way to easily bulk-block accounts; to turn off replies for a tweet; or to prevent a tweet from being quoted.

Why is quote-tweeting such a problem? See this example:

This quote hides the second line of my tweet, which said “Stop saying it. Here’s why👇” making it clear there was more to my argument. It goes without saying that plenty of people don’t read past the headline, but it’s also clear that truncated quote-tweets made that problem even worse in my case.

Interesting, Mastodon deliberately doesn’t allow quote-tooting:

…it inevitably adds toxicity to people’s behaviours. You are tempted to quote when you should be replying, and so you speak at your audience instead of with the person you are talking to. It becomes performative. Even when doing it for “good” like ridiculing awful comments, you are giving awful comments more eyeballs that way.

I only had a quick blast of this, other people get it far worse than me, and there is no doubt in my mind Twitter’s lack of action is contributing towards significant mental harm.

“Better 100 rapists go free than one man be falsely accused of rape”

Some people care more about theoretical justice than actual justice. They are essentially fantasists, and you can spot them quite easily, because they often say something like this:

Better that 100 guilty men go free than 1 innocent man goes to jail. That applies to rape, murder, and any other crime.

I imagine the author of that (real) tweet muttering “checkmate” as they typed their version of Blackstone’s Ratio (the exact ratio varies between 10-100 “guilty”, with Benjamin Franklin at the top end of the scale).

Now, before we get into it, let’s be clear: invoking “innocent until proven guilty in a court” is applying a legal standard to a social situation, like shouting “first amendment!” doesn’t and shouldn’t protect you from people getting angry about what you’re saying. It is quite normal for people to earn reputations as bullies or liars amongst groups without proving anything in a court of law. It is only when a severe sanction is to be applied, such as a fine or imprisonment, that this legal standard is relevant.

Putting that aside, the problem with Blackstone is that not everyone agrees with it. Just because it’s a long-standing principle doesn’t mean it’s a good principle for today. As Daniel Epps writes:

English jurors [in the 1700s] literally feared damnation if they erroneously sent an innocent defendant to death. Moreover, given that English judges and juries faced binary punishment options, the Blackstone principle was consistent with simple notions of deterrence. Where only one level of punishment for crime is available, and where that punishment is quite harsh, following something like the Blackstone ratio would make good sense, especially when relatively minor crimes are at issue.

Thankfully, we live in a very different world now. So while we can have a calm and productive conversation about the validity of Blackstone in countries with English-derived legal systems, it doesn’t come close to being a “checkmate” rhetorical move when it comes to talking about allegations of abuse. Even if you believe in Blackstone, even if you don’t care much about convicting guilty people, it’s clear that we absolutely don’t care about preventing certain types of innocent people from being convicted.

Anyway, if you’re interested in learning more about the problems with Blackstone, check out The Consequences of Error in Criminal Justice: Why convicting one innocent person may be better than letting ten guilty people go free (PDF) in the Harvard Law Review by Daniel Epps, former law clerk to Supreme Court Justice Anthony Kennedy and associate professor of law at Washington University in St. Louis. Epps isn’t the first scholar to doubt the the validity of Blackstone, but he expresses its problems well. It’s 87 pages long, but it’s surprisingly readable!

tl,dr: Epps believes:

  • Blackstone doesn’t help the innocent as much as you might imagine, because if 10 or 100 guilty men are routinely going free, being declared “not guilty” in a court isn’t going to convince many people of your true innocence.
  • By letting 100 guilty go free, prosecutors and politicians can feel pressured to harshly punish the few who are actually convicted. This directly leads to overuse of plea bargains and false convictions.
  • Blackstone can lead to more crime and less confidence in the justice system as a whole, due to high-profile false negatives (i.e. celebrities who are wrongly declared not guilty).

So: I reject Blackstone’s Ratio, and you should feel comfortable doing so as well.

The truth is, I don’t think those quoting Blackstone at me have considered its implications for a single second – and if they have, they’re clearly more invested in abstract principles of justice than the lived experience of countless sexual abuse victims. This allows them to ignore the utter lack of justice I mentioned in my first post.

Some people think women aren’t people

Disturbing beliefs from those attacking me:

  • A false rape accusation is worse than being raped.
  • One false accusation of rape is one too many, but one rape is fine. Also, one rapist going free is fine. In fact, a hundred rapists going free is fine, even if presumably they go on to rape other people.
  • By implication, men are naturally truthful, whereas women should be distrusted.
  • All actual cases of rape always go to trial.
  • All actual rapists get convicted (and even if they don’t, better 100 guilty… etc.)
  • Most rape accusations are false.
  • …but even if (say) 5% are false, that’s enough to dismiss all rape accusations as false (because better 100 guilty… etc)
  • The reason many women wait years before talking about rape or abuse means they’re lying, because there’s no conceivable reason why anyone wouldn’t report instantly – it’s not true that police don’t believe women, rape kits don’t get tested, and rapists get light sentences if convicted, or that women wait because they’ve been threatened or are waiting for corroboration from other victims.
  • All evidence and statistics contradicting these beliefs are wrong, including those published by various governments.

Not all accusations are equal

Many attackers believe all accusations of abuse are essentially equal, which explains why so anonymous Twitter users made various accusations against me, also no doubt muttering “checkmate”. But the recent allegations made against videogame developers are notable for their detail:

  • Victims have given their real names, putting themselves at genuine risk.
  • The victims have documented connections to the abuser, often as an employee or contractor.
  • They provide details, dates, and locations.
  • They are often corroborated by people with documented connections to both parties.
  • They match allegations made by other victims.

Is this incontrovertible proof? No. But to me it constitutes a credible accusation, whereas simply tweeting “Adrian attacked me yesterday in London” is … not. I can’t believe I have to say this, but there you go.

We all need to do more – and yes, that means you

I am fortunate enough that I am not in fear of losing my livelihood at the hands of a Twitter mob. No-one can complain to my boss or demand for me to be fired. And yet, even from my unimaginably secure position compared to most, even I thought twice about speaking out in support of victims whom I know personally. Life absolutely would have been easier if I just kept my mouth shut.

But it’s the responsibility of people in positions of power and security to speak out, especially when it makes you scared. If you aren’t scared, then you aren’t really helping – you’re just making yourself feel better. But it’s OK to be scared. Because if you weren’t scared, you’re not being brave, you’re just an idiot.

So you need to speak up. Don’t just say “I believe her.” Say “I will never work with ABC again.” Say “I will only work for companies that have strictly enforced rules against harassment and abuse.” Say, “I will donate to RAINN and The Survivor’s Trust.”

And if you see someone being attacked on Twitter, help them, even with just a few kind words. You know who gave me the most messages of support amid the flood of abusive tweets? Women – not men. Talk about emotional labour…

“Innocent until proven guilty in a court” isn’t the end of the story

“Innocent until proven guilty in a court” is, at best, childishly naive and, at worst, dangerously obtuse when it comes to accusations of abuse (which I fully believe) against game devs like @alexiskennedy

Stop saying it.

Just 1.5% of rape cases lead to a charge or summons in the UK. The proportion will be even lower for other kinds of sexual abuse and harassment. And trials take years to complete.

So victims barely ever get to have their day in court. It’s cripplingly expensive and disruptive to your life and work. Unfortunately, this is a direct effect of the UK justice system being dismantled, largely by the Tories, as detailed in Stories of the Law and How It’s Broken by The Secret Barrister.

The real world isn’t like The Wire. People don’t wear bodycams and tape recorders 24/7, and they shouldn’t have to. So there’s often precious little evidence. Why even go to the police when the odds are so comprehensively stacked against you?

You think the goal of these accusations is to ruin someone’s life or put them behind bars? I can’t speak for the victims, but the fact that @Babylonian is still YouTubing suggests that’s not a realistic expectation. These people’s lives go on.

But the one assured outcome of public accusations is that other women can protect themselves in future. The fact that so many were surprised by the news yesterday means whisper networks can’t do it all.

I will help fund ongoing efforts to support and protect victims of sexual abuse and harassment in the games industry. If your group needs money, hit me up.

Apple Rules*, Flat vs. Threaded Discussions

Issue 11 of my newsletter – subscribe here

I spotted this ad on the tube recently, and it made me think about Apple’s unappreciated dominance over the lives of hundreds of millions of people:

600 free prints a year – what a deal! Clearly there must be a catch, and sure enough, there is a tiny asterisk next to “App!” I wasn’t able to see it from my side of the carriage, but there’s an even-tinier line of text right at the bottom of the ad that doubtless discloses just how much you’ll have to pay to get those 600 “free” prints.

This kind of chicanery is prohibited by Apple on its App Store. There are very strict rules about disclosure of payments, and in particular, of the auto-renewing subscriptions that so many developers (including my company) are so fond of, since it’s kind of the only way we can people to actually, you know, pay for a service these days.

How strict? Well, there is a very specific set of wording you must use to describe how subscriptions auto-renew. The text labels on the subscription buttons must be over a certain size. Prices must be the biggest element of text, so you can’t just say “FREE!*” as the tube ad does:

In the purchase flow, the amount that will be billed must be the most prominent pricing element in the layout. For example, an annual subscription should clearly display the total amount that will be billed upon purchase. While you may also present a breakdown price that the annual amount is equivalent to or a savings when compared to weekly or monthly subscriptions, these additional elements should be displayed in a subordinate position and size to the annual price. This ensures that users are not misled.

It’s all very consumer-friendly, so I honestly don’t have a big problem with the intent of Apple’s rules. But just consider what’s happening here – Apple dictates the precise way in which you design key parts of your app. That’s unusual, to say the least. It’d be like the London Underground dictating that you couldn’t design adverts with asterisks and tiny disclaimer text. Maybe they should? But maybe not.

If you don’t agree with Apple’s rules, well, you don’t need to be on their platform, right? It’s not like you have to make iOS apps, after all. Why, you can just… completely ignore the majority of the smartphone app revenue in the US and UK!

So that’s a big problem. Apple may not command the lion’s share of smartphone users as Google does, but in many markets, iOS users spend far more money on apps than Android users.

A bigger problem is that Apple breaks its own rules by not disclosing pricing and subscription terms; by using push notifications for marketing (third parties apps are prohibited from doing so); by immediately cutting off trial subscriptions rather than letting them run for their whole term; and so on. It matters because Apple is now competing directly against many of the companies whose apps it hosts and rules over, from Apple Music to Apple Arcade to Apple TV+ and Apple News.

Apple would claim it tries to be reasonable when reviewing third party apps, and I don’t doubt the motives of their review team. However, the mere chance that your app update could be rejected, or worse yet, removed, whether for good reasons or just out of a misunderstanding? It’s literally terrifying for people whose livelihoods depend on Apple’s whims – like me.

All of this has introduced a chilling effect on third party app developers. I’ve seen many otherwise outspoken developers genuinely scared of voicing even mild criticism of Apple on private forums and chat rooms, let alone on Twitter or podcasts. They really think they could be blacklisted for criticising Apple, and while I am 100% certain that isn’t the case – because I and others have been far more critical in public and have never experienced any blowback – I understand their fear.

Now, as far as benevolent dictators go, you could do much worse than Apple. I’ve always had pleasant dealings with them in person. But no company should have that kind of complete and terrible power over tens of thousands of companies and hundreds of millions of users. We don’t allow it in the “real” world and so we shouldn’t allow it in the digital world.

I’ve been thinking a lot about the merits of flat versus threaded discussions online. Flat discussions are like those on Metafilter and other old-school forums where each reply follows directly from the one above it, while threaded discussions are like what you see on Reddit where replies can be in response to other specific replies, creating multiple “threads” branches off from one another.

For a long time, I believed flat discussions were generally superior to threaded ones. They helped promote coherent debate rather than encouraging groups of users to spin off on wild tangents, never to be seen again. In other words, Metafilter good, Reddit bad.

But I’ll confess – I’m spending a lot more time on Reddit than Metafilter these days (although I’m making a concerted effort to change!) That imbalance is partly because Reddit, having far more users than Metafilter, can cater towards more specific interests. There’s just no way I can get all the latest news about Edinburgh or Apple on Metafilter, that’s not what the site was designed for.

There’s another reason, though: I don’t think that flat discussions, as typically implemented by forums like Discourse or Metafilter, are working well for communities where there is serious disagreement about how discussions should unfold.

I’m being really unspecific here because I don’t just mean “communities with political disagreements”, I mean “communities where some people like making jokey comments and others dislike them” and also “…where some like tangents and others hate them.”

Flat discussions makes problems stick out more because everyone is forced to read them. In flat discussions, it’s possible for one or two people to completely dominate a discussion, or for irrelevant, misinformed, or extreme opinions to derail a conversation because it’s just impossible to ignore them; you’re literally posting right underneath them. In threaded forums, these sidetracks would end up downvoted and quarantined in their own threads.

Metafilter and other “flat” forums are trying to preserve a unitary community where we hope that just by talking enough, people can come to an agreement about how to run a conversation. Reddit assumes that’s impossible.

Watching

🎞️ Dr. Strangeglove. Fantastically funny and tightly-written masterpiece.

Reading

📖 Ways of Seeing by John Berger. A classic, which means you’ve probably heard most of its arguments in other books by now, but still very much worth reading. It doesn’t waste your time.

📖 The Calculating Stars by Mary Robinette Kowal. Don’t be put off by the misleading “Lady Astronaut” tagline – it’s impeccably researched but packed with smart, interesting characters. A bit like Stephen Baxter crossed with someone who understand human emotions.

Visiting

🏛️ Michael Rakowitz at the Whitechapel Gallery. Funny and political and smart, an artist for millennials.

🏛️ Manga at The British Museum. A workmanlike exhibition that’s nonetheless essential if you have any interest in Manga whatsoever.

🏛️ AI: More than Human at the Barbican. Very much for the layperson who doesn’t mind watching a bunch of Google and Deepmind marketing materials that are wholly unquestioned. There was one video, made and filmed by Deepmind, about the work they’re doing with Moorfield Eye Hospital – but nothing about the many controversies about their use of patient data.

It is an occupational hazard of curating an exhibition about AI that you quickly run out of good stuff to display. Sure enough, two thirds of the way through, the exhibition devolved into a grab bag of “future stuff” that had nothing to do with AI. Must try harder!

Actually, one more thing – there was a machine-generated poem installation at the Barbican. It wasn’t any good. I’m pretty tired of this AI fetishism and star-struck curators who pretend machine-generated stuff is worth exhibition.