, ,

I rather like this piece by Alex Gabriel.

I don’t know if I lived with a rapist, or someone who’d have liked to be. None of these incidents proves anything, but what if that was the idea? Was F, I wonder now, scoping me out the way queer kids scope out their mum and dad, as I’d scoped him out with mention of feminists? Did he laugh about rape because it amused him, or because what might be a joke is always plausibly deniable, like a sexual advance veiled as an invitation for coffee? One’s instinct is to award the benefit of the doubt, but maybe that’s the point.

It echoes something that’s been rattling in my own head since I first heard of the Iterated Prisoner’s Dilemma.

The last two-thirds of that phrase are a game that’s puzzled game theorists for some time. Long story short, two people are given the option to help or exploit one another. If they both chose to help, both benefit; if they both chose to exploit, both suffer. But if one helps while the other exploits, the helper suffers greatly while the other party benefits greatly. The puzzle was that, logically, the best action is to always exploit the other person, yet when this problem is handed to people they almost always chose to help. This resulted in much nervous murmuring among game theorists.

Enter Robert Axelrod, and the “Iterated” part. Rather than consider a one-off choice, he wondered what would happen if people were forced to make this choice continuously with their neighbours. What emerged as the best strategy to take, Tit for Tat, was also one of the simplest: do unto your partner what they did to you last time. What do you do when there was no last time? Simple, your first move is to help. The puzzle was neatly solved.

But notice that Axelrod’s tournaments carry certain assumptions with them. What if you had some control over your opponents, and could ensure you only played them a fixed number of times? Now it’s in your best interest to stab them in the back when you see them last. We’ll call this strategy “the Grifter:” move from place to place, taking advantage of people’s initial offers of generosity but then bailing before they can return the non-favor. There’s an obvious counter-strategy, “Fear Outsiders,” which shouldn’t need elaboration.

There’s no law stating you’ve gotta exploit people when you first meet them, though. And unlike most of Axelrod’s simulations, mistakes and misunderstandings happen. Rather than have a sole help/exploit mental checkbox, people tend to store a history of past interactions and overlook what they consider one-time mistakes or accidents. So a more sophisticated strategy is “Trust Farming:” build up a history of helping, then either exploit that for a big betrayal or make repeated small “mistakes” and ask for the benefit of the doubt. This, however, could build up “A Reputation” which is readily shared across community lines.

A subtle variant comes from observing that not all interactions are the same. It’s possible for someone to be helpful in situations A to Y, but exploitative in Z. If you’re not a part of Z or effected by it, all you see is benefit from interacting with this person. So why rock the boat and call them out? By carefully picking their Z’s, and exploiting our non-omniscience, someone can form a “Chronically Problematic” strategy. A decent response is “Empathy,” that we should care about and listen to the problems of other people even when they don’t effect us.

Real-life examples are easy to come by, especially in an environment that makes it trivial to change your name.

A strong defense is precisely what Gabriel gently hints at: don’t reach for steel every time you’re confronted with straw. Deal with the words as spoken, as it better challenges the thoughts which may be behind them.

What if the very appeal for charity is cause to withhold it? Where doubt exists, people who deserve generosity rarely rely on it, nailing their colours to the mast instead and making what they think clear. Those who evade, equivocate and require others to assume the best sometimes sow doubt for their own benefit: F asked me to believe his daydreams of torturing women were a joke, yet never offered any reason to, sheltering in uncertainty. Without a means of dispelling the fog, I wonder what charity might have forced me to ignore.

[HJH 2015-08-31: Spotted some grammer in need of a light scrub.]