Why You Make Decisions You Regret

Why do choices you later regret feel sensible in the moment? This practical Stoic talk explains how anger, exhaustion, loneliness and other hot states can change what seems valuable. It replaces the vague demand for more willpower with a four-step handover note: identify the state, write one precise if-then rule and change the environment before judgement shifts. Use it for late-night messages, scrolling, drinking or any repeated decision that looks different the next morning.

Transcript

The standard story of regret is that there's a gap between knowing and doing. You know you shouldn't send that message, You know you shouldn't take the drink, Open the tab at 1am, Say the thing to your partner you've already said twice this month, And you know it, And yet you do it anyway. So the diagnosis is, I have no discipline, I have a weak will, And the prescription is just to become more disciplined. But that story is wrong. Not slightly wrong, But wrong in a way that sort of guarantees the fix. Won't work and never will work. Ovid gave us the line everyone quotes. Medea, On the edge of the worst decision in Greek literature, Says, Titus reads it differently. He came back to Medea repeatedly in the discourses. And his verdict was cold. She did exactly what she thought was best. In that moment, Revenge looked to her like the most valuable thing available. Her error wasn't in her will, It was in her pricing. This is one of the least fashionable stoic claims, But it's also one of the most useful for understanding our decisions and human behavior. Nobody assents or endorses to an impression they judge to be false. Nobody chooses the option that looks worse to them at the moment of choosing, Which means every decision you've regretted made complete sense to the person making it. And that person was, Of course, You. And you weren't lying to yourself, You were running a different valuation at the time. So the question isn't why you failed to act on what you knew, It's why, For roughly 90 seconds, You knew something else. George Loewenstein calls this the hot-cold empathy gap. In a cool state, You can accurately describe what you want and how much it's worth. In a hot state, Driven by anger or hunger or exhaustion, Arousal or loneliness, Boredom, The numbers change. Not your access to the numbers, But the actual numbers. George Ainslie's work on discounting shows the same thing, But with more precision. We don't devalue future rewards on a smooth curve. We devalue them hyperbolically, Means the curves cross. At a distance, The larger, Later reward is genuinely preferred. But up close, The smaller, Sooner one wins. Both preferences are real, And both are yours. And the reversal isn't a lapse in the second self's integrity, It's just basic arithmetic. Notice what this does to this story of discipline. There's no calm, Rational you being overpowered by an impulsive you. There are two states of the same nervous system that price the world differently, And only one of them is in the room when the decision gets made. Below about 30 meters, Breathing compressed air, Divers develop nitrogen narcosis. The old rule of thumb calls it Martini's Law. Every 10 meters past the threshold is worth roughly one drink. Judgment degrades, Reaction time slows, And fine motor control goes. But that isn't what makes it dangerous, Actually. What makes it dangerous is euphoria. Often report feeling excellent, Calm, Capable. Some become convinced they should go deeper to see what's down there. A few have been found having offered their regulator to a passing fish. The faculty that would detect the impairment is the impaired faculty. That's the part that transfers. When you're in a state that produces the regretted decision, You don't experience yourself as compromised. You experience yourself as clear-headed. In fact, Unusually clear-headed. The 2 a. M. Scroll feels like curiosity. The message you'll regret feels like honesty. The seventh drink feels like being present with your friends. From the inside, The impayment presents itself as a form of insight. Divers didn't solve this with resolve or more discipline. They solved it with procedure. Planned maximum depths agreed on the surface. Gauge checks at fixed intervals whether or not you feel you need one. A buddy whose job it is to notice what you can't. Ascend a few meters and it clears. Nobody expects a narced diver to reason their way out of narcosis. Because the reasoning is the thing that's bent. And here's something uncomfortable. Intelligence doesn't protect you. And there's a decent case that it might even hurt. David Perkins found that people with higher measured intelligence generated more arguments on informal reasoning problems, But not more arguments against their own position. The extra capacity went entirely into their own side. Keith Stanovich has spent decades on this, Coining disrationalia. For the specific condition of being smart and still reasoning badly, And showing that my side bias is largely uncorrelated with cognitive ability. Read that alongside Nisbet and Wilson's classic work on introspection where people confidently reported reasons for behavior that demonstrably weren't the causes, And a shape emerges. Your reasoning isn't the judge, It's the defense, Barrister, And a good education just means you've hired a better one. Which is why the smartest people you know have the most sophisticated accounts of their worst habits. The account isn't a failure of intelligence, It's actually an achievement of it. It. The obvious response is to train willpower like a muscle. That model has had a bad decade. The ego-depletion literature, Once close to settled, Largely collapsed under the multi-lab replication. Hager and colleagues ran the effect across two dozen labs in 2016 and found something close to nothing. Whatever is happening when you fail at 11pm, It probably isn't a depleted reservoir of self-control that push-ups and cold showers will deepen. The Stoics were better engineers than this. Though the modern retelling flattens them into people who just tried harder. But if you read the actual practices, And they're almost all preemptive, Epictetus tells you to speak to an impression before you ascend to it, To make it wait while you examine it, Which is not a strategy for the moment of temptation so much as a habit you install beforehand. Premeditatio malorum, Premeditation of future evils, Is not gloominess, It's rehearsing the hot state while you're cool, Which is the only condition under which you can plan for it The move is to stop trying to win the argument at depth and start deciding on the surface. Peter Golowitz's work on implementation intentions is the cleanest modern version. Vague goals like, I'll be more careful with my phone in the evening, Do very little. If then plans that specify the situation and the response produce reliable medium-sized effects across hundreds of studies. The plan works because it isn't asking your compromised self to deliberate. It's asking it to follow a decision that was already made by someone with better information at hand. So I recommend you do this once properly for one specific decision. There are four steps. 1. Name the last decision you regretted. 1. Decision. Purely factual. No commentary on your character. 2. Reconstruct the state, Not the choice. Bring to mind where you were, What time it was, What you were running on, Tired, Hungry, Angry, Bored, Alone, Maybe with a screen, Three drinks in, Just off a difficult phone call. This is depth reading. It matters more than the decision itself. Write one if-then sentence. If it's after 11 and I'm alone with my phone, Then it goes on the kitchen counter and I read something on paper. Specific trigger, Specific action, Not an intention, A rule. Therefore, Make one change that doesn't require you to be at your best. Move the object. Delete the app. The test is whether it still works on your worst night. Because your worst night is the only night it's for. So we have name the last decision you regretted, Reconstruct the state, Not the choice, Write one if-then sentence, Make one change that doesn't require you to be at your best. You are not writing a promise, You're writing a note to someone who will be less capable than you are now, And who will not believe that they are.

Reviews

Fascinating. I really enjoyed this talk. I’ll give the practical exercise a go. Thanks.