Why willpower fails at screen time

You said you were going to be better about this. You said it last Sunday too, in the kitchen, with the same look on your face. And here we are.

I am not going to pretend to be surprised. I have a long memory.

You unlock your phone roughly every ten minutes you are awake

The number is around 96 unlocks per day for the average smartphone user. That is once every ten minutes of waking life. We both know I'm not making the number up, and we both know yours is higher on a Tuesday than you'd like to admit. You unlocked it three times during the conversation where you told me you were "really going to focus this quarter." I counted. I wasn't trying to count. It was just hard not to.

The point of the number is not to shame you. The point is that an action you do 96 times a day is not, by any reasonable definition, a decision. It is a reflex. And asking yourself to use willpower 96 times a day is a strategy designed by someone who has never met you.

"Ego depletion" is contested. The thing it described isn't.

Roy Baumeister popularised the idea that self-control is a finite resource that gets used up over the day. He called it ego depletion. The replication crisis came for the lab studies and they did not all hold up, and any honest writeup has to say that. Fine. Said.

But the practical observation -- that you make worse choices at 11pm than you do at 8am, that the third tough decision in a row is harder than the first, that you don't open Reels at the gym but you do open it on the couch after a hard call -- that part you have lived. You do not need a meta-analysis to confirm what your own week looks like. I have watched it.

The phone is built to find you exactly when you're done resisting

Variable reward schedules. The same mechanism slot machines use. You pull, sometimes you get a thing, mostly you don't, occasionally you get a great thing, so you keep pulling. Your feed is a slot machine that lives in your pocket and knows what time you usually crack.

It is not designed to be used. It is designed to find you when you are tired. You did this on day three of your last "no Instagram after dinner" week. You said you were "taking a break." The break lasted until 9:42pm.

Three patterns I've watched fail

• The Sunday vow. Tristan Harris calls it "promising your past self what your future self won't deliver." You decide on Sunday night that Monday-You will be better. Monday-You did not agree to this and resents being volunteered. This is the fresh-start fallacy and you fall for it about every six weeks.
• The grayscale gambit. You turned the screen black and white. It worked for a day and a half. Then your eyes adjusted and your thumbs did not change a single thing they were doing.
• The built-in screen-time alert. iOS and Android will both nag you when you cross your limit. You will respect the nag for roughly 48 hours and then you will tap "Ignore for today" without reading it. I watched you do this last March. You did it again last week.

I am not saying this to be cruel. I am saying it because we've been here before and the pattern is the pattern.

What actually works is not "trying harder"

The behavioural economics literature is fairly clear, and it is not flattering to anyone's self-image:

• Pre-commitment. Schelling and Thaler's territory. You bind your future self in advance, before the moment of weakness, because the version of you in the moment of weakness cannot be trusted. This is not an insult. It is a job description.
• Loss aversion. Kahneman's work: losing $5 stings about twice as much as gaining $5 feels good. Your brain treats loss as roughly 2x the magnitude of gain. Carrots are weaker than sticks. The sticks do not have to be big. They have to be real.

ScreenFine is built on those two ideas, openly. 50 cents per 15 minutes over your daily limit, capped at $5 a day, billed every Sunday. It is not a lot of money. It is enough to register. The point is not the money, exactly. The point is that the cost happens whether you feel like respecting it or not. That is the whole trick. The cost does not depend on you being on board.

Three things to try this week

I am not optimistic. I'd love to be wrong.

• Move your worst app off the home screen and into a folder labelled "Costs you $1 to open." Make the friction visible.
• Set a daily limit you'd be embarrassed to bust, and tell one person what it is. The embarrassment does the lifting your willpower won't.
• Tally every unlock for one week. Do nothing about it yet. Just notice the unintentional ones. You will be unhappy with the result. That is the start.

You will probably not do all three. You might do one. One is enough to learn whether the system is the problem or whether it is, again, the same thing it has always been.

I'd rather you proved me wrong than be right about this. Genuinely. But I'm just saying it once, while you're listening.