"I can't stop halfway"

Here's the objection from experience. You've tried. You've opened a bag meaning to have half and put it away, and you didn't, because halfway is a place with no signal and no partition, and once you're in the bag you're in the bag.

You're right, and the method agrees with you. Don't stop halfway. Nobody stops halfway. Halfway through a bag is a place with no markers, in the dark, with a screen on, and expecting to stop there is expecting willpower to do the thing chapter 8 said it can't.

Start with half instead. That's the whole difference. The decision about how much to eat is made once, at the cupboard, in the light, before the screen is on and before the first one has been tasted, and it's made by pouring. After that there's no stopping to do, because the bowl stops. The bowl is a partition placed at the start, where you're strong, instead of a resolution placed in the middle, where you're a hand.

Cheema and Soman's partitions worked this way. The chocolates were divided before the eating began, and the division did the controlling. The participants didn't have to decide to stop at chocolate six. They reached the end of a partition and had a moment, and the moment was where the decision got made, and it was made by a person who'd been given somewhere to make it. The researchers also found that partitions wear off with habituation, and work best when they change. That's worth holding onto for chapter 21. A bowl every night becomes invisible. A small bag one night and a bowl the next, or a different bowl, keeps the partition doing its job.

So the belief is right about halfway and wrong about the conclusion. You can't stop halfway through a bag. You don't need to. You can start with a bowl, and the bowl is the amount, and the stopping is already done by the time you sit down. What's left is the eating, and the eating is the good part.