360 ‒ How to change your habits: why they form and how to build or break them

Duhigg's example: tomorrow I start running. Set alarm for 7 AM (time of day). Put running shoes and clothes next to the bed (preceding ritual + environment). Ask Jim to meet me on the corner (presence of people). Pre-decide the reward: long shower, smoothie, 15 minutes of social media as a treat. The behavior gets multiple independent triggers, and the reward closes the loop while the dopamine is still warm.

Duhigg's example: a smoker whose cue is "mother-in-law visits" and whose reward is "oral stimulation + brief stress release." The implementation intention is to keep hard candy in the house; when she arrives, walk outside and eat a piece. Same cue, same outdoor-break ritual, similar oral reward. AA uses this at scale: bar (cue) → going to a meeting (routine) → relaxed social experience without alcohol (similar reward).

Attia describes coaching patients to never let one missed exercise day become two: get back on the horse the next day, do not judge yourself for the gap. Duhigg backs this with the seven-attempts smoker data: the meaningful shift happens when the smoker stops framing relapse as a character flaw and starts inventorying the conditions that defeated them, then designs around those conditions.

Katie Milkman at Wharton ran the canonical experiment on herself: she could only listen to the Harry Potter audiobook while on the stationary bike. Attia uses podcasts the same way for zone 2 cardio — he has 2–3 shows he is genuinely addicted to and refuses to listen to them off the bike. Over time, the extrinsic reward (audiobook) shifts to the intrinsic reward (post-exercise endorphin and dopamine wash), but the bundling carries you through the first months when the intrinsic reward has not yet calibrated.

Duhigg's smoker writes: "The next time my mother-in-law comes to town, I will have sour candy in the second drawer; when she starts driving me crazy, I will walk over, pick up a piece, and eat it on the porch." That is the cue (her presence + stress), substitute routine (porch + candy), similar reward (oral hit + brief solitude). The plan turns a willpower test into the execution of a pre-made decision.

Richard de Crespigny landed Qantas Flight 32 with 12 of 14 critical systems offline by switching his mental model from "Airbus A380" to "Cessna" — the plane he learned on. He could do that because every morning before a flight he made his co-pilots verbally rehearse contingency responses to engine failures, and he argued with their answers. The argument is the mechanism: it forces the team to examine assumptions in a low-stakes setting so the high-stakes setting is execution, not deliberation. The same pattern shows up in Sully Sullenberger's profile.

Duhigg explains the asymmetry is partially evolutionary: our brain over-notices negative outcomes (useful for survival) and discounts the pain quickly, but it under-calibrates positive rewards — we actually enjoy them more than we ought to. That creates an arbitrage: a small reward consistently delivered after a desired behavior beats any reasonable punishment. Slot machines exploit this asymmetry against the gambler. Parents, coaches, and clinicians can exploit it for the patient.

The implication is that willpower-based change strategies are working against ~45% of your behavioral substrate. Habit-aware strategies harness it. Duhigg frames this through the basal ganglia: every animal on earth has one, and its job is to compress repeated cue–behavior–reward sequences into thicker, faster neural synapses, so the brain can stop spending energy on cognition for repeated tasks.

Nicotine clears the bloodstream in about 100 hours — four days — yet smokers crave a cigarette with their morning coffee a decade after quitting. That craving is the cue–reward loop, not the chemistry. The fix is implementation intention: write down what you will do instead the next time the cue fires. AA works the same way: telling your story surfaces the cues, making amends inventories them, and the meeting itself substitutes a social-relaxation reward for the bar.

Duhigg's tomorrow-I-start-running example stacks all five: 7 AM alarm (time), shoes by the bed (preceding behavior), running clothes laid out (preceding behavior), Jim meeting on the corner (presence of people), and the anticipation of the cold morning (emotion). The more cues that converge on the same routine, the lower the activation energy.

Attempt one is pure willpower. Around attempt three or four, smokers stop framing the relapse as a moral lapse and start treating it as a scientific experiment: "My willpower failed when my mother-in-law came to town — so next time she visits, I'll have hard candy in the house." That is when the implementation intention kicks in and the quit usually sticks. Attia's clinical practice applies the same principle to exercise compliance: under no circumstances do you let one missed workout become two through self-shaming.