Almost nobody fails a diet because the maths was wrong. They fail because the plan assumed a version of them that does not exist: one who sleeps eight hours, never eats out, and finds hunger interesting rather than maddening.
A deficit is a stress you are choosing to apply for months. The right size is the largest one you can sustain without your training, sleep or mood collapsing — which is usually smaller than the internet suggests.
Start with rate of change, not a percentage
The most robust way to pick a deficit is to decide how fast you want to lose, then work backwards.
| Weekly loss | Share of body weight | Who it suits |
|---|---|---|
| 0.25% | Very slow | Already lean, or running a long recomposition |
| 0.5% | Moderate | Most people, most of the time |
| 0.75% | Faster | Higher body fat, short timeline, experienced |
| 1%+ | Aggressive | High body fat only, and not for long |
For an 80 kg person, 0.5% is 0.4 kg per week. Fat holds roughly 7,700 kcal per kilogram, so a 0.4 kg weekly loss implies a daily deficit near 440 kcal. That is a 20% cut from a 2,200 kcal maintenance — meaningful, and survivable.
A 1,000 kcal deficit looks heroic on paper and collapses in week three, usually taking muscle and adherence with it.
Protect the two things that make the result look good
Protein. Aim for 1.6–2.2 g per kilogram of body weight per day while cutting. It preserves lean mass, and it is the most satiating macronutrient per calorie. An 80 kg person lands at 130–175 g.
Heavy loading. Two to four sessions a week with progressive resistance tells your body the muscle is still needed. Cardio is for energy expenditure and health; it is not what keeps muscle in a deficit.
Everything else — meal timing, fasting windows, food choices — is a compliance tool. Pick whatever makes the plan boring enough to keep.
Read the trend, not the scale
Weight fluctuates 1–2 kg within a day from sodium, carbohydrate intake, glycogen and digestion. A single reading is noise.
- Weigh daily, at the same point in your routine
- Compare weekly averages, not individual days
- Judge a change only after three weeks of consistent data
If the weekly average has not moved in 21 days and your logging is honest, the deficit is no longer a deficit. Cut 150–200 kcal or add 2,000 steps, then wait another two weeks before touching it again. Constant tinkering destroys the signal you are trying to read.
Plateaus are usually one of four things
- Logging drift — portions creep, oils and drinks get missed. Re-measure for a week.
- Activity decay — you move less without noticing. Steps fall first.
- Real adaptation — output genuinely drops a few percent as you get lighter and leaner.
- Water masking fat loss — a new training block or a high-salt week can hide three weeks of progress, then release it overnight.
Only the third is a reason to change the plan. The first two are reasons to tighten measurement.
When to stop dieting
Dieting is not a state you maintain indefinitely. Take a break when:
- Weekly loss has stalled for three weeks despite honest logging and a real deficit
- Sleep quality, mood or training performance are clearly degrading
- You have been in a deficit for 12–16 weeks without a pause
- Body fat is low enough that further loss costs more muscle than fat
A diet break at maintenance for one to two weeks restores adherence, refills glycogen and lets you check whether the plan still fits. It is not failure; it is maintenance practice, which is the skill you actually need afterwards.
The unglamorous version
Eat 15%–20% below maintenance, get 1.6 g/kg of protein, lift twice a week, walk daily, weigh yourself every morning, and change one variable every three weeks. That is the whole method. It works because it survives contact with a normal life.