Chapter 31 — Meal Planning: Why Plans Fail, and the Ones That Don't

The Hook: Sunday's plan, Thursday's bin

Kwabena Asante plans on Sunday evenings. He is 38, works in software, lives with his partner, and enjoys cooking.

⚠️ He has time, money, a decent kitchen and a freezer. He is exactly who this chapter is for.

Here is the plan he wrote on a Sunday in October, and here is what actually happened.

⚠️ Three of seven nights went as planned.

And what went in the bin on Thursday week: ⚠️ half a bunch of coriander, most of a bag of pak choi, two leeks, a red chilli, half a tub of crème fraîche, the salmon, and a bag of spinach he bought "because it's good to have."

⚠️ Roughly £14 of food, in a household that plans.

⚠️ Kwabena's diagnosis of the problem was the wrong one.

"I need to be more disciplined about sticking to it."

⚠️ He does not. The plan was the problem, and specifically it was a plan written by Sunday-evening Kwabena for a week that Wednesday-evening Kwabena was going to have to live in.

Look at the failures:

Wednesday failed because of a late meeting — ⚠️ which he knew about on Sunday and did not plan around. Thursday failed because Wednesday failed — ⚠️ the chicken pie hadn't been made, so the leftovers were still there, and the plan had no way to absorb a one-day slip. Friday failed because he was tired and the plan called for a NEW RECIPE — ⚠️ the highest-effort option, on the lowest-energy night of the week.

⚠️ Three of the four failures were predictable on Sunday. None of them was a discipline problem.

⚠️ This chapter is written for people who have some time, some money, some kitchen and some storage.

⚠️ If you don't — if the binding constraints are money, hours, a shared kitchen, no freezer, or a shop you can't get to — Chapter 32 is your chapter, and it is not a lesser version of this one. Chapter 22 §22.11 named those constraints and could not solve them, and pretending this chapter is universal would be the same failure in a friendlier register.

🏃 Fast Track: §31.2 (⚠️ the four failure modes), §31.6 (⚠️ the list), §31.8 (⚠️ components), §31.11 (⚠️ the bad night). Twenty-five minutes.

🔬 Deep Dive: §31.4 (working backwards), §31.7 (batch cooking, honestly), §31.14 (⚠️ who this isn't for).

31.1 Does planning actually help?

⚠️ Briefly, and honestly, because Chapter 30 conceded that label reading was defensive and this chapter has to earn its claim.

What the evidence supports: ⚠️ meal planning is associated with better diet quality, greater food variety, and lower prevalence of overweight and obesity — most prominently in analyses of the French NutriNet-Santé cohort.

⚠️ And it is observational, with everything Chapter 2 §2.3 says about that. People who plan meals differ from people who don't in income, education, time, household structure and cooking confidence.

⚠️ The honest position: planning is plausibly useful, the causal evidence is weak, and the mechanism is not mysterious.

Planning changes what is IN THE HOUSE, and what is in the house determines a great deal of what gets eaten — which is Chapter 22's argument arriving in a kitchen, and it is better supported than the planning literature itself.

⚠️ Which means the useful framing is: planning is a delivery mechanism for the shopping list, and the shopping list is the intervention (§31.6).

31.2 ⚠️ The four failure modes

Almost every abandoned meal plan fails in one of four ways, and none of them is discipline.

⚠️ 1. Planned for the week you wanted, not the week you have

Kwabena's Wednesday. ⚠️ The late meeting was in his calendar on Sunday.

A plan that assumes seven evenings of equal capacity will fail on the first unequal one — and every week contains unequal ones.

⚠️ 2. Too many novel recipes

Each new recipe carries hidden costs: ⚠️ unfamiliar ingredients to source, a recipe to read while cooking, unpredictable timing, and no muscle memory.

⚠️ Kwabena planned two new recipes in one week and put one of them on a Friday.

⚠️ A new recipe is a hobby. A repertoire is a system. §31.5.

⚠️ 3. No fallback, so one failure ends the plan

Thursday failed BECAUSE Wednesday failed. ⚠️ The plan was a chain, and chains break at their weakest link and stay broken.

And once two nights had gone, the plan was "not working," so it was abandoned — which is the all-or-nothing pattern Chapter 10 §10.4 identified in diets and which applies identically here.

⚠️ 4. Ingredient orphans

⚠️ The half-bunch of coriander. The remaining two-thirds of a tin. The pak choi bought for one dish.

Recipes are written for their own sake, not for a week. ⚠️ Seven unrelated recipes generate maximum orphans, because almost nothing is shared.

💡 ⚠️ Notice that all four are properties of the PLAN, not of the person.

A plan that assumes uniform capacity, front-loads novelty, has no slack, and shares no ingredients will fail for almost anybody.

⚠️ Which is good news, because plans are much easier to change than people.

31.3 Map the week before you plan it

⚠️ The step almost nobody does, and it takes four minutes.

Before choosing any food, write down what the week actually contains.

⚠️ Then classify each night by CAPACITY, not by cuisine:

⚠️ Most weeks contain one or two HIGH nights, three MEDIUM, one or two LOW, and one ZERO.

A plan with seven medium-to-high nights in it is not a plan. It is a wish with days attached.

31.4 ⚠️ Work backwards, not forwards

The structural correction, and it changes everything downstream.

⚠️ Working backwards means the constraint comes first and the food comes last — which feels backwards to anyone who enjoys cooking, and is the whole point.

⚠️ Kwabena's week, replanned:

⚠️ Same amount of cooking. Same enjoyment. Two failure modes eliminated by rearrangement alone.

31.5 A repertoire, not a menu

⚠️ The goal is not variety. The goal is ten to fifteen meals you can cook without reading anything.

⚠️ What a repertoire meal is:

• You know it without a recipe
• ⚠️ You know roughly how long it takes, accurately
• You have most of the ingredients most of the time
• ⚠️ It survives substitution — no coriander, use parsley; no chicken, use chickpeas
• You can cook it tired

⚠️ Novel recipes are how you ADD to the repertoire. They are not how you feed yourself on a Wednesday.

One new recipe a week, on a high-capacity night, cooked three or four times before it counts as repertoire. ⚠️ At that rate you add roughly twelve meals a year, which is more than enough.

⚠️ And the cultural observation worth making, because it corrects an assumption:

Most cuisines, historically and currently, run on a SMALL repertoire. ⚠️ Households across the world eat a rotating set of perhaps fifteen to twenty dishes, seasonally adjusted. The expectation of constant novelty is recent, is largely a product of recipe media, and is one of the main reasons planning feels exhausting.

⚠️ Rotation is not failure. It is how eating has always worked.

31.5b ⚠️ What a real repertoire looks like, written down

The instruction "have ten to fifteen meals you can cook without thinking" is easy to write and strangely hard to act on, ⚠️ because almost nobody has ever written theirs down. So here is one, in full, in the form that makes it usable.

⚠️ The organizing principle is the one from §31.3: sort by CAPACITY, not by cuisine. A repertoire sorted by cuisine tells you nothing on a Wednesday. A repertoire sorted by how much you have left in you answers the actual question.

Fifteen meals. ⚠️ Four of them need no thought and no fresh ingredient; seven cover the middle of the week; three are for when you actually want to cook; one is the fallback.

⚠️ Look at the distribution, because it is the point of the exercise. Most people's mental repertoire is heavily weighted toward the MEDIUM and HIGH rows — those are the meals they think of as "meals" — ⚠️ and the LOW row is the one that decides whether the week holds.

⚠️ You do not need more dinner-party food. You need four more things you can produce at 19:45 with your coat still on.

⚠️ Three properties of the list above that are doing more work than they look like they are:

1. ⚠️ It is written by capacity, so it can be used at the moment of decision. The moment of decision is tired, hungry and standing in a kitchen — and "what have I got the energy for" is the question actually being asked, not "what cuisine do I fancy."

2. ⚠️ Ingredient overlap is visible. Onions, tinned tomatoes, a pulse, eggs, rice, pasta and frozen vegetables appear across most rows. ⚠️ That is what makes a standing list of twelve to twenty items possible (§31.6), and it is the direct fix for §31.2's fourth failure mode.

3. ⚠️ It contains repeats and does not apologize. Two pasta rows, two egg rows. ⚠️ A repertoire is not a menu for a restaurant; nobody is grading it.

🔄 Retrieval practice. ⚠️ Stop and write yours. Not aspirationally — actually.

⚠️ Three columns: capacity, meal, honest time. Most people get to six or seven and stall, then remember four more, then find they were at eleven all along and had never counted.

⚠️ The stall point is informative. If you stall in the LOW row, that is the gap, and it is the cheapest one in this chapter to fill — a low-capacity meal takes one shopping decision and no skill.

⚠️ And the honest caveat about the fifteen above: they are mine, or near enough, and yours should not look like them. They are here as a FORMAT, not a prescription. ⚠️ A repertoire that doesn't match what you like eating, what you were taught to cook, or what your household will accept is a list you won't use — which is Chapter 10's adherence point applied to a spreadsheet.

31.6 ⚠️ The shopping list is the intervention

Everything above exists to produce this, and this is the part that actually changes what you eat.

⚠️ A plan without a list is a wish. The list is what converts intention into food in the house — and Chapter 22 established that what's in the house determines a great deal of what gets eaten.

⚠️ What the evidence and practice support:

• ⚠️ Shopping without a list produces more unplanned purchases, and unplanned purchases skew heavily toward Group 4 (Chapter 22)
• Shopping hungry is worse — a small, real, replicated effect
• ⚠️ Online ordering removes the aisle, which cuts impulse purchasing substantially — and it is one of the few genuinely effective structural interventions available to an individual

⚠️ How to build one that works:

1. ⚠️ The standing list. The twelve to twenty things you always buy — Chapter 30's decoded shelf. These need no thought and should be a saved list, not a weekly decision.

2. ⚠️ The variable list, generated from the week's plan.

3. ⚠️ Ordered by shop layout, not by meal. Walking the shop once beats walking it four times, and every extra lap is an exposure.

4. ⚠️ Checked against the fridge and freezer first (§31.13).

5. ⚠️ And nothing on it "just in case." Kwabena's spinach — bought "because it's good to have" — went in the bin. ⚠️ Aspirational purchases are the single most reliable waste category.

31.7 Batch cooking, honestly

⚠️ Recommended constantly, assessed rarely. Here are both halves.

⚠️ What it genuinely delivers:

• Economies of scale in preparation — chopping four onions takes barely longer than one
• Fewer cooking events per week
• ⚠️ A meal available on a zero-capacity night, which is §31.11's whole problem
• Better ingredient use — a whole bunch of coriander gets used

⚠️ What it costs, and this is the part usually omitted:

⚠️ Palatability decay is why batch cooking fails, and it is almost never mentioned.

The intuitive fix — make a big batch of one thing — maximizes the problem.

⚠️ The actual fix is §31.8, and it is the single most useful practical idea in this chapter.

31.8 ⚠️ Batch components, not complete meals

⚠️ Cook building blocks, not dinners.

A single session produces four to six components, and they COMBINE.

💡 Aha moment. ⚠️ Four components make far more than four meals.

Rice + chicken + roasted vegetables + yoghurt sauce is a bowl. The same chicken + slaw + a wrap is a wrap. The same roasted vegetables + lentils + vinaigrette is a salad. The same lentils + tomato sauce + pasta is a pasta dish.

⚠️ The combinatorics defeat palatability decay, because no two dinners are the same dinner — and the total cooking time is a fraction of cooking four dinners.

⚠️ Two rules that make it work:

1. ⚠️ Keep components UNSEASONED or lightly seasoned. A chilli-and-cumin chickpea is a chickpea that can only go in one direction. A plain cooked chickpea goes anywhere. ⚠️ Season at assembly, not at batch.

2. ⚠️ Store them separately. Combined food degrades faster and cannot be recombined.

⚠️ A worked week from one session

Because "batch components" is abstract until you see a week fall out of it.

⚠️ Sunday, roughly 100 minutes, mostly unattended:

⚠️ Everything unseasoned or lightly seasoned. Everything stored separately.

⚠️ And here is the week:

💡 ⚠️ Five dinners from one session, no two the same, and the longest weekday assembly is fifteen minutes.

Compare the failure mode it replaces: a big pot of chilli, eaten Monday, Tuesday, Wednesday with declining enthusiasm, and thrown out on Thursday.

⚠️ Same Sunday. Same ingredients cost. Completely different week.

⚠️ And note what makes it work: the components are boring on their own and interesting in combination, the sauces do the differentiating, and nothing was committed to a single dish at the point of cooking.

31.9 The freezer, assessed

⚠️ The most under-used appliance in most kitchens and the binding constraint in many.

⚠️ Four practical rules:

1. ⚠️ Freeze in the portion you will actually use. A single block of four portions requires you to want four portions. 2. ⚠️ Label and date everything. Unlabelled frozen food becomes archaeology and then waste. 3. ⚠️ You can freeze up to the use-by date (Chapter 30 §30.11b), which is the single most useful freezer fact and is not widely known. 4. ⚠️ Keep a list on the door. Otherwise you buy things you already have and the bottom stratum is never excavated.

⚠️ Equipment that earns its space, and equipment that doesn't

Briefly, because kitchen advice is usually a shopping list and most of it isn't needed.

⚠️ Two that genuinely divide people:

⚠️ A slow cooker or pressure cooker. Excellent for the batch-component session, cheap to run, and it turns cheap cuts and dried legumes into food while you do something else. ⚠️ If your constraint is attention rather than time, this is the one item worth buying.

⚠️ A microwave. Underrated for reheating components, cooking grains, and steaming vegetables — and Chapter 17 §17.9 established that it preserves nutrients better than boiling.

⚠️ The general rule: buy equipment that removes friction from something you already do, not equipment that requires a new habit.

A sharp knife removes friction. A spiralizer requires a habit.

31.10 Defaults, and the cost of deciding

⚠️ Every meal decision has a cost, and you make a lot of them.

⚠️ The fix is defaults — and defaults feel like a failure of imagination until you notice what they buy.

⚠️ The observation people resist: variety is a value, not a nutritional requirement, above a fairly low floor.

You need a varied enough diet to cover micronutrients (Part III) and a fibre-diverse enough one to feed a microbiome (Chapter 27). ⚠️ You do not need a different dinner every night, and treating that as a requirement is a major reason planning collapses.

31.11 ⚠️ The bad-night plan

⚠️ The single highest-yield element in this chapter, because it is what breaks plans.

⚠️ Every week contains at least one night with nothing left. The plan must contain that night, in advance, by name.

⚠️ What a bad-night plan is:

• Decided in advance, so no decision is required when you're least able to make one
• ⚠️ Made entirely from shelf-stable or frozen ingredients, so it is always possible
• Under fifteen minutes
• ⚠️ Genuinely acceptable to you — not a punishment
• Bought as part of the standing list, and NOT eaten on good nights

⚠️ Examples, and they should be your examples:

⚠️ The point is not that these are optimal meals. It is that they exist, require nothing, and are better than the alternative.

The alternative is a takeaway, and the takeaway isn't the problem — ⚠️ the problem is that the takeaway means the plan "failed," and a failed plan gets abandoned (§31.2's third failure mode).

⚠️ A bad-night plan converts a plan failure into a plan feature.

31.12 Leftovers, and cooking once

⚠️ "Planned-overs" — cooking deliberately more than you need — is different from having leftovers.

⚠️ Food safety, briefly and properly (Chapter 20 §20.12b): cool leftovers quickly, refrigerate within about two hours, use within a few days, and reheat thoroughly. Rice specifically should be cooled fast and refrigerated promptly — ⚠️ Bacillus cereus is the reason, and it is a real if uncommon cause of illness.

31.12b ⚠️ Breakfast and lunch, which nobody plans

Meal planning is almost always dinner planning. ⚠️ Which is odd, because breakfast and lunch are roughly two-thirds of eating occasions and are where most people's diets actually go wrong.

⚠️ Chapter 22 found it: for most households, Group 4 intake is concentrated in DRINKS, BREAKFAST and SNACKS — not dinner.

Breakfast

⚠️ This is where defaults belong (§31.10). Nobody needs breakfast variety, and the decision cost is paid every single morning.

⚠️ Pick two. Rotate them. Stop deciding.

Lunch

⚠️ The harder one, and the one most often bought rather than made.

⚠️ The costing that changes behaviour: a bought lunch at £6–9, five days a week, is £1,500–2,200 a year. The same lunch from planned-overs is close to free, because the food was already bought.

⚠️ And this is the largest single sum in Part VI, from the meal people plan least.

31.13 Waste is a planning problem

⚠️ Household food waste is substantial, and it is largely a consequence of the four failure modes.

⚠️ That last rule is worth its own line, because it solves several problems at once.

Fresh produce for Monday to Wednesday; frozen and tinned from Thursday onwards. Nothing spoils, the late-week nights are the low-capacity ones anyway, and frozen vegetables are nutritionally comparable (Chapter 20 §20.15).

31.13b Planning for more than one person

⚠️ Almost all meal-planning advice is written for a single decision-maker feeding themselves, and almost nobody is in that situation.

⚠️ The additional problems a household creates:

⚠️ What actually works:

1. ⚠️ Deconstructed meals. The components approach solves this too (§31.8) — the same elements on the table, assembled differently by each person. Tacos, bowls, baked potatoes, pasta with the sauce separate. One cook, several dinners.

2. ⚠️ One meal, with a guaranteed acceptable element. Chapter 25 §25.6's division of responsibility, applied to adults as well as children: something on the table that each person reliably eats, and no alternative meal produced.

3. ⚠️ Split the roles explicitly. Planning, shopping and cooking are three jobs. A household where one person does all three and everyone else eats is not a system — it is a person, and it fails when that person has a bad week.

4. ⚠️ And a household veto list, agreed once. Three or four things each person genuinely won't eat. Short lists are workable; unlimited veto isn't.

⚠️ The thing to avoid, because it is where households actually break: the short-order cook.

Producing a different meal per person converts one cooking event into three, permanently, and it is the commonest reason a planning system collapses in a family — not a lack of recipes.

31.14 ⚠️ Who this chapter is for — and who it isn't

Stated plainly, because Chapter 22 §22.11 named these constraints and this chapter cannot solve them.

⚠️ If the right-hand column describes you, this chapter's advice is not wrong — it is addressed to someone else, and following it will mostly produce a sense of failure.

⚠️ Chapter 32 is written for the right-hand column, deliberately, and it is not a simplified version of this one. It is a different problem.

⚠️ And a middle case worth naming: many households are in the right-hand column on some rows and the left on others. Kwabena has time and money and no cooking confidence in his partner; Marisol has skill and no money; someone else has both and a freezer compartment the size of a shoebox.

⚠️ Read the table row by row rather than as a verdict, and take the parts of this chapter that match your left-hand rows.

31.15 The nine slots — the whole system on one page

⚠️ If you do nothing else in this chapter, do this.

⚠️ 1. Map the week (§31.3). Four minutes. Who's in, when you're back, energy, nights out.

⚠️ 2. Label each night's CAPACITY — high, medium, low, zero, out.

⚠️ 3. Put the ZERO night in first, as the bad-night meal (§31.11).

⚠️ 4. Put the HIGH night in second, as the cook-once night — deliberately oversized (§31.12).

⚠️ 5. Put the night AFTER it in third, as assembly from the surplus.

⚠️ 6. Fill the MEDIUM nights from your repertoire (§31.5). No new recipes.

⚠️ 7. One new recipe maximum, on a high-capacity night.

⚠️ 8. Shop your fridge and freezer (§31.13), then write the list — standing list plus variable, ordered by shop layout (§31.6).

⚠️ 9. Buy fresh for the first half of the week and frozen or tinned for the second.

⚠️ Fifteen minutes on a Sunday, and it removes the four failure modes structurally rather than requiring you to overcome them.

🧾 What planning is actually worth

⚠️ The numbers are larger than people expect, and one line dominates.

⚠️ The bought-lunch line is the largest single number in Part VI, and it comes from the meal nobody plans (§31.12b).

⚠️ And the takeaway line is bought by ONE decision — naming a bad-night meal in advance (§31.11). Not by discipline, and not by cooking more.

⚠️ Fifteenth chapter where the better-evidenced option is also the cheaper one — though here the currency is partly time, and §31.14 is about who has it.

31.16 What we don't know, and how firmly I hold this

What we don't know

⚠️ Whether planning causes better diet quality or accompanies it. The evidence is observational and the confounding is obvious (§31.1). Whether batch cooking survives past a few months in most households — I have not seen good longitudinal data and I would like to. ⚠️ And whether any of this transfers to households in §31.14's right-hand column, which is the question Chapter 32 has to answer.

⚠️ How firmly I hold these

⚠️ A note on evidence in Part VI. This chapter is closer to craft than to science, and I would rather say so than dress practical advice in the vocabulary of trials.

What it rests on: Chapter 22's finding that the food environment determines intake, Chapter 10's finding that adherence beats composition, and a set of observations about why plans break that I think are correct and cannot prove.

Spaced Review

1. (Chapter 22) Why does §31.6 call the shopping list the intervention?

⚠️ Because planning's mechanism is changing what is IN THE HOUSE, and Chapter 22 established that what's in the house determines a great deal of what gets eaten. The plan is a delivery mechanism for the list; the list is what changes the food environment. ⚠️ Which is better supported than the planning literature itself.

2. (Chapter 10) Why does one failed night end a whole meal plan?

⚠️ The all-or-nothing pattern. Chapter 10 §10.4 found that adherence beats composition and that plans with no tolerance for deviation get abandoned entirely rather than partially. Kwabena's Thursday failed because Wednesday did, and then the plan was "not working." ⚠️ §31.11's bad-night plan exists to convert a plan failure into a plan feature.

3. (Chapter 30) What is the single most useful freezer fact from the previous chapter?

⚠️ You can freeze food up to its use-by date, which resets the clock. "Use by" is safety, "best before" is quality (Chapter 30 §30.11b) — and freezing on the use-by date is permitted, widely applicable, and not widely known.

Project Checkpoint: Map Your Week

Component thirty-one. ⚠️ Do this before you plan anything.

Step 1 — The map. ⚠️ Four minutes, for next week.

⚠️ Count them: how many HIGH nights do you actually have?

Step 2 — ⚠️ Name your bad-night meal. Now, before you need it.

My bad-night meal is: __ Everything it needs is shelf-stable or frozen: _ ⚠️ It's on my standing shopping list: _

Step 3 — Count your repertoire. ⚠️ Meals you can cook without a recipe.

I have ____ of them. (⚠️ Ten to fifteen is the target. Most people have six to eight and underestimate it.)

⚠️ Which one am I going to add this month? __

Step 4 — ⚠️ The orphan audit. Look in your fridge right now.

Half-used ingredients: ____ Things bought "just in case": ____ ⚠️ Estimated value of what will be thrown away this week: ____

Step 5 — Build the standing list. ⚠️ The 12–20 things you always buy (Chapter 30's decoded shelf). Save it in your phone.

Step 6 — And run §31.15's nine slots for one week.

⚠️ Then compare: how many nights went as planned?

Kwabena went from three of seven to six of seven, with the same amount of cooking.

Next checkpoint (Chapter 32): the real budget — what you actually spend, and where it goes.

Chapter Summary

⚠️ Kwabena's plan achieved three nights of seven and £14 of waste, in a household with time, money, a kitchen and a freezer. His diagnosis — "I need to be more disciplined" — was wrong.

⚠️ THE FOUR FAILURE MODES, and none is discipline:

⚠️ All four are properties of the PLAN, not the person — which is good news, because plans are much easier to change than people.

⚠️ §31.3–31.4: MAP THE WEEK BEFORE YOU PLAN IT, and WORK BACKWARDS. Classify each night by CAPACITY — high, medium, low, zero, out — then assign a meal TYPE, then choose food. ⚠️ "A plan with seven medium-to-high nights in it is not a plan. It is a wish with days attached."

⚠️ §31.5: aim for 10–15 meals you can cook without reading anything. One new recipe a week, on a high night, cooked three or four times before it counts. ⚠️ Most cuisines run on a small repertoire — "rotation is not failure; it is how eating has always worked."

⚠️ §31.6: THE SHOPPING LIST IS THE INTERVENTION. A plan without a list is a wish. Standing list + variable list · ordered by shop layout · checked against the fridge first · ⚠️ and nothing "just in case." Online ordering removes the aisle and is one of the few effective structural interventions available to an individual.

⚠️ §31.7–31.8: batch cooking's real cost is PALATABILITY DECAY — the same chilli four nights running fails and the rest goes in the bin. The fix is ⚠️ BATCH COMPONENTS, NOT COMPLETE MEALS: a grain, a protein, a roasted vegetable tray, a quick vegetable, a sauce, a flavour hit. ⚠️ Four components make far more than four meals, and no two dinners are the same dinner. Keep them unseasoned; season at assembly; store separately.

⚠️ §31.11: THE BAD-NIGHT PLAN is the highest-yield item in the chapter. Decided in advance, shelf-stable, under fifteen minutes, genuinely acceptable, on the standing list. ⚠️ It converts a plan failure into a plan feature.

⚠️ §31.13: buy fresh for the first half of the week and frozen or tinned for the second. Nothing spoils, the late-week nights are the low-capacity ones anyway, and frozen vegetables are nutritionally comparable.

⚠️ §31.14 states the limit honestly: this chapter is for households with some time, kitchen, freezer, money and confidence. If the binding constraints are the other column, Chapter 32 is the chapter — and it is a different problem, not a simplified version of this one.

⚠️ And §31.16 concedes the evidence: this chapter is closer to craft than to science. The planning literature is observational and confounded; what it rests on is Chapter 22's food-environment finding and Chapter 10's adherence finding.

The one thing to remember: ⚠️ Three of Kwabena's four failures were predictable on Sunday, and rearranging the same seven meals fixed two of them without changing a single ingredient.

What's Next

Chapter 32 is ★ the chapter this one has been deferring to, and it is the harder problem.

⚠️ Eating well on a genuinely constrained budget — not as a lifestyle choice, and not by shopping at farmers' markets. What food actually costs per calorie, per gram of protein, and per unit of fibre. Why cheap food is energy-dense, and what that does to Chapter 22's argument. Food deserts and food swamps. Time as a currency. The cost of a small fridge, an unreliable cooker, and buying day to day.

⚠️ And the question this book has been circling since Chapter 22: whether "eat better" is advice or a description of an income bracket.