Why do you keep cooking the same few meals?

It is a strange kind of contradiction: a phone full of saved recipes, a shelf of cookbooks, months of half-remembered things a friend cooked once — and dinner is still pasta, stir-fry, tacos, a curry, eggs on toast, on rotation, most weeks of the year. That is not a failure of imagination and it is not a lack of options. It is what happens when a decision has to be made quickly, under time pressure, using only what can be confirmed from memory — and most of what you know how to cook cannot be confirmed that way.

The rotation is made of recipes you can vouch for without looking

At six o'clock, tired, deciding what to cook is not really a search through everything you know how to make. It is a much narrower search through everything you can be reasonably sure you can make right now, without a trip to the fridge to check three separate things first. A stir-fry works because it is built from a short list of things a stocked kitchen almost always has some version of — an onion, a protein, rice, a sauce from the door of the fridge. You do not need to verify any of that. You already know it is probably true. A dish that depends on one specific thing — the aubergine you bought on Tuesday, the tin of coconut milk from a few weeks ago, the block of feta sitting in the door — cannot be vouched for the same way, even when every one of those things is actually there. It requires a check, and a check is exactly the kind of extra step a tired brain skips at the moment it is trying to close the question, not open it further.

Every recipe you do not cook this way is invisible, not rejected

This is the part that makes the rut feel confusing from the inside. It is not that the other recipes lost some kind of vote — they were never actually considered, because considering them would have meant walking over and looking. A saved recipe that needs a specific ingredient gets silently skipped, night after night, not because anyone decided against it, but because nobody's mental shortlist at 6pm includes a step where you go and verify the fridge before choosing. Over months, that quietly trains the rotation: the handful of recipes that never require checking anything get cooked constantly, and everything else — including dishes you genuinely like, and paid for the ingredients of — sits unconsidered, not because it was worse, but because it was never confirmable from the sofa.

Buying for a new recipe does not widen the rotation on its own

The instinctive fix is to try something new — buy the ingredients for an unfamiliar dish, cook it once, feel briefly like the rut is broken. But a single new recipe rarely survives, because it has exactly the same problem as all the others once the novelty wears off: it depends on something specific, and specific things fall out of the mental shortlist the moment they stop being fresh in memory. A jar bought for one dish two weeks ago is no more checkable at 6pm than it ever was. This is a big part of why kitchens end up with ingredients bought "for" a recipe that never gets repeated — the rotation did not actually widen, it just borrowed one dish for one night and then closed back up to the same handful it always relies on.

A written meal plan usually does not fix this either

A weekly plan looks like the obvious answer, but it is built from the same short mental list, just filled in a few days ahead of time rather than at 6pm. Planning on Sunday does not change which dishes you can confirm are makeable — it only moves the guess earlier, to a moment when you are relying on memory of what is in the fridge rather than an actual look at it. That is why a meal plan so often ends up listing the same rotation anyway, just written down with more confidence: the underlying limitation, not knowing at a glance what specific things are actually on hand, was never addressed. It was just scheduled.

What actually widens the rotation

The fix is not more recipes, more planning, or more willpower to try new things. It is removing the need to check from memory in the first place. If what is actually in the fridge is visible at a glance — not recalled, not guessed, just looked at — then a dish that depends on one specific thing becomes exactly as confirmable as a stir-fry. The aubergine stops being invisible the moment you can see it is there without walking over to check, and a recipe built around it moves from "probably not tonight" to "yes, actually" without anyone having to remember anything.

• Keep a list of what is actually in the house, not a memory of what is usually there. The rotation is powered by which dishes you can vouch for instantly — widening what you can vouch for widens the rotation.
• Let a recipe be "one ingredient away" instead of simply excluded. Most of what gets skipped is not missing everything, just one thing you were not sure you had — a completely different, much smaller gap to close.
• Notice which ingredient keeps sitting unused, and ask why, instead of quietly rebuying the same handful of staples. The pattern usually points straight at what never gets checked, not at what you do not like cooking.
• Let the thing that has been in there longest suggest tonight's dish sometimes, rather than always starting from a dish name. Starting from the ingredient instead of the recipe is often what actually breaks a rotation that starting from more recipes never does.

This is close to what Munch's recipe matching is built to do. It groups recipes against your actual fridge, freezer and pantry — what you can already cook in full, and what you are only one ingredient short of — rather than leaving it to memory to guess. A dish that depends on the aubergine sitting in the crisper surfaces the same way the stir-fry always did, because the app already knows the aubergine is there. Nobody has to remember to check.

None of this means the same five meals are a problem to be solved through sheer effort, or that the rotation says anything about how adventurous a cook you are. It is what happens, reliably, whenever a decision has to be fast and the only information available for it is memory. Give that decision something more reliable to check than memory, and the rotation widens on its own — not because you tried harder, but because more of what you already own became visible at the moment it mattered.