Buy The Pot Worth Handing Down.

I set out to score three frying pans. And ended up rebuilding the ruler.

A day inside the Shake method scoring the cookware category.

Updated 11 September 2026: two more pans have since been scored under the rebuilt model — Lodge at 88.75% and Xtrema at 75%. The ruler held.

---

A daily ritual of gathering starts in the heart of my home when I take out the pots. Pets appear at my feet, kids start homework at the kitchen island, granny is there to chit-chat, and a glass of wine materialises from my better half. Not sure if it’s the socialising or a hint of hurry-up, out of pure hunger. Maybe both.

The pots are the oldest things in my kitchen and the only ones I expect to outlive me, a strange thing to notice about an object you mostly use for supper. Not because the right pot makes you a better cook, but because the good ones stay, and what gathers around them stays with them. So buy the pot worth handing down.

There’s a version of my job that looks simple from the outside. You pick three pans people actually buy. You read what the brands publish. You put the numbers into the scoring model. You publish. Somebody in Atlanta reads it on their phone in a shop and makes a slightly better decision than they would have.

That was the plan for my Tuesday. It took about forty minutes to fall apart and the rest of the day to understand why — which turned out to be the more useful outcome, though it did not feel like it at eleven in the morning.

The problem with a perfect score

Shake rates every product on four pillars.

• People — what it means for the person using it.
• Animals — cruelty in sourcing, harm to wildlife.
• Gaia — what it costs the planet, raw material to bin.
• Frequency — how minimal, whole and intentional the thing actually is.

Each pillar starts at 100. Evidenced concerns take points off, verified virtues add some back, and the final number is the average, and pops out a percentage.

First run, all three pans scored over 90. Lovely numbers. Completely meaningless.

Because three of the four pillars had scored everything 100 — not because the pans were flawless, but because we had never written down what to look for. Our materials repository, the list of every substance we can prove is a concern, had two entries that applied to cookware. Both under People. Animals, Gaia and Frequency had nothing at all.

We have a rule that a concern cannot count against a product unless it’s in that list with an evidence tier and a named source. It’s a good rule — it’s why we can rate a brand that never asked to be rated and defend it afterwards. Applied to a category we hadn’t properly scoped, it produced a score that was arithmetically correct and said nothing true.

The sharpest illustration was Le Creuset. A brand that publishes almost nothing about what the enamel your food touches is made of, has no leach testing anywhere, and couldn’t answer two of the three questions that applied to it — landing in the Aligned band. That didn’t sit right.

The animal in the kitchen

The rebuild started with a question that had nothing to do with methodology. We sometimes cook for our animals. And they’re always at our feet while we cook — getting affected by the off-gas, maybe.

It’s the kind of thing you only think of with a dog in your kitchen rather than a spreadsheet in front of you. And it exposed something in a pillar we’d been using for months.

Our Animals pillar was carrying three ideas under one name. Cruelty in sourcing. Harm to wildlife and waterways. And nothing at all for the animal actually in the room.

Thank goodness I don’t own an elephant.

The rules had a reason. Exposure route and vulnerability — the things that measure how much of a substance reaches a body — don’t apply to Animals. Which is correct, for a river. A river has no lungs. But a cat in a kitchen is not a river. It’s a body, in a room, with a dose and a route and a vulnerability, and everything those multipliers were built to measure applies to it exactly as it applies to you.

So we went looking for evidence and found something we should have known. Overheated non-stick coatings kill birds. Not in a contested, more-research-needed way — it’s a named condition in the standard veterinary reference, documented for decades, and the usual clinical sign is that the bird is simply dead. Cornell’s veterinary school recorded three ducks dying about twelve hours after someone installed a PTFE-coated heat lamp in a closed room.

“Canary in a coal mine” is not a metaphor. A bird’s lungs pull more out of every breath than ours do. That’s the entire mechanism.

So the most useful sentence on our new cookware page cost nothing and took an afternoon to find: if you keep a bird, keep it out of the kitchen, and never heat an empty non-stick pan. An empty pan on a high flame gets there in minutes. Almost nobody has ever been told.

The word that was wrong

The second question was about the planet pillar. How biodegradable is this product?

Right instinct, wrong word — and cookware shows why most clearly. You don’t want a pan to biodegrade. A pan that breaks down is a pan that failed. A cast-iron skillet from 1940 still in use and still iron in 2140 is the best possible outcome, and it’s the precise opposite of biodegradable.

The honest question isn’t does it break down. It’s how long does it last, and where does the material go when it finally stops.

That reframed the pillar. A coated aluminium pan with a four-year life over a body nobody can separate for recycling is a poor result — not because it’s made of terrible things, but because it’s a short-lived object made of materials that never come back. Enamelled cast iron is a coated product with a hundred-year life and infinitely recyclable iron underneath. It’s a very good one.

How long does it last and where does the material go is answerable for iron, steel, aluminium and ceramic alike. Nothing hangs on which one we happened to research first.

Le Creuset still wins that pillar outright, at 100. But now it wins because cast iron genuinely is the most durable and recoverable object on the page — a reason we can put in writing.

The thing we nearly got badly wrong

Somewhere in the middle of this we very nearly published something untrue.

The reasoning went: aluminium has a bad reputation where health is concerned, everyone knows that, so aluminium leaching into a person is probably bad too.

It’s one of the most widely believed health claims in circulation. The European Food Safety Authority assessed it directly and concluded that exposure to aluminium through food does not present a risk for developing Alzheimer’s disease. Had we published that link, we’d have done precisely the thing our first principle exists to prevent — rating by a frightening name instead of by evidence. It would have read well. It would have been shared. It would have been wrong, and we’d have deserved the correction.

But there’s a real concern underneath the myth, and it’s a better one. EFSA’s tolerable weekly limit is one milligram of aluminium per kilogram of bodyweight, and intake may exceed that in a significant part of the European population — children sitting higher against the same line. Most of that is food, not pans. But aluminium migrates more into acidic and salty cooking, and considerably more where a coating has worn through to bare metal. And there’s a group who should genuinely pay attention: people with impaired kidney function.

Smaller, quieter and more useful than aluminium is bad for you. Also true.

It produced the clearest line on the finished page: a pan with aluminium inside it is not a pan with aluminium in your food. In a bonded pan it’s sealed between layers of steel and never touches anything you eat. It’s there to spread heat. What matters is the surface — which is exactly why we ask brands to publish their construction layer by layer.

Grading effort, or grading evidence

The last decision was the hardest, and it cost the most likeable brand on the page about seven points.

Made In publishes more about its pans than anyone else in the category — five layers, every one named, thickness, country of manufacture. Genuinely more than we usually get. It also says its pans are made without PFAS, and we could not find a test report behind it.

We had a choice. Cap only the brands that can’t tell you what the surface is made of, and let Made In through on everything else it discloses. Or hold every unevidenced claim to the same line.

What decided it was a brand we’d assessed a fortnight earlier, which published its composition too, couldn’t back its PFAS-free claim either, and got capped. If Made In went through, we’d have treated two brands differently for identical behaviour — and the only visible difference would have been that we liked Made In more.

Not a position anyone would want to defend in public. So: an unverified claim is unverified. Shake grades evidence.

Where it landed

What the day was actually about

We published three numbers. It took eleven hours, and most of that wasn’t spent on pans.

It was spent finding out that our own ruler could measure a quarter of the question, that a pillar named after animals had no room for the animal in the room, and that we were three sentences away from repeating a health claim the evidence doesn’t support.

Every one of those would have been invisible in the finished piece. The scores would have looked exactly as confident. That’s the uncomfortable thing about this work — a wrong answer and a right one are the same shape from the outside, and the only difference is whether somebody went and checked.

So we go and check. And when the checking says the scoring model is broken, we fix the model rather than publishing around it.

It’s slower. It’s why there are five pans on that page and not forty. We think it’s the only version of this worth doing.

#ADifferentLens