How wrong is your calorie count?

Wrong enough to matter, and not in the way that stops it working. What the error actually is, and why consistency beats accuracy.

Around 7 minutes · Calories

A food diary presents every entry to four significant figures. 486 kcal. 32g of protein. The precision is real in the sense that the arithmetic was done correctly, and largely fictional in the sense that the inputs were never that good.

This is not an argument against counting. It is an argument for knowing which parts of the number to trust, because that turns out to decide whether counting works for you at all.

The label is a tolerance, not a measurement

Packaged food is the most trustworthy thing in your diary, and it is still approximate. Nutrition labels are produced from a mix of laboratory analysis and calculation from recipe, then checked against a permitted tolerance band rather than an exact figure. The bands are not tiny — guidance in the range of plus or minus a fifth on energy is commonly cited for many products.

There is a real reason for that. A batch of crisps is fried in oil whose uptake varies; a loaf's moisture varies with the day. A regulator demanding an exact figure would be demanding something food cannot deliver.

So a 500 kcal ready meal is a ready meal that is probably somewhere near 500. That is fine. It is also worth knowing before you agonise over a 20-calorie difference between two yoghurts.

Where the error is actually large

Labels are the good case. The error in a real diary mostly comes from three other places.

Portions

"A bowl of pasta" covers a range of about two hundred calories. Eyeballing is worse than most people think, and it is worse in a consistent direction: portions of things we like get underestimated, and the error grows with the size of the serving. Scales fix this and almost nobody uses them forever — which is fine, as long as you know that the estimate is where the uncertainty lives.

Oil, butter, sauces and dressings

Fat is nine calories a gram and effectively invisible once it's in the pan. A tablespoon of olive oil is around 120 kcal. A "healthy" salad can carry more calories in its dressing than in everything else on the plate. Most large unexplained gaps in a diary are here.

The things that never get logged

The mouthful of the kids' dinner. Three chips off someone else's plate. The milk in four coffees. A handful of nuts while cooking. None of it feels like eating, and a few hundred calories a day of it is entirely normal.

Why this doesn't sink the whole exercise

Here is the part that matters, and it is genuinely good news.

If your logging is wrong by roughly the same amount every day, the error largely cancels out of the thing you actually use the diary for. You are not trying to learn the true calorie content of a shepherd's pie. You are trying to learn what happens to your weight when you eat a certain amount of food — and the diary's job is to give you a consistent unit of measurement, not a true one.

Work out your maintenance level from your own logged intake and your own weight trend, and any systematic error is baked into both sides of that arithmetic. If you log 15% low, the maintenance figure you derive is 15% low too, and eating to it works exactly as intended. The numbers are wrong and the decisions are right.

Which reframes what a good diary is. Not the one with the most precise entries — the one you keep filling in on the bad days.

What to actually do about it

• Be consistent before you're accurate. Same foods logged the same way, every day, including the ones you'd rather not.
• Weigh the calorie-dense things. If you only ever weigh three categories, make them oils and butter, nuts and nut butters, and cheese. That is where the big invisible errors are.
• Log the bits that don't feel like eating. The handful, the mouthful, the milk. That's the gap between a diary that explains your weight and one that doesn't.
• Don't re-litigate 20 calories. Two entries that differ by 4% are the same entry. Spend the attention on the portion size instead.
• Let your own data correct the bias. Four weeks of logging against your weight trend tells you what your numbers mean in practice, whatever they say on the label.

Why we mark our own estimates

QuicKcals labels every entry as either verified or estimated, and shows the portion an estimate assumed. That isn't modesty. A number you can see the reasoning behind is one you can correct in a tap; a number presented as fact is one you either accept or abandon the app over. Given everything above, pretending to a precision the data can't support would be the actual dishonesty.