>I have a theory that 'calorie counters' are shockingly inaccurate (one other poster mentioned a 25% margin is 'fine'!)

Accurate or inaccurate is only meaningful related to the task and its requirements. 25% can be totally acceptable margin of error for the task. We use even bigger margins in lots of ventures (determining which startup will have a succesful exit to fund, for one).

And yes, the mere act of quantifying helps. But quantifying with even 25% error is still better than just writing down "ate 5 cakes", especially if one doesn't eat too many repeats of the same food.

Not sure in what reasoning one can complain for a method with 25% margin of error (say), but be OK with a method like "5 cakes" which still applies quantification, just in an even more vague and hazy sense.

5 cakes is much worse than 5 carrots, for example, but with merely writing down how many you ate, you have to rely on a far more relative guesstimation of their relative harm than you would be if you were counting their calories and being off by 25%).

> Not sure in what reasoning one can complain for a method with 25% margin of error (say), but be OK with a method like "5 cakes" which still applies quantification, just in an even more vague and hazy sense.

One method implies rigour and the other one is honest about what it is setting out to do

But you seem to be the only one assuming and/or bringing up rigor in this discussion.

Everybody said it's a quick ballpark figure / back of the envelope style calculation.

The mere fact "cake bad, carrots good" everybody knows. It's not much information concerning "Did I ate too much today?". One can have a caloric budget and stay within it (more or less) without having to be perfect in measurements (or sticking to carrot because it's easier to know its light).

I actually agree with you - to a large degree, counting the calories is enough to change behavior in itself, without even attempting to modify your diet at all.

The point remains, though, that it really doesn't matter how accurate or precise you are, simply that you're gathering data and using it to make measurable changes in your diet.

Your way is perfectly fine too! The nice things about calories (just like money) is that they're a universal medium of comparison, so you can compare your cakes to steaks. But if you simply want to look at a category and say "I am eating N of these, I need to eat N-1 to lose the weight", that's perfectly functional!

The thing I hate is when people make totally non-empirical diet changes, and then lament that they aren't losing weight. You just have to measure and adjust.

My point though is what you're "measuring" is quite difficult to do in a clinical setting, let alone every day living your life.

The above poster talking about the deli sandwich mentally breaking down all the ingredients... I'm happy to be proven wrong but I simply don't believe you can do that with a reasonable degree of accuracy.

I suppose this is turning in to quite a pedantic argument about the definition of "count". I'd call what everyone has said they're doing more accurately a "calorie estimating food diary".

My initial post was genuinely curious about how people can be so accurate when eating foods from a variety of non pre-measured sources. It turns out they're not being accurate.

>The above poster talking about the deli sandwich mentally breaking down all the ingredients... I'm happy to be proven wrong but I simply don't believe you can do that with a reasonable degree of accuracy.

You are just overestimating both the difficulty and the degree of accuracy required.

While at the same time, still relying on even more vague terms, to determine if you ate too many cakes etc (as per your other example).

I'm not claiming any degree of accuracy in the estimates, just that without tricking yourself in to think you're being rigorous you can achieve pretty much the same results. I outlined the mechanism that I think these 'counters' are actually achieving results by in spite of their atrocious data collection.

I can't believe people are defending such poor data collection practices. You'd be all over it if someone else was selling their results based on such inaccurate data but in this case it's fine?

If you're going to call it counting you need to be accurate.

>I outlined the mechanism that I think these 'counters' are actually achieving results by in spite of their atrocious data collection.

The mechanism is simple: they reduce their caloric intake, because they can track how much they eat. More or less: it doesn't have to be perfect, nor is it "atrocious" if it isn't. And you can easily just round the numbers up ("I calculated 500 for this thing, but let's say it's 600 just to be safe").

You seem to believe that any kind of "back of the envelope" / "ballpark" calculation is useless. Or that people only eat complex multi-part meals with no nutricion information, and have to gauge everything from zero all the time.

So I went looking for any studies about peoples ability to estimate calories, since any under/over estimation compounds either way in terms of results.

It turns out like everything, that's a tricky thing to study. When people are aware that their meals are going to be scrutinised they change what they eat for the duration of the study.

I wasn't able to find a recent study that investigates people's ability to estimate calories and the variance in their estimations.

> You seem to believe that any kind of "back of the envelope" / "ballpark" calculation is useless.

I do think that anything above a "good, bad, not sure" estimate is probably going to be so inaccurate as to not be worth the effort. However the act of trying to count calories itself promotes a mindfulness of what we're eating and that can induce change.

it doesn't matter if it's accurate, it matters if it's precise, and consistent. kind of like the scale you stand on, or weigh your food with.

I disagree that you can possibly be precise and consistent across foods without preparing them yourself.