How accurate is MFP supposed to be?

makotron · June 2018

I've been consistently losing an average of 2 lbs per week, since the beginning of March. Lost a total of 40 lbs. Started using MFP and a Polar M600 in May, and so far it's been a very frustrating experience.

The M600 is a mess of its own, but it gives me somewhat accurate step count and calories burned. Which leaves most of the problems to MFP.

According to MFP, in order to lose 2 lbs I need to cut 7350 calories, which is approx. 7,35 calories per gram of fat. When I convert my weekly weight loss, it's always off by a LOT of calories. I created a random food just to add/remove calories according to my weight loss in a week. I've been doing this for almost two months. One week I had to add 5,000 calories, the next one I had to remove 1,000 calories. It's totally random, which means I can't trust the food database at all.

There's also the polar calorie adjustment, which I thought would solve these problems, but it only made it worse. If it's supposed to represent the difference between MFP and Polar daily calories, then why would you add exercise data from Polar to MFPs daily projection? Doesn't make sense and gives me weird numbers. Also, for some reason it randomly miss Polar training sessions. So one day I get 1600 calories burned in exercise, the next day I get only 280. But my workout has been THE SAME since day one - 3 to 4 hours of walk with the dog, daily.

Maybe I'm doing something wrong, but the 40 lbs I lost tell me otherwise. Is there a way to obtain more accurate numbers in MFP?

RAinWA · June 2018

How are you measuring your intake? Are you using a food scale for all solids?

You adding random food to make your intake line up with your actual loss is confusing. Weight loss is not going to be the same week to week and isn't always going to line up perfectly with your intake. You could have a perfect 7000 calorie deficit one week and lose nothing or 1.1 or 2 or 2.2 or any other random number of pounds. You have to allow for fluctuations too.

I would suggest you read the most helpful posts at the top of the forum. Lots of good information there on how MFP works.

livingleanlivingclean · June 2018

Mfp is as accurate as you make it....

You also need to realise that weight loss is different to fat loss. Your weight is affected by many factors that are put of your control.

100_PROOF_ · June 2018

Mfp is as accurate as you make it.
You enter the info ( or your connected devices do)
If something isn't adding up then you need to backtrack and see where and what is going wrong.
If you are not losing then you are not creating a calorie deficit.

makotron · June 2018

Thanks for the reply. Yes I've been using a scale for everything I eat. I'm also allowing for fluctuations, but I think it's weird when MFP tells me that I'm still 5.000 calories UNDER in the week - meaning I could've eat those extra 5000 calories according to MFP, but if I did that I would probably have added some weight. So I just add the random food to fill the number gap in MFP.

If it was off by a thousand or less calories every week, that's acceptable. But if it's going to give me such random and large numbers, I wonder what's the point in using MFP as a guide. At this point, I'm just trying to find a reason to keep using it. Sorry if I asked in the wrong place, or sorry if this is a stupid question to ask anyway.

Kim_S_G · June 2018

It sounds to me like you do not have MFP set up correctly and/or know how to use it. Did you enter all your stats (age, gender, height, weight, activity level) into MFP, along with your weight loss goal of 2 pounds per week? Once you enter all that information, you will be given your daily calorie goal (how many calories you should be eating per day). Record what you eat in MFP.

I do not understand why you are adding /removing calories.

rickdkitson · June 2018

There are so many sources of error in any calorie in and out plus weight gain loss system to make this whole effort not much more than a WAG. Some are specific to MFP and some are inherent in all nutrition and exercise measurements.

The data base of user added foods is absolute crap. Just about any food you look at has multiple user entered data and it can vary significantly especially for stuff “homemade” or generic like fruits.

MFP should as a minimum delete all entries with less than a couple of verifications within a month or so of being entered. Then take a look at the foods in the data base with wide ranges of values for the same food. These should be verified by a knowledgeable person.

New foods added, except for an individual’s own use, should be published in a separate section and independently verified by other users before being published to everyone.

I can understand why MFP will not do that, they use the number of records in their data base as an advertising feature, not the accuracy of the data just the volume.

Even commercial foods can be way off on their calorie totals. They are allowed a 20% error rate in published totals and no one from any government agency ever checks so a food manufacturer will understate the actual amounts by 20% and when caught due to random manufacturing variations can simply shrug and walk away from it. Makes for better advertising saying only 80 Calories per serving when there are actually 100, especially for "diet" foods.

User error, unless you are using a scale and ensuring that you are using the correct serving size, you can be way off as well. (Who says your kitchen scale is accurate anyway?)

Remember we are measuring a property of the dead carcass of a once living thing. The normal and natural variations in those properties can and will be significant, all we can do is take averages.

Measuring calorie burn is not an exact science. There are lots of reports out there of a person wearing two different fitness trackers and getting wide ranging variations in calorie burn rates for the same exercise. Look up expected calorie burn rates on line and see how much variation you can have if not using a tracker. One person's moderate exercise rate is another's extreme and a third persons low level.

Your bathroom scales are not all that accurate, they can easily be out several % points. Stand on it and shift your weight around a little, lean forward or lean back and you will see the needle move. You can do this even on a triple beam balance generally considered the best and most accurate scale out there. Placing the scale in a different spot can also impact the readings, your floors are not exactly level you know.

The best you can do is track the numbers and take it as a rough guide. In a week someone with a 2,500 cal per day base would burn 17,500 calories so being out 5,000 based on weight gain or loss is less than a 30% difference. Take all the variables together and I doubt if anyone can measure weight, calories in and calories burned that accurately. To be "out" 1000 calories in a week would be more accuracy than I'd expect to see outside of a lab setting.

It is good to see where the bulk of your calories are coming from. I noticed quickly that I was a boredom eater, I'd eat when bored not when hungry and this was a significant portion of my daily intake. Now any tracking method, pen and paper, or electronic would have shown that but I just happened to notice it on MFP.

Finally it is not an exact relationship between calories burned and weight loss. Water gain/loss can easily overtake the actual weight loss gain in a week. It may take a while for the residue of breaking down the food molecules in your cells for it to be eliminated from your body.

This is a long term effort, as long as you are heading in the right direction, using the MFP data and other data points as a guide, you will come out OK at the other side.

heybales · June 2018

makotron wrote: »

I've been consistently losing an average of 2 lbs per week, since the beginning of March. Lost a total of 40 lbs. Started using MFP and a Polar M600 in May, and so far it's been a very frustrating experience.

The M600 is a mess of its own, but it gives me somewhat accurate step count and calories burned. Which leaves most of the problems to MFP.

According to MFP, in order to lose 2 lbs I need to cut 7350 calories, which is approx. 7,35 calories per gram of fat. When I convert my weekly weight loss, it's always off by a LOT of calories. I created a random food just to add/remove calories according to my weight loss in a week. I've been doing this for almost two months. One week I had to add 5,000 calories, the next one I had to remove 1,000 calories. It's totally random, which means I can't trust the food database at all.

There's also the polar calorie adjustment, which I thought would solve these problems, but it only made it worse. If it's supposed to represent the difference between MFP and Polar daily calories, then why would you add exercise data from Polar to MFPs daily projection? Doesn't make sense and gives me weird numbers. Also, for some reason it randomly miss Polar training sessions. So one day I get 1600 calories burned in exercise, the next day I get only 280. But my workout has been THE SAME since day one - 3 to 4 hours of walk with the dog, daily.

Maybe I'm doing something wrong, but the 40 lbs I lost tell me otherwise. Is there a way to obtain more accurate numbers in MFP?

With the sync - MFP is now correcting itself to the Polar supplied figures.
Workout calories burned, and total daily calories burned.

So it's not a question of obtaining more accurate info in MFP - it's getting more accurate info in Polar first.
Oh, MFP uses traditional 3500 cal fat per lb, are you doing metric conversion?

How do you know the Polar is giving accurate calories burned just because the step count is accurate?
What are your workouts besides walking?

Frankly, if walking is the main or only exercise - step-based calorie burn would be more accurate than HR-based.

HR-based formula for calorie burn (it's a formula that estimates, it's not actually measured) is least accurate and most inflated at the edges of the range for aerobic exercise.
And walking is at the bottom edge of that range.

So if your Polar is slipping into HR-based calorie burn when it would be better to use step-based like for rest of the day - inflated calorie burn.

But even your daily stuff - steps can be accurate - but is the distance it leads to accurate?
Because calories there based on distance, not steps only.

You ever walked a known distance at your average daily pace (not grocery store shuffle, not exercise level pace) and confirmed your stride length setting is correct?

That can start inflated all your daily activity moving time.

With losing 2 lbs weekly - you've likely done the traditional 20% LBM and some muscle mass - which means you burn less than Polar is probably estimating for just resting time.

Unless you were lifting to help maintain muscle - in which case HR-based calorie burn for lifting is inflated too.

So yes - several ways to tweak Polar to be more accurate.

And therefore when MFP correctly itself, for MFP to be more accurate.

Oh - MFP is adding the workout because Polar sent it. Unless you are manually logging it.
And either way, it's removing that from the daily calories burned that Polar sends - because it already knows about it.
You might supply some numbers there and I can show you what is actually happening.
Click the "i" on the Adjustment for more info, and supply the time stamp and daily calories burned, and projection, and the MFP daily burn, and exercise calories mentioned.

Or if you like fun with numbers, read the 2nd section here - the math works the same. (if the syncing is working reliably anyway)
http://community.myfitnesspal.com/en/discussion/10098937/faq-syncing-logging-food-exercise-calorie-adjustments-activity-levels-accuracy/p1

h1udd · June 2018

Its not at all accurate ... everyone is different, it just logs and applys couple of equations to give you a number .... depending on what you do what inputs you give it will depend on how accurate it is for you ....

(1) Calorie info on food is not accurate
(2) Your guestimate to your activity level is not accurate
(3) Your daily calorie burn is not the same every day
(4) how many calories you burn doing exercise is certainly not accurate
(5) Your metabolism vs the average metabolism is probably not accurate
(6) Your scales are not accurate.

This is just for fat loss .... but All of this is masked by the simple fact that your scale measures body weight. ... so your fat, muscle, bones, organs, contents of your stomach and bowels, water retention, glycogen levels

its a great big random guess

But .... it all evens out in the long run .. the trend.

How accurate is MFP supposed to be?

Replies

Categories