MFP predicted weight loss accuracy...

PaulThomp23
PaulThomp23 Posts: 38 Member
in Getting Started
So, when you complete your diary and it gives you the 'if every day was like today' thing and it calculates your future weight loss based on that days calorie intake.... how accurate is it? Like, can it be relied upon to even be remotely close?

Yes there are days i havent tracked meals due to work, but i know 100% that ive never over eaten my daily allowance to the point it would negatively affect my weight loss and thus far it has been pretty consistent.

But say for example 5 weeks ago it tells me at the end of my day 'in 5 weeks time' following the same intake i would weigh 10lbs less than i currently do and even though I've not done anything that would have been so drastic as to put on 10lbs, how accurate is that prediction and if it is relied upon by others what can i do to get on that same track?

Replies

  • L1zardQueen
    L1zardQueen Posts: 8,753 Member
    It’s a gimmick.

    http://community.myfitnesspal.com/en/discussion/1318741/in-5-weeks-youll-lose-10lbs-why-is-it-not-working/p1
  • wunderkindking
    wunderkindking Posts: 1,615 Member
    edited March 2021
    It isn't designed to really predict. I think it's just meant to show the effect your choices have. No day is EVER (hormones, step numbers, how much you fidget, actual exercise, food, whatever) exactly like another one.
  • PaulThomp23
    PaulThomp23 Posts: 38 Member
    wunderkindking wrote: »
    It isn't designed to really predict. I think it's just meant to show the effect your choices have. No day is EVER (hormones, step numbers, how much you fidget, actual exercise, food, whatever) exactly like another one.

    Perfect. Thank you. Thought i was really doing something wrong
  • callsitlikeiseeit
    callsitlikeiseeit Posts: 8,626 Member
    been here almost a decade.

    Have never EVER seen it be accurate for ANYONE.
  • AmunahSki
    AmunahSki Posts: 211 Member
    I’ve been logging for 68 days, and when I ‘complete’ my diary each day I duly log that prediction on a spreadsheet. I now have 33 days of data to compare my actual weight against what MFP predicted I’d weigh 5 weeks ago.

    In the early weeks I was ahead of the prediction by three or so pounds (as I dropped water and food weight at the beginning), but that early headstart slowed and it now differs by a pound or so either way. The pattern over the last couple of weeks is that I am ‘under’ for three days, then ‘over’ for three days... Yesterday, I was ‘over’ the prediction by 0.9lb, today I’m ‘under’ by 2.6lb. Weight loss is not linear, as we all know too well!

    However, as an average, my daily weight is only above the prediction by 0.7lb, so although I don’t have anywhere near a decade of data I’d argue that it’s actually pretty close. I am of course very interested to see how the pattern changes as I get closer to goal (and into maintenance), hence the reason I’m collecting the data in the first place.
  • ALZ14
    ALZ14 Posts: 202 Member
    Over the summer I actually tracked this!

    I was pretty diligent at eating the same amount of calories each day and getting roughly the same amount exercise.

    After 5 weeks I started checking where I was compared to the prediction. Sometimes it was within a pound, sometimes it was off by 5 or more. Those were usually the days where I had burned a lot more calories than usual.

    It is nearly impossible to make “every day like today” and it is only as accurate as your logging. If you aren’t logging 100 calories here or there then the prediction will be further off.

    I used it as motivation to keep up the work and to be honest with myself about what I was actually eating.
  • AnnPT77
    AnnPT77 Posts: 34,377 Member
    To the comments above, with which I'd agree, I'd add this: The MFP calculations are statistical estimates based on reasonable research. The calorie goal it spits out is essentially the average calorie requirement for people similar to you in age, weight, activity level, etc. But individuals can vary from statistical averages.

    Most people will find that the calorie level MFP calculates for them, used in the way intended, will result in very close to the weight loss rate that they put in their profile (as long as they don't over-reach and get the minimum of 1200 for women, 1500 for men). A few people will find the estimate to be a bit off, high or low. A very few rare people will find the estimate to be surprisingly far off. That's the nature of statistical estimates. (And the same is true of other calorie-needs calculators, too.)

    What we should do is track carefully for the first 4-6 weeks (full menstrual cycle for premenopausal women), then compare results to expectations (being honest about any deviations from plan). That will reveal whether we're one of the many close-to-average people when it comes to calorie needs. If we are, all is well. If we're far off, we may need to adjust calorie goal, going forward. (If someone is non-average, it may not be evident why.)

    The people who turn out to be non-average will find that the "in 5 weeks" prediction is farther off from right for them, too.

  • FitAgainBy55
    FitAgainBy55 Posts: 179 Member
    edited March 2021
    callsitlikeiseeit wrote: »
    been here almost a decade.

    Have never EVER seen it be accurate for ANYONE.

    I don't use MFP for tracking but I'm sure it's based on the same models that I'm currently using. This is my actual weight loss versus a linear 1.5 lb per week weight loss. The blue line is linear projected 1.5 lbs per week. The dots are weigh ins. The redish line is my trend weight. I discussed this in another thread but the only divergence I've seen with the model I'm using is that my fitness tracker underestimates my TDEE so I actually eat more than the projected amount to lose:

    xbo7i1rdhwq5.png

Categories