Which is better at predicting calories - Fitbit or MFP?

2»

Replies

  • WinoGelato
    WinoGelato Posts: 13,454 Member
    girlinahat wrote: »
    girlinahat wrote: »
    WinoGelato wrote: »

    Well the other thing to keep in mind when you have the two synced is that there is often an end of the day adjustment which can result in a drop in the calories you have left for the day. FitBit prorates your calories, assuming you are going to be as active all day as you are at certain points in time, so if you go to bed early for example, and you aren't burning as many calories for the last 3 hours of the day, then when you wake up, you will often see that there are less calories (sometimes putting you in the negative) on MFP. Once you've been using the two systems for a while you get used to it, but I usually try to leave 50-100 calories available by the time I end my day, since I know many of those will be gone when I wake up.

    Also - I never look at how many calories FitBit says I have used or have left. I log all my food in MFP and let FitBit tell it how many calories I burned and how much "credit" I get in the form of exercise adjustments. When I first started using FitBit I drove myself batty trying to understand why the numbers were always different, then I realized it's different systems, different technology, different algorithms. They work well together, but they aren't identical.

    question - I understand about fitbit prorata-ing my daily calories, but what I can't get my head around yet is how to set up both systems. Do I let MFP calculate a calorie goal? What about fitbit? how do I set it? Do I set a calorie goal there? On fitbit it won't show me the base rate of calories it is using prior to exercise so I can't be certain it is starting from the same point as MFP.

    Do you have a day that you forgot to wear your Fitbit? If you go back to that day on your dashboard it will have your days calories without any extra movement added. Mine is 1138 which is also very close to that calculated by a bodytrax machine at my gym which is reassuring if disappointingly low

    I guess it's more about how to set each device - MFP and fitbit. I want my fitbit to sync with MFP, and I log my food intake on MFP. However I am finding fitbit is adding too many calories. Should I have a calorie goal in fitbit?

    Why do you think FitBit is adding too many calories?
  • macchiatto
    macchiatto Posts: 2,890 Member
    I've had my Charge 2 for about a month now. I wish it were accurate. ;) It's saying I average around 2300 calories/day (ranges from 2100-2700 depending on my activity level) but I've been eating around 1700. Initially I was still slooowly losing weight but now it seems to be stabilizing so 1700 might actually be my TDEE.

    I'm a 40 y/o female, 5'7", 131 lbs and I've been running about 3x/wk and PT exercises 2x/wk, averaging 13k steps/day. I do have PCOS, most likely with insulin resistance, so I may be one of the unlucky ducks for whom PCOS causes a slower metabolism.
  • ChelzFit
    ChelzFit Posts: 292 Member
    I have my FITBIT synced to MFP and I find MFP usually estimates around 50-100 calories more compared to my FITBIT. I usually go off my FITBIT for total calorie burn.
  • MiniMexxxxx
    MiniMexxxxx Posts: 43 Member
    cheldadex wrote: »
    Mfp gives me less than 2100 as my tdee, I know from experience that's cutting calories for me.

    Fitbit tend to hover between 2300 and 2800, so in my case fitbit is more accurate. Ymmv

    Thanks. My Fitbit app and MFP actually ended up about 10 calories apart in the end yesterday so pretty close.

  • MiniMexxxxx
    MiniMexxxxx Posts: 43 Member
    macchiatto wrote: »
    I've had my Charge 2 for about a month now. I wish it were accurate. ;) It's saying I average around 2300 calories/day (ranges from 2100-2700 depending on my activity level) but I've been eating around 1700. Initially I was still slooowly losing weight but now it seems to be stabilizing so 1700 might actually be my TDEE.

    I'm a 40 y/o female, 5'7", 131 lbs and I've been running about 3x/wk and PT exercises 2x/wk, averaging 13k steps/day. I do have PCOS, most likely with insulin resistance, so I may be one of the unlucky ducks for whom PCOS causes a slower metabolism.

    Could it be your calorie requirements have gone down as you've lost weight. Mind you, Fitbit and MFP should make that allowance for you so perhaps not! :smile:
  • macchiatto
    macchiatto Posts: 2,890 Member
    macchiatto wrote: »
    I've had my Charge 2 for about a month now. I wish it were accurate. ;) It's saying I average around 2300 calories/day (ranges from 2100-2700 depending on my activity level) but I've been eating around 1700. Initially I was still slooowly losing weight but now it seems to be stabilizing so 1700 might actually be my TDEE.

    I'm a 40 y/o female, 5'7", 131 lbs and I've been running about 3x/wk and PT exercises 2x/wk, averaging 13k steps/day. I do have PCOS, most likely with insulin resistance, so I may be one of the unlucky ducks for whom PCOS causes a slower metabolism.

    Could it be your calorie requirements have gone down as you've lost weight. Mind you, Fitbit and MFP should make that allowance for you so perhaps not! :smile:

    Yeah, MFP and Fitbit are updated with my current stats. In recent years I only lost 20 lbs and that was over a year ago. (I did have to re-lose 5 of it in Jan/Feb.) I just got the Fitbit Charge 2 a month ago. For my current stats with my activity level I just would have expected a TDEE a little higher, and it seems MFP and Fitbit both would, too. ;)
  • SoDamnHungry
    SoDamnHungry Posts: 6,998 Member
    I use fitness frog I think to determine my TDEE as sedentary. Fit Bit seems to agree, and its added calories burned for steps taken seem pretty accurate as I haven't gained.

    I have the Fit Bit One which I like a little better because I don't get "steps" added from moving my arm or something.
  • heybales
    heybales Posts: 18,842 Member
    MFP where you are guessing from 4 non-exercise daily activity levels, and hoping you add exercise correctly?

    or

    A device that is on you all day, using the same basis as MFP but applied to what it actually sees you doing?

    (and yes to someone's question of is Garmin the same).

    You could totally luck out and MFP is right on correct. Go buy lottery tickets too.
    Fitbit's going to have much better chance of getting close.

    Calorie goals, weight loss goals, daily burn goals - none of those in Fitbit will effect the basic math that is done.
    Fitbit reported calorie burn minus MFP estimated calorie burn.

    The only one that could matter is the Fitbit activity level, for times you do NOT sync your device during the day.
    Fitbit is going to estimate a per minute burn rate if there is no real data from the device to use. After it goes up by 100 calories from last sync, new amount is sent to MFP.
    Fitbit has 2 options - Sedentary which is just barely above BMR, and Personal which is based on historical averages.
    If you device sync to phone often - those don't matter except when sleeping then. And really doesn't matter then either.
    If like me and you don't sync device until close to dinner time as first time, it could matter. Except I don't look at MFP goals during the day anyway, as meals are pre-planned. So the math done doesn't matter to me anyway.

    @MiniMexxxxx - it could also easily be that a poorly logged 1700 cal day is actually closer to 2000 or more if it was accurately done.

    You could also have your stride length for walking set to exercise pace that is done how many hours a day - but then that inflates ALL the other walking you do much slower for how many hours a day?

    You could also have workouts where HR-based calorie burn calculations are incorrect and inflated.

    And you even could have had many diets in the past where you've lost muscle mass, so your BMR is lower than what it uses as basis for everything else. Or the PCOS combined with that makes it lower.
    Even those with un-medicated thyroid issues in studies have shown a pretty tight bell curve of vast majority being no worse than 5% slower.
    It's rather the being tired feeling and not moving as much daily that's the kicker.
  • WinoGelato
    WinoGelato Posts: 13,454 Member
    heybales wrote: »
    MFP where you are guessing from 4 non-exercise daily activity levels, and hoping you add exercise correctly?

    or

    A device that is on you all day, using the same basis as MFP but applied to what it actually sees you doing?

    (and yes to someone's question of is Garmin the same).

    You could totally luck out and MFP is right on correct. Go buy lottery tickets too.
    Fitbit's going to have much better chance of getting close.

    Calorie goals, weight loss goals, daily burn goals - none of those in Fitbit will effect the basic math that is done.
    Fitbit reported calorie burn minus MFP estimated calorie burn.

    The only one that could matter is the Fitbit activity level, for times you do NOT sync your device during the day.
    Fitbit is going to estimate a per minute burn rate if there is no real data from the device to use. After it goes up by 100 calories from last sync, new amount is sent to MFP.
    Fitbit has 2 options - Sedentary which is just barely above BMR, and Personal which is based on historical averages.
    If you device sync to phone often - those don't matter except when sleeping then. And really doesn't matter then either.
    If like me and you don't sync device until close to dinner time as first time, it could matter. Except I don't look at MFP goals during the day anyway, as meals are pre-planned. So the math done doesn't matter to me anyway.

    @MiniMexxxxx - it could also easily be that a poorly logged 1700 cal day is actually closer to 2000 or more if it was accurately done.

    You could also have your stride length for walking set to exercise pace that is done how many hours a day - but then that inflates ALL the other walking you do much slower for how many hours a day?

    You could also have workouts where HR-based calorie burn calculations are incorrect and inflated.

    And you even could have had many diets in the past where you've lost muscle mass, so your BMR is lower than what it uses as basis for everything else. Or the PCOS combined with that makes it lower.
    Even those with un-medicated thyroid issues in studies have shown a pretty tight bell curve of vast majority being no worse than 5% slower.
    It's rather the being tired feeling and not moving as much daily that's the kicker.

    Agree with all of this, but I think it was @macchiatto that suggested that the 1700 from MFP is more in line with their losing pattern, not @MiniMexxxxx who was the OP.

  • macchiatto
    macchiatto Posts: 2,890 Member
    @heybales thanks for the insights. I admit it's a little early for me to jump to the conclusion that 1700 is TDEE since weight loss has only seemed to stabilize for a very short time now. (I lost 1.2 lbs in 3 weeks at that intake level but only 0.1 the past week.) Just keeping an eye on the data. I do weigh my food and log carefully at home but meals out (usually 1-2x/wk) might throw things off a bit.

    I've never had much muscle mass (and did some yoyo dieting in the past in a 20-lb range) so the decreased BMR theory may be a factor.

    Re walking stride length, I followed the instructions in the Fitbit group for calculating running and walking stride length. My stride lengths do seem to vary a good bit by speed, which also varies a good bit throughout the day (or the type of run, "easy run" vs speed work, etc.). I'll try bumping that down a little and see what happens.
  • heybales
    heybales Posts: 18,842 Member
    Many people see the advice (and so I could see the Fitbit advice being such) of setting the walking stride length to their exercise pace - which is only used for say 1 hr daily, or less often.

    The Fitbit does dynamically adjust used stride length based on the figure stride length given and your weight, and it calculates from each step impact what that one stride length must have been, length and time is pace, pace and weight is calorie burn.

    But as you might guess, if the stride length figure is set to the extreme side of the range of fast pace, it's ability to adjust way down to grocery store shuffle is going to be off.

    So it's best to head to a known distance track for a good 1/4 mile, and walk it at normal average daily pace, which is usually going to feel very slow in that setting. Start the workout at the mark, end it at the mark after 1 lap. (confirm the distance marks since many are meter tracks now).
    Now you have what Fitbit thought the distance as, and most important the steps actually done.

    1320 / steps = stride length decimal feet.inches.

    Feet in that setting.
    0.inches x 12 = decimal inches in that setting.

    Running same thing - pick the middle pace you are likely to go between the extremes.

    Of course the more steps you do daily the more important those figures are.
    The average user attempting as goals to get to 10K steps and walking as only exercise might do fine on default values, but it can't hurt to attempt more accuracy.