Does anyone have a formula to project weight loss?
MegaMooseEsq
Posts: 3,118 Member
MFP appeals to me because I'm a numbers nut, but unfortunately, I'm only a talented novice when it comes to spreadsheets. I've been trying to create a formula to predict my weight loss going forward, and I thought I had it about right, but when I applied it to my last seven months of tracking it was a little *too* accurate, which makes me think I'm goofing up somewhere. Can any numbers nuts help me out?
Here's what I came up with: (-[yesterday's weight]*13+[today's net calories])/3500+[yesterday's weight]
I'm estimating TDEE using the formula found here: http://www.bodyrecomposition.com/fat-loss/how-to-estimate-maintenance-caloric-intake.html/. I've used the low estimate for women to get 13 kcal/lb.
ETA: Oof, apparently I came across way more woo than intended - I suppose it is Friday. I am definitely not trying to predict what the scale will say tomorrow - I've been losing for several months now and I know how this works. However, because I set different calorie goals for different days of the week, I have a hard time with online projectors such as the one MFP uses when you close out your diary. I'm trying to reverse-engineer that formula for goal-setting purposes and to save myself having to type in my stats over and over. Thanks!
Here's what I came up with: (-[yesterday's weight]*13+[today's net calories])/3500+[yesterday's weight]
I'm estimating TDEE using the formula found here: http://www.bodyrecomposition.com/fat-loss/how-to-estimate-maintenance-caloric-intake.html/. I've used the low estimate for women to get 13 kcal/lb.
ETA: Oof, apparently I came across way more woo than intended - I suppose it is Friday. I am definitely not trying to predict what the scale will say tomorrow - I've been losing for several months now and I know how this works. However, because I set different calorie goals for different days of the week, I have a hard time with online projectors such as the one MFP uses when you close out your diary. I'm trying to reverse-engineer that formula for goal-setting purposes and to save myself having to type in my stats over and over. Thanks!
0
Replies
-
How can it be too accurate? I don't understand...if it's accurate then it's working, right???1
-
It's pretty much impossible to calculate.3
-
Weight loss looks like uneven steps on a graph, not a smooth downward line. The body lets go of weight, but not in an orderly and predictable fashion.2
-
I know that there are lots of variables and it won't be at all perfect - I'm just looking for estimated trend lines to help set my weekly and monthly goals. Like MFP uses when you close out your food diary. Plus I figure if I can figure out the "ideal" formula, I can use it to backtrack and see what my actual numbers were. Then I've got all sorts of data to play with.0
-
-
ISweat4This wrote: »
Yes, this is exactly what I'm trying to reverse engineer. It's got to be possible, right?
1 -
I was hoping it would download to an excel file with formulas but no luck, but it is possible.
1 -
tcunbeliever wrote: »How can it be too accurate? I don't understand...if it's accurate then it's working, right???
My concern is that I would not expect a predicted weight loss to track all that closely with an actual weight loss. The formula is taking one (mostly) objective number - the reading on my scale - estimating my TDEE off of a rough formula, then using my estimated monthly calories (in this case, the calories I logged over the first half of the year, and we all know how precise logging can be), and the estimated calories it takes to burn off a pound, and predicting my weight a month (or day) later. I would not expect those numbers to line up all that closely, since that would seem to say that all of those estimates were pretty much spot on. Or maybe that is just how it works when you look at trends and I'm just overthinking this?0 -
Why not use a trending app like Libra or Happy Scale?4
-
I have a spreadsheet where I log my weight every week, calculate the average pounds lost per week (starting from when I first started keeping track a few months ago), then use that average to predict my weight for each week for the rest of the year. Would that help?1
-
MegaMooseEsq wrote: »tcunbeliever wrote: »How can it be too accurate? I don't understand...if it's accurate then it's working, right???
My concern is that I would not expect a predicted weight loss to track all that closely with an actual weight loss. The formula is taking one (mostly) objective number - the reading on my scale - estimating my TDEE off of a rough formula, then using my estimated monthly calories (in this case, the calories I logged over the first half of the year, and we all know how precise logging can be), and the estimated calories it takes to burn off a pound, and predicting my weight a month (or day) later. I would not expect those numbers to line up all that closely, since that would seem to say that all of those estimates were pretty much spot on. Or maybe that is just how it works when you look at trends and I'm just overthinking this?
I still don't know what you are going for here. It sounds like you want to predict your weight loss specifically because you know that the prediction will not align with reality.0 -
If I Lost One Pound per Week
By Valentine's Day, I'd be 7 pounds lighter.
I'm loving it!
By St. Patrick's Day, I'd be 11 pounds lighter.
With the luck of the Irish, I'll make my goal!
By Easter, I'd be 16 pounds lighter.
I'd be a cute bunny!
By Memorial Day, I'd be 23 pounds lighter!
Won't let it rain on my parade!
By Flag Day, I'd be 25 pounds lighter.
Boy, could I wave my flag, then!
By Independence Day, I'd be 27 pounds lighter.
That's a great independence from over eating!
By Labor Day, I'd be 36 pounds lighter.
What a wonderful reward for my hard work!
By Columbus Day, I'd be 41 pounds lighter.
What a joy to discover what I can do!
By Halloween, I'd be 44 pounds lighter.
I didn't think I had a ghost of a chance!
By Thanksgiving, I'd be 47 pounds lighter.
I have so much to be thankful for!
By Christmas Eve, I'd be 51 pounds lighter.
Talk about being merry!
By New Year's Eve, I'd be 52 pounds lighter.
It's not just a new year, it's a new ME!2 -
I don't have links at have links at hand, but over time there have been pretty sophisticated spreadsheets developed to estimate TDEE from your MFP data, with links in forum posts. That'd be a better start - way better than the TDEE calculators if you vary much from the norm.
Maybe someone else has the links, or you can find them by forum search. I'm sorry I don't have them - they're on a device that's bricked right now.
Of course, using something like this depends on accurate logging of food, exercise and scale weight over quite a period of time. But if you're a data geek, you probably have that.
Libra gave me decent projections most of the time when I was in a consistent deficit. Now, in maintenance, I get more of a "rolling hills" effect. You can see a sample snip in the ticker in my MFP profile page - at the left, a small piece of the sort of smooth descent to goal, then the mini-hills start.
0 -
I don't have links at have links at hand, but over time there have been pretty sophisticated spreadsheets developed to estimate TDEE from your MFP data, with links in forum posts. That'd be a better start - way better than the TDEE calculators if you vary much from the norm.
Maybe someone else has the links, or you can find them by forum search. I'm sorry I don't have them - they're on a device that's bricked right now.
Of course, using something like this depends on accurate logging of food, exercise and scale weight over quite a period of time. But if you're a data geek, you probably have that.
Libra gave me decent projections most of the time when I was in a consistent deficit. Now, in maintenance, I get more of a "rolling hills" effect. You can see a sample snip in the ticker in my MFP profile page - at the left, a small piece of the sort of smooth descent to goal, then the mini-hills start.
That's exactly the kind of thing I'm looking for! I tried to search the forum archives earlier but got distracted - I'll keep looking, or maybe someone will pop up who remembers a link.MegaMooseEsq wrote: »tcunbeliever wrote: »How can it be too accurate? I don't understand...if it's accurate then it's working, right???
My concern is that I would not expect a predicted weight loss to track all that closely with an actual weight loss. The formula is taking one (mostly) objective number - the reading on my scale - estimating my TDEE off of a rough formula, then using my estimated monthly calories (in this case, the calories I logged over the first half of the year, and we all know how precise logging can be), and the estimated calories it takes to burn off a pound, and predicting my weight a month (or day) later. I would not expect those numbers to line up all that closely, since that would seem to say that all of those estimates were pretty much spot on. Or maybe that is just how it works when you look at trends and I'm just overthinking this?
I still don't know what you are going for here. It sounds like you want to predict your weight loss specifically because you know that the prediction will not align with reality.
Well, sort of - what I'm trying to do is identify, based on my past data, where (if anywhere) I'm going wrong with my logging based on how my numbers deviate from a predictable formula.Why not use a trending app like Libra or Happy Scale?
Mostly I want to figure out how to do it to see if I can, but I'm also hoping to use the formula to play with my data outside of the restrictions of an app.mebelfanti wrote: »I have a spreadsheet where I log my weight every week, calculate the average pounds lost per week (starting from when I first started keeping track a few months ago), then use that average to predict my weight for each week for the rest of the year. Would that help?
That would probably work just fine - weirdly I didn't think of that since I was so preoccupied with trying to work out the standard formula. I'd have to tweak it to figure out the effect of future calorie changes, but I think I know how to do that. Thanks!0 -
Try this:
http://community.myfitnesspal.com/en/discussion/813720
I have no idea whether the spreadsheet is compatible with current MFP, but it still exists and may be a source for ideas, at least.1 -
Try this:
http://community.myfitnesspal.com/en/discussion/813720
I have no idea whether the spreadsheet is compatible with current MFP, but it still exists and may be a source for ideas, at least.
Oooooh, thank you! So many ideas!0 -
Or this, with the same caveats:
http://www.myfitnesspal.com/blog/EvgeniZyntx/view/mfp-data-export-tool-the-overview-6599271 -
How good are you at calculous? The formula is a bit daunting. As your daily weight changes via CICO, this is a double variable equation. One is weight, the other is calorie deficit per day based on a recalculation of daily weight changes. You're maintenance calories are the start, and you need to use the Harris-benedict equation as part of your equation.2
-
I use an excel spreadsheet with a column of weigh-in dates (each Friday), a column of actual weight on that day, and a column of estimated weight (expectation of 1 lb loss a week). I use those columns to create a line graph in the spreadsheet with a trending line. The expected weight line vs. the trending line just shows me how I'm trending compared to how I had estimated. The actual weight line just gives me a nice visual of my progress. I like this because I can also see exactly where TOM has affected weight fluctuations. Would something like that work for you? This won't really predict future weight though.1
-
Show the calculus. Please. Pretty please.2
-
VeronicaA76 wrote: »How good are you at calculous? The formula is a bit daunting. As your daily weight changes via CICO, this is a double variable equation. One is weight, the other is calorie deficit per day based on a recalculation of daily weight changes. You're maintenance calories are the start, and you need to use the Harris-benedict equation as part of your equation.
My calculus is "I only made it to Trig I because a teacher told me I was bad at math and that was 16 years ago", but I would love to see the formula if you've got it.1 -
If I understood correctly what you're trying to say, you're starting from the fallacy that there exists an equation that is more accurate than your own loging.
In other words you're trying to figure out if you have errors in your logging by figuring out your diviation from an impartial equation.
But such an equation that is specific to you does not exist.
You can calculate your deviation from a standard formula... But that doesn't mean that you're logging incorrectly.
There is no guarantee that you match the "standard".
What you can do however is be CONSISTENT in how you log
And if you're consistent in how you log, and possibly use a trending weight app or other way to estimate your weight level, you may be able to calculate how far off the predicted equations you are located.
And of course you would still have to account for your potential body composition changes.
If I do understand what you're trying to do, you may be better off thinking of it as a budget problem.
You need to save 35,000 calories to lose 10lbs. Your savings each day depend on how many calories you ate and how many calories you spent. Ideally you shouldnot have a savings of more than 20% of calories spent in a day.... now over to someone who can translate this into something useful
3 -
I am no mathematician, and I don't really understand what you want to do so ignore this if I am misunderstanding. But you can find the line of best fit to predict where your weight would be on a future day. Of course, it will only be accurate to a point since you won't get down to zero weight. This would give you the linear line of your weight loss, although since weight loss isn't linear it won't be 100 percent accurate. But I think scale apps probably do the same thing, so I don't know that you'd want to do the math.
Is that what you want to do? If so I can post the formula. It is just the slope formula and slope-intercept equation of a line.0 -
So, it's not a formula per se, but I think that you may be over complicating things.
If you simply want to predict your loss over time, you can look at your historical data and figure out what your current average rate of loss is over the past say, 6 weeks(42 days): (SW-CW)/42 = Rate of Loss (RoL)
For future predictions:
RoL x 35 = Future Loss (FL) over 5 weeks (which is that the MPF app tells you)
CW-FL = Predicted weight in 5 weeks.
Of course, this only works for mid term predictions, because as you lose more weight the average RoL will slow down. But, if all you're looking for is an easy way to see if you're generally staying on track, this will probably work. You can run the simulation weekly or so, which will update your Future Weight as your other variables change.0
This discussion has been closed.
Categories
- All Categories
- 1.4M Health, Wellness and Goals
- 393.7K Introduce Yourself
- 43.8K Getting Started
- 260.3K Health and Weight Loss
- 176K Food and Nutrition
- 47.5K Recipes
- 232.6K Fitness and Exercise
- 431 Sleep, Mindfulness and Overall Wellness
- 6.5K Goal: Maintaining Weight
- 8.6K Goal: Gaining Weight and Body Building
- 153K Motivation and Support
- 8.1K Challenges
- 1.3K Debate Club
- 96.4K Chit-Chat
- 2.5K Fun and Games
- 3.8K MyFitnessPal Information
- 23 News and Announcements
- 1.2K Feature Suggestions and Ideas
- 2.6K MyFitnessPal Tech Support Questions