Trendweight stats - how accurate is it?

elaineamj

I've just started tracking my weight 2 weeks ago. Upon reading a lot of advice on here, I started using trendweight. Since I am very much in the beginning, I am assuming that I have lost so rapidly in the beginning primarily due to water weight and that this pace will slow down in the upcoming weeks.

I have also been logging everything into MFP. Doing my best to measure but still guesstimating a lot(I ordered a scale and should get it in another couple of weeks). To counteract this, I have been estimating as aggressively as possible.

I studied my stats in trendweight and it says:



It says I am burning 346 cals/day beyond my plan (losing 1lb/week). How much should I trust this? Or is this something I should trust more after it collects 4 weeks of stats?

I am 36yr f, 148lbs (right now) and 5ft tall. I work at a desk job and do about 15 mins of strength/flexibility training plus 15-30 mins of low intensity cardio daily. MFP gives me 1200 cals a day and I try to stick with that although sometimes I go over. Because I don't have a food scale and still struggling with accuracy in calorie tracking, I have not been trying to eat back exercise calories. Overall I feel satisfied with my food intake and rarely get hungry. If I get truly hungry, I do usually eat something like a fruit/chickpeas/hardboiled egg. I have no interest in starving myself and want this to be a long-term thing.

This is what my net calories have looked like over the past 2 weeks:


A little worried I am doing this too drastically, even though I feel satisfied and am not hungry. That said, my thoughts are that I have given trendweight too little time and it can't really predict accurately without at least 4 weeks of data. Thoughts? Should I increase my calorie intake (an extra 346 cals/day is a lot!) or stick to my current strategy?

elaineamj

I should add that I don't really trust my exercise calories in MFP though - I think they're pretty exaggerated. If you look at my total calories - I usually eat at least 1200cals a day. Been undereating a little this week to save calories for an expected big dinner on Saturday:

nordlead2005

Trendweight works by using a weighted average over the past 20 days, so anything over 20 days prior doesn't get calculated in. The more consistent you are over those 20 days the more accurate trendweight will be.

So, a week of maintenance will throw off trendweight, but after 2 weeks of losing weight it should be relatively accurate.

Since it looks like you just started losing weight it may be slightly over stating your calorie deficit, but honestly you probably do need to eat a tiny bit more if you want to only lose 1lb/week. I'd keep doing what you are doing for another week and then evaluate as that will push that initial drop to the tail end of the 20 day window.

WalkingAlong

Water weight loss will throw it off since that is not related to calorie deficit. I would take those stats with a grain of salt, especially in the first month. If you're feeling good at your current levels, I would continue. Your weight loss will slow down. Enjoy it while it lasts.

Sued0nim

Give it more time

But over time using the 4 week and 3 month graphs it gives a good indication

The numbers are averages they're a good guideline

But it's all just estimates so as long as that is in your head and your weight is going in the right direction it's a great tool

elaineamj

nordlead2005 wrote: »

Trendweight works by using a weighted average over the past 20 days, so anything over 20 days prior doesn't get calculated in. The more consistent you are over those 20 days the more accurate trendweight will be.

So, a week of maintenance will throw off trendweight, but after 2 weeks of losing weight it should be relatively accurate.

Since it looks like you just started losing weight it may be slightly over stating your calorie deficit, but honestly you probably do need to eat a tiny bit more if you want to only lose 1lb/week. I'd keep doing what you are doing for another week and then evaluate as that will push that initial drop to the tail end of the 20 day window.

Very good to know about the 20 day window. I couldn't figure that out from what I was reading. Your advice sounds really good.

I've started reading "The Hacker's Diet" that the trendweight calculator is based on. Many things are starting to make more sense. I don't grasp all the math, but I understand more about certain little things. I like that over time, I'll get better feedback (based on my own body) on what my BMR is and can adjust my caloric intake accordingly. I want to be in this for life and I don't want to restrict my diet too severely and yo-yo later. I'm very short with a small frame so every calorie makes a pretty big difference. Based on the charts, it is likely that my maintenance will only be 1500 or so calories a day.

Plus, in the spring I will be in training for more intense physical activities and I don't want to deprive my body of much needed fuel.

If, after 4 weeks, I get the same stats I have now, does this mean I can be eating an extra 346 cals/day extra and still lose 1lb a week? Does this mean that using trendweight gives me a much more accurate calculation of the extra calories I earn through exercise?

Statistics:
You are losing 1.7 lbs/week of total weight.
You are burning 846 cal/day more than you eat.

You have been tracking your weight for 2 week.
You have 20.7 lbs to lose to reach your goal weight.
You will reach your goal around April 16, 2016.
You are burning 346 cal/day beyond your plan

PAV8888

Hmm: Trendweight's programmer states, somewhere: The current weight trend uses an exponential moving average that is roughly equivalent to a 20 day weighted moving average.

Your logic about how to use Trendweight is correct, except that you should keep in mind that even Trendweight is an estimate, albeit a better than average one.


Our body composition is not really known to us, so by convention we take each lb lost/gained to be 3500 Cal, as if our trending weight change is 100% caused by changes to our fat deposits.

The reality is that all of us are losing both fat and lean mass.

While in some cases the ratio is as bad as 1:1, for most of us, especially while obese, we obtain much more favourable, but ultimately unknown*, ratios.

*based on DXA scans a 9+:1 ratio of fat to lean mass lost at 38.8% body fat became a 2:1 ratio between 27% and 25% body fat. N=1. YMMV.

robertw486

elaineamj wrote: »

If, after 4 weeks, I get the same stats I have now, does this mean I can be eating an extra 346 cals/day extra and still lose 1lb a week? Does this mean that using trendweight gives me a much more accurate calculation of the extra calories I earn through exercise?

Statistics:
You are losing 1.7 lbs/week of total weight.
You are burning 846 cal/day more than you eat.

You have been tracking your weight for 2 week.
You have 20.7 lbs to lose to reach your goal weight.
You will reach your goal around April 16, 2016.
You are burning 346 cal/day beyond your plan

Almost all calculators such as the one here at MFP, TDEE calculators, etc use estimates, as that is the best they can do. Trendweight and other such apps use actual trends. So really they provide data that is more useful for you vs data based on averages.

Do keep in mind though, that your water weight fluctuations, as well as hormone levels, can and do impact weights. Diet composition can also affect the weight, as it somewhat changes food absorbtion efficiency and other small factors become involved. As an example, a high fiber diet often impacts to less efficient absorbtion as compared to low fiber.

But if your trends continue, it is a good gauge. Just keep in mind that it will rarely be something you can pin down to exacts, and small changes here and there will happen. If you use the weekly calorie intake on the app, you could log it and compare how close you are to expected weight loss goals. Since it can give the net calorie intake week by week, it gives you a solid comparison to what Trendweight is showing.


And since nobody mentioned it that I noticed, DO NOT freak out if you run into a few days here and there where your trend goes up for reasons you can't explain. At times it happens, even when you remain in deficit. Water retention due to exercise, hormones, sodium, etc, as well as food intake timing and such can all impact your weight. My weight can change 5-6 pounds over a day. If it does make you nervous when it happens, just weigh yourself multiple times in one day to see with your own eyes how quickly it can change.

elaineamj

This is all really great stuff - really helps me understand how I can use trend weight as a tool to help me figure out the optimal amounts I should be eating and whether I am in the zone I want to be in. I also like that it will help me not focus on the daily fluctuations. Also I sure wouldn't mind having a few extra calories to work with every day Just not going to do it until I get more data. Will stick to my current plans for another 2-3 weeks and see where I am at then.

MoreBean13

These are great, please keep them for later. me and my people are moving to vegqas

elaineamj

An update - it's been 2 more weeks and now trendweight's average should no longer include my initial water weight loss.



So instead of burning 346 calories beyond my plan, it now says I am 241 cals/day beyond my plan (of 1lb loss/week). I typically burn 400-500 calories a day in my workouts (according to MFP) and have only been eating half of that. Looks like I should just eat all of it to be in the range of my planned loss.

Plus, my coworker commented a few days ago that I looked tired and pale (ha - I say it was just not enough makeup LOL!). Then again, DH said I look sluggish and that I am more scattered than normal. He does place very close attention to my body signals and usually knows how I am feeling before I do, so I do trust him. I have been feeling sleepier in the mornings for a while now. At first I wondered if it was not enough sleep - but now I am reassessing since I have improved things on the sleep front. I do feel fine and energetic after noon (I'm a night own by nature).

So - the numbers are staring me in the face and now I'm scared of eating SO MUCH MORE. 200-250 calories extra a day is a LOT. MFP sets me at 1200 cals a day (only 5ft tall) and this past week or so with eating half my exercise calories, I've been averaging close to 1400 calories a day. The thought of 1600-1700 calories a day is rather terrifying.

macgurlnet

elaineamj wrote: »

An update - it's been 2 more weeks and now trendweight's average should no longer include my initial water weight loss.



So instead of burning 346 calories beyond my plan, it now says I am 241 cals/day beyond my plan (of 1lb loss/week). I typically burn 400-500 calories a day in my workouts (according to MFP) and have only been eating half of that. Looks like I should just eat all of it to be in the range of my planned loss.

Plus, my coworker commented a few days ago that I looked tired and pale (ha - I say it was just not enough makeup LOL!). Then again, DH said I look sluggish and that I am more scattered than normal. He does place very close attention to my body signals and usually knows how I am feeling before I do, so I do trust him.

So - the numbers are staring me in the face and now I'm scared of eating SO MUCH MORE. 200-250 calories extra a day is a LOT. MFP sets me at 1200 cals a day (only 5ft tall) and this past week or so with eating half my exercise calories, I've been averaging close to 1400 calories a day. The thought of 1600-1700 calories a day is rather terrifying.

For what it's worth, I'm same height as you and weigh ~113. If I get in enough activity (8k steps minimum), I'll burn 1800 calories that day, which in turn means I could eat ~1600 and be at a pretty reasonable deficit.

Eat the calories. Nom nom nom.

~Lyssa

stealthq

elaineamj wrote: »

An update - it's been 2 more weeks and now trendweight's average should no longer include my initial water weight loss.



So instead of burning 346 calories beyond my plan, it now says I am 241 cals/day beyond my plan (of 1lb loss/week). I typically burn 400-500 calories a day in my workouts (according to MFP) and have only been eating half of that. Looks like I should just eat all of it to be in the range of my planned loss.

Plus, my coworker commented a few days ago that I looked tired and pale (ha - I say it was just not enough makeup LOL!). Then again, DH said I look sluggish and that I am more scattered than normal. He does place very close attention to my body signals and usually knows how I am feeling before I do, so I do trust him. I have been feeling sleepier in the mornings for a while now. At first I wondered if it was not enough sleep - but now I am reassessing since I have improved things on the sleep front. I do feel fine and energetic after noon (I'm a night own by nature).

So - the numbers are staring me in the face and now I'm scared of eating SO MUCH MORE. 200-250 calories extra a day is a LOT. MFP sets me at 1200 cals a day (only 5ft tall) and this past week or so with eating half my exercise calories, I've been averaging close to 1400 calories a day. The thought of 1600-1700 calories a day is rather terrifying.

Good Lord, 200-250 cals per day is nothing. It's 2 tablespoons of butter or olive oil OR 2 oz of cheese OR a serving of a good ice cream OR a 2 oz piece of baguette. It's slightly more than one serving of rice.

The only way 200-250 cals is a lot is if you're eating only very low cal foods like steamed vegetables. 200 cals of broccoli is definitely a lot.

So, either add something in that's nice and calorie dense, or swap out some of your current choices for slightly more calorie dense options.

elaineamj

LOL! When you've been eating on such a low cal diet, 250 calories is my entire breakfast and many times, my entire lunch! It sounds like SO MUCH FOOD!!

nordlead2005

If you don't want to just jump up another 250 calories, then add 100 calories to your diet (or eat back 100 more calories from your exercise, whatever mentally works for you). You'll probably gain some water/waste weight, especially if your 100 calories are carbs. This is normal, but it'll cause trendweight to possibly swing too far the other way for a short period of time. Do that for 2 weeks and see if you are still losing more than intended. Then increase again by another 100 calories until you reach the rate you want.

blues4miles

Thanks for posting. I like to see people's numbers/graphs. I do my own graphs manually in excel. I was using a 5-day average but it does cause the thing to be all over the place. I just did 20-day average and it looks a lot smoother like yours does, and makes it a bit more obvious what I'm really losing. Feel like I learned something today.

elaineamj

Well, I'm going for it. I had a tiny dinner planned for tonight (super delicious chicken and wild rice soup for a measly 250 cals). Since I am going out with a friend for lunch , I am going to go hog wild and order a fried shrimp dish. It's a local Vietnamese place so I'm going to go with the closest thing from PF Chang's menu and overestimating. It will end up a whopping 1000 calories lol. And I'll STILL be under my calorie goal for the day by 73 cals! (mostly because I am also upping my workout today - which I had already planned to do a few days ago).

I considered lower cal choices - but eating in a restaurant means a lot of oil and ingredients I can't control anyway and my best projections will save me a measly 200 cals or so. I pre-logged, deleted, and pre-logged again to try out different scenarios.

@bluesformiles Glad this post helped! The 20 day average makes sense - although someone did point out to me that because it's an average, it takes about 1-2 weeks before trendweight catches up to any changes to diet/exercise. So I won't know for at about 2 weeks if this calorie increase is working. Patience is the name of the game!!

WalkingAlong

Nice work, elaine! The new level will feel tough again in a month, I bet. Enjoy it feeling indulgent! And enjoy your Chang-like lunch.

blues4miles

elaineamj wrote: »

@bluesformiles Glad this post helped! The 20 day average makes sense - although someone did point out to me that because it's an average, it takes about 1-2 weeks before trendweight catches up to any changes to diet/exercise. So I won't know for at about 2 weeks if this calorie increase is working. Patience is the name of the game!!

Yeah it keeps me motivated to have more data. I usually don't enter in my info over the weekend (I weigh myself and keep a food log, I just don't copy the info anywhere). So it's exciting on Monday morning when I have 3 morning weights to report in. I only really started to get on point with weighing every day Jan 4, so my trend line doesn't start until Jan 23rd. But it's still cool to see the line forming in such an obviously downward way.

elaineamj

I love my data. I feel so much more confident knowing my math is reasonably accurate.