What's the best way to know what your body weight trend is? I mean say you weigh yourself every day. How do you know from a week's entries what your weight change per month is?

Skip to content
# What's the best way to know what your body weight trend is? I mean say you weigh yourself every day.

###

What's the best way to know what your body weight trend is? I mean say you weigh yourself every day. How do you know from a week's entries what your weight change per month is?

I take the median of each week and use that to track trends because it levels out daily weight fluctuations. Only good piece of advice I've gotten from Doucette outside of him recommending Joseph's lavash bread for pizzas and wraps

What, so the median of one week compared to the median of the following week?

Yeah. The median of each week is what I use to track the long term progress I'm making

Shifts in water weight can make that an issue. If you have low water 4 days out of the week then that shows a huge problem outside of typical weight loss

Why median and not average? And by this method you need at least 2 weeks before you get a value. That's a bit too long I think, when you change your calorie intake one week should be enough to see the trend, and make adjustments if necessary.

>best way to know your body weight trend

weigh yourself at regular intervals, under the same conditions, for example:

>interval: every Monday

>conditions: after you've taken a shit, and wearing only underwear

>how do you know what your weight change per month is?

Let's say you start weighing yourself on the first Monday of each week in April. To know how much weight you lost that month, subtract your weight on the first Monday of May from your weight on the first Monday of April.

e.g. If I weigh 80kg on the first Monday of April, and weigh 76kg on the first Monday of May, then I've lost 4kg in April.

The point is not to see your weight change after the fact, then you already know your weight change, I said you want to see the trend. To see the monthly change after one week of entries as I said, not after one month of entries.

what the frick is the point of knowing the 'trend'

I experimented with two methods right now on my computer.

1) A scatter plot with a trend line and trend line equation. The "k" value, i.e. the number before the x, is the slope of the trend line, which is the rate of change. Take that number and multiply by 30 and you should get the weight change per month.

2) Put weight entries in cells in a spreadsheet. Next column, each cell contains the formula for the cell to the left minus the cell above that, so it shows the change for one day from the previous day. Next column, cells contain the formula for averaging 7 cells in the column to the left of that, and multiplying by 30. Now you should have the projected 30 day weight change from a week's entries.

The weird thing is that the k-value in 1) and the average change in 2) were not the same number, don't know why, they should be at least close, if not identical.

for me it's daily weigh-ins after taking a shower then using the average for the week.

As I said then you need at least two weeks before you can see what your rate of change is.

ok? and?

I weight myself everyday after waking up, not drinking water, drinking some coffee to shit my guts and before showering, and of course in underwear. If you are using a goy smartscale, drinking water, not shitting before, eating and taking a shower will frick up and vary your bioimpedance results as frick. As the other anons say, do it always under the same conditions, but i add the factor of not doing anything that can add random numbers due to biology (like a shower, your body will absorb a random amount of water thru your skin, thus fricking up your bodyweight and bioimpedance analysis)

>How do you know from a week's entries what your weight change per month is?

Last day minus first day weight, I guess.

No because there are fluctuations. These two methods are attempts to deal with fluctuations:

Something is wrong, because after one week of entries it shows a trend of -7.2 kg/month even though the trend line is clearly almost totally horizontal.

3 day moving average.

Why would you do anything else?

So after 4 days of entries you take the average of 1, 2, 3 compared to the average of 2, 3, 4 times 30? Or after 6 days you take 1, 2, 3 and 4, 5, 6 times 10?

wake up, take a piss, weigh yourself. Do that whenever.

why after waking up? Because changes happen as you sleep, including weight lost. You exhale the mass.

why piss? to reduce the variable.

why whenever? because I only care about my current weight and my goal weight. Trends? Monthly change? Who cares, man. The only math that matters is goal-current.

>What's the best way to know what your body weight trend is

take garage spring of known k constant, measure displacement hanging from it, backcalculate mass, plot on 1 long piece of paper scrolled around the meridian of your room

>all these autistic calculations

homie literally just get an app that has charts and moving averages

find out you weight's function

take its derivatives

How to do this unironically? Mathlet here

Quadratic regression so you have the curve, then get derivative using online calculator (I failed calculus)

Linear regression is cleaner, might represent the real life progress better and you don't need to derivate to project yourself into the future.

>calculus

3500kcal = 1lb

current average weekly deficit = 250

target weight = current weight-20lbs

you only need algebra to solve for how long this average, if sustained, will get you to your target

what I do is show how many months it is to my goal by deficits of lesser and greater severity, and then show the immediate correction I would need to achieve those deficits. Easy to calculate that if I took a walk I'd be back on track, or that if I fasted I could get onto the more severe track.

seismograph data

Mirror

In the morning after you relieve yourself and before breakfast.

graph it

Weight daily at the same time and take a rolling average from the last two weeks.

i have an excel sheet with a polynomial trend line

Weight yourself naked first thing when you wake up, before drinking any water. This doesn't take into account shit, of course. In the case that you shit at some point, don't drink water after shitting and weight yourself several hours later. Recor the minimum weight each day

Alternatively, you could just weight yourself naked multiples times each day and record the minimum value

Train a NN to predict your weight based on past data. I suggest quadratic error as loss function and lookback window as hyperparameter.

An ARIMA model would probably work better in practice, assuming we are only tracking weight. NN would allow capturing non linear relationships with other factors, but we'd need to actually have data on them.

Haha doing math to lose weight

Kys

>weight every morning

>put data in excel

>plot data

>add line of best fit/3-7 day averages

90% of replies:

>weigh yourself when you wake up

Do you all have fricking down syndrome here?

yes, because all you do to "Calculate your weight trend" is plot data points, by weighing yourself. Are you a high school dropout or just functionally moronic?

fricking brainlet

Slope = (x2-x1)/(y2-y1)

I used myfitnesspal to get an estimate of how many calories I should eat. Did that for a week. It said I should gain 2 kg/month, but I seem to be close to maintenance, so I'm adding 500 calories to what myfitnesspal says I should eat, starting from today.

nobody fricking cares about your blog post

I use the minimum value. Yeah, sometimes I go up 2kg, others I go down 1kg, but if the minimum values are decreasing over time, then that means I'm losing weight

I’ve noticed my training day weight also differs quite substantially from my off day weight.

Right after a workout I’m depleted and weigh and easy 2 lbs less. Weirdly I also weigh less on my off day and usually weigh the most on a training day after I’ve eaten, even more than on my off days.

Weigh yourself every day then set up a trendline

>3 day moving average

>6 day moving average

>9 day moving average

>12 day moving average

>observe slopes

Listen kid, this shit isn’t scalping 0DTEs on the spooz 40 minutes before FOMC using a 5 minute chart. There is no volatility, if you can’t observe a trend with the aforementioned moving averages you are not following a trend aka not cutting/bulking