Why I built a calculator that gives slower answers.

WeightTimeline is a free weight loss timeline calculator. You enter your current weight, your goal, your height, age, activity level and planned daily intake, and it returns a projected date along with the week your progress is likely to stall. It is a small, single-purpose tool with no account, no upload and no paywall.

The problem I set out to fix

A kilogram of body mass stores roughly 7,700 kcal, or about 3,500 kcal per pound. That figure is real, and it is the foundation of nearly every calculator you will find. Divide your daily deficit into it, count the days, and you have a date.

The trouble is what that arithmetic assumes: that the body doing the burning never changes. It does. Losing 10 kg lowers your basal metabolic rate by around 100 kcal a day from the weight term alone, so the deficit you started with is not the deficit you are running a month later. Sustained dieting also suppresses non-exercise activity — you sit more and fidget less without ever deciding to — and that effect is large enough to erase a modest deficit by itself. A fixed-constant projection cannot see either of these, so it draws a straight line and quietly accumulates error for as long as you keep following it.

That gap is not academic. It is the reason a plan that looked like twelve weeks turns into twenty, and it is a common reason people conclude they have failed when in fact the projection was wrong from the start.

What it does instead

The engine runs a day-by-day simulation across a 365-day horizon rather than performing a single division. Every day gets a fresh Mifflin-St Jeor BMR computed from the previous day's projected weight, so the curve flattens on its own as you get lighter. After 30 consecutive days in a deficit it applies a NEAT penalty against expenditure, modelling the drop in spontaneous movement that comes with sustained dieting.

It then runs the naive fixed-constant projection alongside the adaptive one and draws both. The distance between the two lines is the error the conventional approach accumulates, and the shaded band is the week your progress is projected to stall. Showing them together is deliberate: the comparison is the argument.

The methodology page lists every assumption the engine makes and where each one comes from, and tests the projection against the model behind the NIH Body Weight Planner — including where this calculator comes out too optimistic.

Its limits, stated plainly

This is a model, not a measurement, and I would rather say so plainly than have you discover it later. Mifflin-St Jeor carries a meaningful error margin on any individual. The adaptation figures are my own estimates, informed by the research rather than taken from it, and none of them is fitted to you. Nothing here can account for a medical condition, a medication, thyroid function or body composition.

So treat the shape of the curve and the position of the plateau as the useful output, and the exact date as an estimate. If your projection and your bathroom scale disagree by a couple of weeks, the scale is right.

The calculator also refuses to stay quiet about aggressive inputs. A danger indicator fires when intake falls below 90% of your current BMR, when it drops under the commonly cited floors of roughly 1,200 or 1,500 kcal depending on profile, or when the projected rate of loss exceeds 1% of bodyweight per week. You can still model those scenarios — it just will not hand you a clean timeline that omits what they cost.

How it is built

The whole simulation is one side-effect-free TypeScript function that runs in your browser. A full 365-day projection completes in under a millisecond, which is why the chart redraws while you are still dragging a slider, and it is also why nothing about your body ever needs to reach a server. There is no backend to send it to.

That is a privacy position as much as an engineering one. Weight, height and age are intimate figures, and the safest way to handle them is to never receive them. The privacy policy sets out exactly what that means in practice.

What this is not

WeightTimeline is an information tool, not a medical service, and nothing on it is medical advice. It is not a diagnosis, a treatment plan or a substitute for professional care. If you are pregnant or breastfeeding, managing an eating disorder, living with diabetes or another metabolic condition, taking medication that affects weight or appetite, or planning a large deficit, that is a conversation to have with a healthcare provider rather than with a web page. The terms and conditions cover this in full.

The research behind it

The two load-bearing references are the Mifflin-St Jeor equation (1990), which supplies the resting metabolic rate, and the NIH Body Weight Planner, whose dynamic-model approach established that adaptation belongs in a projection rather than as a footnote to one. Each guide lists the research it draws on, and the methodology page has the full list for the engine.

Who built this

Bibek Sharma

Software Engineer · B.Tech, Computer Science

I'm a software engineer, and I built WeightTimeline because the standard calculators ignore how your energy expenditure changes as you lose weight. I wrote the simulation engine and the guides on this site. The core equations come from the published research above; where the research gives no single figure, the methodology page says which numbers are my own.

I'm not a clinician or a dietitian, and no clinician has reviewed these pages. WeightTimeline is a planning tool, not medical advice.

You can find me on LinkedIn.

Get in touch

If a projection looks wrong, a number reads oddly, or something is inaccessible with a keyboard or screen reader, I want to know. The contact page has the details, and corrections are genuinely welcome — a calculator that argues for accuracy should be held to it.