HonestSlim - Monte Carlo Weight Loss Calculator
I started my own weight loss journey and quickly noticed something annoying: every calculator gave me a single straight line. "You'll reach your goal in 14 weeks." It never worked out that way. Good weeks, bad weeks, a cold that knocked me out for 5 days, a holiday that wrecked my deficit.
The straight line was always wrong — not because I was failing, but because the model was too simple.
So I built HonestSlim.
How it works
Instead of one prediction, it runs 5,000 Monte Carlo simulations. Each simulation models a different adherence pattern — some weeks you nail your deficit, some weeks you don't. BMR is recalculated weekly using Mifflin-St Jeor to model metabolic adaptation as you lose weight.
The result is a fan chart showing p10 through p90:
- Best 10% of outcomes (blue line)
- Median — what most people experience (green)
- Worst 10% — slow but still moving (red)
You see a realistic spread instead of false precision.
Tech
React + Recharts, no backend, no database. Runs entirely in the
browser. Hosted on Vercel free tier. Zero running costs.
Where I am
Just launched. No marketing budget. Learning how to get traffic for the first time. Goal is €50–100/month from AdSense eventually.
Would love feedback — especially:
- Does the fan chart make sense to non-technical users?
- Is the algorithm explanation clear enough on the site?
Interesting approach with the Monte Carlo simulations. The fan chart makes the variability clear, but many creators or users might still struggle to translate it into actionable steps for their week-to-week progress.
The real question is: which part of the output do users actually act on, and how can that become a no-brainer insight?
If useful, I can put together a short written guidance plan showing how to highlight actionable insights from the simulations for first users. It would focus on clarity, the core signals, and quick decisions — not a full tutorial or feature redesign.
Thank you so much for your feedback. And you are right, it is hard to translate to actions. But the original intent was to help users develop realistic expectations, so they don't feel like they've failed when progress temporarily slows down or fluctuates.
That makes sense.
Then the core issue is not “what should users do next?” It is whether the simulation helps them understand that slowdown, fluctuation, or uneven progress is normal.
That is a useful positioning angle because it turns the chart from a prediction tool into an expectation-setting tool.
If useful, I can write the tighter guidance layer for this: how to frame the output so users do not just see uncertainty, but understand what it means for their progress.