How Ember works
Simulations and Monte Carlo confidence
Your dashboard headline is a single, deterministic projection. The Simulations page (Ember plan and above) asks a different question: across thousands of possible market histories, how often does that plan actually survive?
Deterministic projection vs. Monte Carlo
The dashboard’s FIRE date and journey chart use a single fixed set of assumptions — your chosen expected real return, inflation, and withdrawal rate — and project one path forward. It answers “given these assumptions, when am I independent, and does the pot hold up?”
Monte Carlo simulation instead runs your plan thousands of times, each time sampling a different sequence of historical market returns, to see how often it survives when real-world variation — good decades, bad decades, currency swings — is allowed to happen in a different order or mix. Both use the same plan and the same engine; the difference is whether the future is treated as one fixed path or many possible ones.
Where confidence shows up
A confidence percentage (the probability your plan survives if you retire at your target age) appears in a few places: as a badge on the Simulations page, and as the resilience axis on the Your Ember radar on the dashboard. Resilience is the one radar axis driven by simulation rather than a single deterministic figure — when a confidence run is available it feeds resilience directly; otherwise the axis falls back to a simpler proxy. See the dashboard guide for the other five axes.
The overdraw & survival panel
The dashboard’s journey line holds your pot on a sustainable draw — the amount a safe-withdrawal-rate calculation says it can pay indefinitely. But your actual spending target might be higher than that sustainable figure in some years, especially early in retirement before a pension kicks in. The overdraw & survival panel simulates drawing your full spending target every year instead, using the same expected-return basis as the journey chart, and shows which age range overdraws the sustainable rate, how far the pot dips at its lowest point, and whether it still lasts through to life expectancy. It’s deterministic, not Monte Carlo — a single expected-return path, not a probability — but it answers a question the sustainable-draw figure alone can’t: does the actual plan you intend to live on hold up.
Real terms, throughout
Every figure on the Simulations page, the confidence badge, and the overdraw panel is in real terms — inflation-adjusted to today’s purchasing power — matching the rest of Ember. If you’re cross-checking against a tool that reports nominal (uninflated) figures, expect its numbers to look larger for an identical plan; that’s the inflation difference, not a disagreement about the plan itself.
What a confidence percentage means — and doesn’t
A confidence figure of, say, “most runs survive” means that share of simulated return sequences — drawn from historical market history via block resampling — left the pot at or above zero through your full horizon. It is not a guarantee, and it isn’t a prediction of which specific future you’ll get: it’s a measure of how sensitive your plan is to the order and shape of returns you might actually experience, given how markets have behaved historically. A change to your assumptions, spending target, or retirement age will move it, and the underlying historical series is versioned data, not something restated in this guide — the exact methodology and any backtest/cohort detail is visible in-app on the Simulations page itself, alongside its own disclaimer that none of this is financial advice.