How Our Simulations Work
A risk tool you cannot trust is worse than none. So here is exactly how ours compute their numbers: the random generator, the formulas, and where the data comes from. No fabricated figures, no hidden assumptions — if a number appears on this site, this is how it was produced.
A seeded, reproducible generator
Every Monte-Carlo tool on the site draws its randomness from the same small, well-known generator — mulberry32, a fast 32-bit pseudo-random function — started from a fixed seed. A fixed seed means the simulation is deterministic: the same inputs always produce the same result, run to run, device to device. That is deliberate. It guarantees the numbers are not cherry-picked from a lucky run — you, and we, get the identical output every time, so the result is reproducible and auditable rather than a one-off.
The Monte-Carlo equity model
The risk simulator models an account starting at $10,000 and runs thousands of independent "lives" of your system. In each trade it multiplies equity by a win or a loss, scaled by your risk per trade f and your reward-to-risk R:
on a loss: equity ×= (1 − f)
It runs a fast preview of about 2,000 paths as you type and a full 10,000 on demand, then reports how many ended in ruin, the median outcome, and the share that finished in profit. An account is counted as "ruined" once equity falls to the ruin floor. Because the streak of wins and losses is random but seeded, the headline numbers are stable — change an input and the change you see is real, not noise.
Cross-checking against the formula
Simulation is the honest way to handle messy cases (uneven payouts, a finite number of trades), but it can hide bugs. So where an exact formula exists, we show it next to the simulation. The risk of ruin calculator runs its own seeded Monte-Carlo and displays the classic even-money gambler's-ruin formula beside it — RoR = ((1−p)/p)^U for a fair payout — so you can see the two agree. Checking a result two independent ways is the habit that keeps a number trustworthy.
Real data, honest errors
The market tools do not invent prices. The volatility / ATR, correlation matrix, RSI screener, DCA and converter pull real historical and live prices from public exchange endpoints (Binance candles, CoinGecko rates). If a request fails or is rate-limited, the tool says so and shows nothing — it never fabricates a fallback number. Only public price requests are sent; nothing you enter is stored or uploaded, and figures are displayed in plain Latin digits to match across languages.
What this means for you
Use the outputs as honest floors on risk, not promises. The models assume a fixed win rate, fixed reward-to-risk and independent trades; reality is rougher — edges drift, losses cluster, fees bite — so the real risk is usually a little worse, never better. That is the point: these tools are built to talk you out of trouble, not into it.
