/ Notes

Notes on verifying backtests

Each note takes one thing a backtest report doesn’t show you and works it through with exact arithmetic.

Best of forty is not a result

How often noise clears the bar, the best figure noise hands you, and where the bar moves after a sweep of forty variants.

The one-bar repaint: why a higher-timeframe value changes after the fact

A daily value pulled onto an intraday chart reads a candle that hasn’t closed on the live bar and only finished ones on history. Two fixes that don’t work, and the one that does.

Six numbers a strategy report hides, and how to recover them

Profit factor, win rate and trade count give you the payoff, the breakeven, an honest interval, the expectancy and the cost that erases it.

RSI(14) is two different indicators, and the label doesn’t tell you which

Wilder’s smoothing versus an EMA: one constant, double the lag, and a different set of threshold crosses on the same chart.

Historical simulations are not live results. Nothing on this site is investment advice.