Fixed fractional is the default professional method and the one most retail rules imitate. You pick a fraction f - commonly 0.25% to 1% - and each trade risks f times current equity.
It has two useful properties. Position size compounds automatically as equity rises, and it de-levers on the way down: after a 20% drawdown each trade risks 20% fewer dollars, which stretches the losing streak you can survive. Mathematically you can never be fully wiped out by stopped-out trades alone, only ground down, which is why risk-of-ruin under fixed fractional is really risk of falling below a threshold.
The costs are real. Recovery is slower than under flat sizing because you are smallest right when the good trades arrive, and the method assumes your stop actually holds. One gap-risk event that blows through the stop breaks the arithmetic entirely.
Related: fixed-lot-sizing, fixed-ratio-sizing, compounding-position-size