A single number pulled from instinct is how projects end up under-quoted. Give each task a best case, a likely case and a worst case, and the spread tells you what to actually charge โ plus the clauses that keep scope from drifting.
Be honest in the worst column โ that's where the value is. A task you're sure about (6 / 8 / 10) barely moves the quote. One you're guessing at (4 / 10 / 40) should, and this is the only way it can.
The expected value is a 50/50 bet โ quote it and you lose money on half your projects by definition. Each step up buys you the odds that the project fits inside the number. For fixed-price work, 90% is the usual landing spot.
Expected hours use the standard PERT weighting โ (best + 4รlikely + worst) รท 6 โ so the likely case dominates while the worst case still pulls the number up. The ยฑ figure is that task's standard deviation; the big ones are where your quote is actually at risk.
Scope creep is consistently named the single biggest problem in freelance work, and it almost never arrives as a big obvious request โ it arrives as a series of small ones, each too minor to argue about. The defence is having written down, before the work starts, what "included" means and what happens when something isn't. This is a starting point, not legal advice; have a lawyer look at anything you rely on.