What is Three-Point Estimating (PERT)?
Three-point estimating replaces a single guess with three: optimistic (O), most likely (M), and pessimistic (P), then blends them. The triangular average weights them equally; the PERT (beta) formula weights the most likely ×4, which tames extreme values.
Its real value is psychological as much as mathematical — asking "what if everything goes wrong?" forces risks into the open that a single-number estimate hides. Exam questions usually just want the beta formula applied cleanly.
Formula
PERT: E = (O + 4M + P) / 6 · Triangular: E = (O + M + P) / 3
Worked example
A developer sizes a payment-gateway integration: best case 4 days, most likely 8, worst case 24 (if the vendor sandbox misbehaves, as vendor sandboxes do). PERT: (4 + 32 + 24) / 6 = 10 days. The single-point "8" would have quietly ignored the sandbox risk everyone knew about.