Course contentsShow
← Software Engineering
Lesson 1652 of 1,93348. Project Management for EngineersFree lesson

Three-Point Estimation

Apply optimistic, pessimistic, and most-likely estimates to calculate weighted averages and understand risk.

Three-Point Estimation

You'll learn how to account for uncertainty by combining optimistic, pessimistic, and most-likely estimates into a single weighted forecast.

What Is Three-Point Estimation?

Three-Point Estimation acknowledges that the future is uncertain. Instead of giving one number (which is often wrong), you provide three scenarios:

  • Optimistic (O): Best-case scenario—everything goes smoothly
  • Most Likely (M): Realistic middle ground—typical conditions
  • Pessimistic (P): Worst-case scenario—complications arise

These three values combine using a weighted formula to produce a more accurate estimate that accounts for risk.

The Calculation

The most common formula is the PERT (Program Evaluation and Review Technique) weighted average:

Expected Estimate = (O + 4M + P) / 6

Notice that the most-likely estimate gets four times the weight—it's usually closer to reality than the extremes.

You can also calculate standard deviation to understand the range of uncertainty:

Standard Deviation = (P - O) / 6

An Example

Imagine estimating a feature to refactor authentication:

  • Optimistic: 3 days (everything's straightforward)
  • Most Likely: 6 days (some edge cases emerge)
  • Pessimistic: 15 days (legacy code fights back)

Expected Estimate = (3 + 4×6 + 15) / 6 = (3 + 24 + 15) / 6 = 7 days

This gives you a more grounded number than just hoping for the best-case 3 days or panicking about the 15-day scenario.

Why It Matters

Three-Point Estimation forces you to think through risks explicitly. The wider the gap between optimistic and pessimistic, the more uncertainty exists—signaling you might need to spike, research, or break down the work further.

Key Takeaway: Three-Point Estimation balances optimism and caution by weighting three scenarios, giving you a statistically grounded forecast that respects uncertainty.