Three-Point Estimation (PERT)

Basic Principles

Three-Point Estimation, also known as PERT (Program Evaluation and Review Technique) estimation, uses three scenarios to calculate expected costs or durations:

Estimate Type	Symbol	Description	Weight
Optimistic	O	Best-case scenario	1
Most Likely	M	Most probable case	4
Pessimistic	P	Worst-case scenario	1

Core Formulas

Expected Value (E)

Standard Deviation (σ)

Variance (σ²)

Confidence Intervals

Confidence Level	Range	Formula
68%	±1σ	E ± σ
95%	±2σ	E ± 2σ
99.7%	±3σ	E ± 3σ

Application Methods

Traditional PERT

Modified PERT

Triangular Distribution

Industry-Specific Applications

Software Development

Component	Optimistic	Most Likely	Pessimistic	Expected
Design	10 days	15 days	25 days	16 days
Development	20 days	30 days	45 days	31 days
Testing	15 days	20 days	30 days	21 days

Construction Projects

Element	Optimistic	Most Likely	Pessimistic	Expected
Foundation	$50K	$65K	$90K	$67K
Structure	$200K	$250K	$350K	$258K
Finishing	$100K	$125K	$175K	$129K

Risk Analysis Integration

Risk Levels and Multipliers

Risk Level	Multiplier Range	Application
Low	1.0-1.1	Well-understood scope
Medium	1.1-1.3	Typical projects
High	1.3-1.5	Complex/new technology
Very High	1.5-2.0	Unprecedented work

Implementation Process

1. Data Collection
   • Expert judgment
   • Historical data
   • Market analysis
   • Risk assessment
2. Estimation
   • Calculate three points
   • Apply formulas
   • Determine confidence
   • Document assumptions
3. Validation
   • Peer review
   • Sensitivity analysis
   • Reality checks
   • Expert validation

Best Practices

Estimation Guidelines

Phase	Key Considerations	Documentation
Planning	Scope definition	Assumptions log
Execution	Regular updates	Change log
Control	Variance analysis	Performance data
Closure	Lessons learned	Final report

Quality Checks

1. Range Validation
   • P > M > O
   • Reasonable spreads
   • Consistent units
   • Logical relationships
2. Expert Review
   • Technical feasibility
   • Market conditions
   • Risk factors
   • Historical comparison

Software Tools and Integration

Popular Tools

Tool	Primary Use	Key Features
Microsoft Project	Project management	Built-in PERT analysis
Oracle Primavera	Enterprise projects	Advanced risk modeling
@Risk	Risk analysis	Monte Carlo simulation
Crystal Ball	Forecasting	Statistical analysis

Statistical Analysis

Distribution Types

Type	Use Case	Formula Modification
Beta	Standard PERT	(O + 4M + P) / 6
Triangular	Simplified	(O + M + P) / 3
Normal	Large datasets	Custom parameters

Advantages and Limitations

Advantages

1. Simple to understand
2. Considers uncertainty
3. Statistical basis
4. Flexible application

Limitations

1. Subjective inputs
2. Assumes beta distribution
3. Limited complexity handling
4. Dependency issues

Industry Standards

PMI Standard

• Required inputs
• Calculation methods
• Documentation requirements
• Quality criteria

AACE International

Class	Accuracy Range	Application
5	-20% to +40%	Screening
4	-15% to +30%	Feasibility
3	-10% to +20%	Budget
2	-5% to +15%	Control
1	-3% to +10%	Check

Future Trends

1. AI Integration
   • Automated range prediction
   • Pattern recognition
   • Historical data analysis
   • Real-time updates
2. Advanced Analytics
   • Machine learning models
   • Predictive analytics
   • Dynamic updating
   • Cloud integration
3. Industry-Specific Developments
   • Custom algorithms
   • Specialized applications
   • Integrated risk analysis
   • Automated validation

Practical Tips

1. Estimation Process
   • Use expert judgment
   • Consider multiple scenarios
   • Document assumptions
   • Review regularly
2. Quality Control
   • Validate ranges
   • Check calculations
   • Compare to actuals
   • Update estimates
3. Risk Management
   • Identify uncertainties
   • Quantify impacts
   • Monitor variations
   • Update projections

Three-Point Estimation remains a valuable tool for project management and cost estimation, providing a structured approach to handling uncertainty while maintaining simplicity and practicality. Its effectiveness depends on quality inputs, proper application of methods, and regular validation of results.