PMP Primer Masterclass Online Training
Shares
Shares

Applying Monte Carlo Analysis – Part 3

Distribution types

The distribution type is an expression of the distribution pattern and may include the following:

Normal or bell curve distribution

normal_distribution_curve Monte Carlo Analysis

This is an asymmetrical distribution with values in the middle been most likely to occur.  It describes many distributions, such as those around well understood and controlled processes or tasks:

Values in the middle near the mean are more likely to occur than in the triangular distribution

This is best used when there is moderate confidence in three-point estimates

Log normal distribution

log_normal_distribution_curve Monte Carlo Analysis

In this distribution, the values are positively skewed, so the characteristics of this curve are:

Values are positively skewed and not symmetrical like a normal distribution

This distribution is used to represent values that do not go below zero per to have a great positive potential

Uniform distribution

uniform_distribution Monte Carlo Analysis

Here, all values have an equal chance of occurring, and the following describes such a distribution:

Useful for estimates that do not appear to show any central tendency

This distribution shows an equally likely chance of occurring somewhere within a particular range

There is no “most likely” value

Triangular distribution

triangular_distribution Monte Carlo Analysis

For this type, the user sets minimum, most likely and maximum estimates:

This distribution is used extensively in risk models

It allows skewed estimates to be modeled

This distribution can be used when there is a low to moderate confidence in three-point estimates, where the risk is higher or unknown, and where global three-point estimates are applied

PERT distribution

pert_distribution Monte Carlo Analysis

This is similar to triangular, but both the minimum and maximum are more likely to occur, so the characteristics of this distribution are:

Values near the mean are more likely to occur than in the triangular or normal distributions

The extremes are not as emphasized

This can be used where there is a high confidence in the three-point estimates.

Discreet distribution

Here, specific values and likely goods are specified.  For example, “there is a 20% chance of occurrence and there will be an effect in five days”

When undertaking a risk analysis, it is important to understand these distribution types and the effect they will have on results.

Go To Part 4 HERE!