Distribution

Parameters

How to use

1

Beta

Alpha and beta parameters

Used for wide range of condition. If Alpha and Beta are 1, Beta distribution will uniform. If Alpha and Beta are the same, the distribution will be symmetrical.

2

BetaPert

Mean parameter

Derived from Beta distribution. Used as “smoother” substitute to Triangular distribution

3

Cauchy

Shape parameter

Used if low duration, cost, or start time represents “best case scenario” value. High value represents a cut off: how long duration, cost, and start time can expend.

4

Custom

Set of points with probabilities

Used to define generic shape of distribution by entering point as a percentage of task duration, start time or cost, and associated relative probability.

5

Discrete

Set of intervals with probabilities

Used to define distribution consisted in multiple segments with associated relative probabilities. Each segment is defined by uniform distribution. Interval is entered as a percent of task duration.

6

Exponential

Mean parameter

Similar to Cauchy distribution

7

Gamma

Alpha and beta parameters

Similar to Cauchy distribution

8

Gumbel

Two share parameters

Skewed distribution. Useful if there is lower chance that task will take longer time then shorter time.

9

Laplace

One parameters

Similar to Cauchy distribution

10

Lognormal

Mean and standard deviation parameter or P1/P99, P5/P95, P10/P90

Similar to Gumbel distribution

11

Normal

Mean parameter or  P1/P99, P5/P95, P10/P90

Simetrial distribution. Low and high values of duration, cost, and start time serve as cut offs.

12

Reyleigh

Standard Deviation parameter

Similar to Gumbel distribution

13

Triangular

Most Likely parameter

Most likely parameter overwrites base value of duration, cost, or start time.

14

Uniform

Equal probability that duration, cost, or start time will be between low and high values.