Page 1 of 1

How are risk probabilities, impacts, and scores calculated?

Posted: Sat May 23, 2009 1:13 pm
by Sornix
How are risk probabilities, impacts, and scores calculated in RiskyProject 3.1?

Posted: Sat May 23, 2009 1:15 pm
by Intaver Support
Risk probabilities, impacts, and scores are calculated the following way:

• RiskyProject tracks the impact of each risk on different project parameters for each iteration of a Monte Carlo simulation. This is done using absolute units. For example, in the first iteration Risk A caused a delay of 2 days.

• RiskyProject also calculates duration, cost, success rate, finish time, and work on each iteration.

• RiskyProject calculates the Spearman Rank Order correlation coefficient between the duration impact for each risk and the project duration. This coefficient is presented in Sensitivity view.

• This correlation coefficient is not a risk impact yet. It only shows how the risk affects the project. If a correlation coefficient between a schedule risk and the project duration is low, it means that some other factors are impacting the project duration rather than this risk. To become an impact, the correlation coefficient must be normalized. For example, you have a schedule, which has only one risk with an outcome increase duration by 2 days. Because there is only one risk, the correlation coefficient will be 100%. If you change the risk outcome to 100 days, the correlation coefficient still will be 100% even though the measurable impact is much higher. Risk impact takes into account the absolute value of duration increase, cost increase, etc. by applying a normalization coefficient to the impact calculation. If a risk causes the duration of the project to increase 100%, the normalization coefficient is 1. Likewise, if the project duration is increased by 50%, the normalization coefficient will be .5. The risk impact equals the correlation coefficient multiplied by the normalization coefficient.

• Risk probability is calculated based on the chance of risk for each assignment. If a risk has only one assignment, calculated probability equals the chance. If the risk as multiple assignments which are not mutually exclusive alternatives (not correlated), the probability is calculated by combining risk results of each iteration.

• Risk score equals probability multiplied by impact.

• For risks affecting All Parameters (combination of duration, cost, safety, quality, etc.) overall impact is calculated using the impact for each risk category, multiplied by relative importance for each risk category. The relative importance can be seen on the Risk Categories and Outcomes dialog box(Risk menu). Note: If you have not modified risk category weights, then this will have not affect on this calculation. For schedule related risk categories, the risk impact on finish time and success rate will not be used to calculate risk impact.

Posted: Fri Jun 19, 2009 11:15 am
by Intaver Support
Here is an important note regarding the difference between chance of occurence and probability in RiskyProject:

Chance of occurence:
Chance that the selected risk, assigned to the task or resource or particular risk alternative would occur. In this document, chance of risk occurrence usually refers to input chance for risk assignment. Risk probability refers to the calculated probabilities for this risk, which can be different.

Risk probability:
Calculated chance that event would occur. Risk Probability is presented in Risk Matrix, Risk Register, and other views and dialog boxes, where risks information can be viewed. In this document, chance of risk occurrence usually refers to input chance for risk assignment. Risk probability refers to the calculated probabilities for this risk, which can be different.

probability clarification

Posted: Tue May 25, 2010 11:05 am
by Daniel001
I have noticed that if I enter x chance that a risk will occur that the probability on the risk register is not the same. This also occurs if I have only one risk and one activity. It is close, but not the same.

Input chance and calculated proability different

Posted: Tue Jul 06, 2010 11:57 am
by Intaver Support
This is a phenomona that is associated with the Monte Carlo calculation and the manner in which overall risk probability is calculated. The input chance is deterministic, yet the calculated probability is probablistic and takes into account the number of risk categories and risk weighting associated with each.