Task success rates are impacted by several different factors.
The first is task deadlines. In this case the success rate is a function of how often a task does not exceed a task deadline. In the Task Information > Deadlines tab you can set deadlines for tasks. Tasks deadlines can have 3 outcomes if the task exceeds the deadline: no impact, end task, and cancel task. Of these the Cancel task outcome will impact the success rate. So if the task exceeds the deadline 30% of the time, it will have a success rate of 70%.
The other success rates measure how often a task is actually executed during a simulation. One through risk outcomes and the other as an output of branching.
Task success rates can be affected is through risk outcomes. When a risk is assigned to a task two of the outcomes can be “cancel task” and “cancel task and all successors”. The simulation will calculate how often a task is cancelled due to the risk and this will impact the success rate.
Finally, if you are using conditional and probabilistic branching, the success rate measures how often specific branches are run which is a function of the conditional or probabilistic criteria used. It is possible that you could have all three of the above impacting the success rates of tasks.
In the branching example below, we can see how success rates (or how often the activities were executed) during the simulation. In addition, Activity 6 – Phase 3 has a deadline (indicated by the red arrow) and as Phase 2b is executed more often and therefore pushing it past the deadline probabilistically, it has a lower success rate. Compare this to activities 12 and 18 which do not have deadlines.