smk45 took the words out of my mouth by mentioning scope. Fully understanding the scope of the project is the first most important thing in reducing risk because unfavorable results can cause delays that can be accounted for when setting the schedule but new factors that require skills or resources outside of the predicted numbers greatly increase the risk of failure. However, fully understanding the scope is not always possible and unexpected events are bound to occur. For rate determining steps in the gant chart, I suggest assuming a critical failure assumption for the LF to give a larger buffer time. Being generous with rate determining steps also, by the nature of rate determining steps, gives leeway to all the other non-rate determining steps and steps needed to be done in parallel.
Risk analysis is also something that can change over the course of a project. As such, good communication with all involved parties to better understand changes in time and cost from factors that the PM was not aware of or overlooked.
In the end, the process of time management and scheduling is not something that is done at the start of the project and fixed for the entire duration (although that is ideal) and monitoring and adjusting based on the flow of task completion relative to time is the time consuming and arduous task of the person responsible for scheduling.