Knowing what a project is really likely to cost is important from every perspective: personal, professional, and organizational. But developing a realistic and achievable cost estimate has two components: first you develop an accurate baseline estimate, then you need to develop a realistic contingency. Most people do step one, very few people even think of step two.
This article focuses on the importance of adequate contingencies and reserves in delivering a successful project: https://mosaicprojects.com.au/Mag_Articles/AA006_Contingencies_are_not_a_soft_option.pdf
@Pat, if you subscribe to the work of Glenn Butts (NASA), https://www.slideshare.net/NASAPMC/glennbutts-mega-projects-estimates; Bent Flybjerg, Oxford https://oxford.academia.edu/BentFlyvbjerg and Ed Merrow, IPA https://www.ipaglobal.com/team/edward-merrow/ et al,
You will see there are two parts to calculation of CONTINGENCIES. They are RANGE or REFERENCE CLASS ESTIMATES PLUS w need to stop using BELL CURVE distributions and start to use “fat” or “long-tail” distributions (log-normal)
For more explanation go HERE https://build-project-management-competency.com/1-4-1-10-unit-10/ and scroll down to Figure 12 and you can see this illustration coming from NASA and the US DoE.