The problem is, neither the "Total Story Points" (nor total story count) or the "Velocity" (or throughput) can be known exactly upfront. There is some level of uncertainty and variability. The team will discover additional work, more details emerge, the delivery team grows (and shrinks), and production issues emerge and bring delivery to a stand-still. By using fixed numbers, we paper over this uncertainty and give a forecast that looks good, but fails to be achievable.

Monte Carlo forecasting is a way to include this uncertainty as input the traditional mathematical formulas. Simply, it is performing mathematics on ranges of values rather than discrete values. By adding this uncertainty, the forecasts given lets YOU control what level of certainty you need to communicate.