Before you start
Risk Contingency Calculator
Work out how much contingency your risk register really calls for, at the confidence you need, and which risks that money is for.
Open the risk register: each risk with a probability and what it would cost if it happened, as one figure or a lowest, most likely and highest. It is read inside your browser; nothing is uploaded.
Risk register
Drop a spreadsheet here
One row per risk: a probability (30% or 0.3, or a level like 4 or High) and what it would cost if it happened, as one figure or a lowest, most likely and highest. Status, owner and days of delay if you have them.
The expected value is an average, not a reserve
The usual way to size contingency is to multiply each risk’s probability by its cost and add them up. That figure, the expected monetary value, is what the risks cost on average over many imaginary projects. Yours is one project. Each risk either happens or it does not, and when a large one happens the cost arrives all at once. A reserve set at the expected value runs out in a large share of outcomes, and this tool says exactly how large.
What the simulation does
Ten thousand times, it decides which risks happen, each at its own probability, and draws the cost of each one that does from its range. The spread of totals gives the reserve at any confidence: the amount that is enough in half the outcomes, in four out of five, in nine out of ten. Put in the contingency you already hold, and the page says what share of outcomes it covers.
Which risks the money is for
Each risk is scored by how much lower the 80% reserve would be without it. That is a better guide to where mitigation spending goes than probability times cost, because it counts the risks big enough to break the budget on their own, not just the likely small ones.
Probabilities as words or levels
Registers often score probability on a scale (1 to 5, or low to high) rather than as a percentage. The page reads either. For levels, each is given the chance at the middle of its band (10%, 30%, 50%, 70% and 90% on five levels), shown so you can change them to whatever your organization’s scale means.
What this does not do
It does not tell you which confidence to fund at; that is a decision about how much risk the sponsor will carry. It treats each risk as independent, which understates the spread when risks share a cause. It adds days of delay only as an upper bound, since delays on parallel work do not all move the finish; the Schedule Risk Analyzer puts them on the schedule itself. And its answer is only as good as the register’s estimates.