Pocket PMOProject forecasting that answers the only two questions that matter.
Probabilistic project forecasting with Pocket PMO: PERT three-point estimates, Monte Carlo delivery probability and the critical path that decides the date, with the maths shown.
Your Gantt chart is not lying to you. It is selling you one scenario.
A Gantt chart draws the plan you hope will happen. A single-point estimate says “fourteen weeks” with the same confidence whether it is based on a precise scope or a guess. Plans built on one optimistic scenario ignore the way work actually unfolds, and then the date slips, the budget grows, and the conversation becomes about trust instead of data. Tasks completed, hours logged, milestones reached, burndown charts: all of it describes what already happened. “Will we hit the date?” “What will it really cost?” “What happens if this slips?” None of those questions are historical. Uncertainty in, probability out: ranges per task, an ensemble of outcomes, and a confidence level attached to every date you name.
Size → Verify → Track
Break the work into tasks. For each one, give three numbers: best case, likely case, worst case. That is ROM and PERT: risk enters the estimate from the first minute instead of arriving later as an unexplained contingency line. Monte Carlo simulation samples every task from its own range, thousands of times, and builds the distribution of whole-project outcomes. The critical path (CPM) shows which chain of tasks actually decides the finish date. As work progresses, the probability of the committed date moves, up or down. Tracking probability, cost, risk and utilisation keeps the forecast alive rather than archived in a kick-off deck.
What comes back is a sentence you can say out loud.
Name any date and get the share of simulated outcomes that finish on or before it. “There is a 74% chance of finishing by 12 June” survives a steering committee. “We should be fine” does not. PERT and ROM estimators convert three-point inputs into a range you can fund, with a visible reason for every unit of contingency and a P80 number if the commitment needs one. CPM highlights exactly which tasks decide the finish. Everything else has float. When reality moves, you know what to move with it. EVM and risk scoring roll up per project and across the portfolio, so sign-off conversations are about interventions, not about whether the numbers are real.
P50, P80, P90: what the percentile on a forecast actually means
The percentile is the share of simulated outcomes that finish on or before that date. P50 is the median outcome; a P80 date means roughly one in five simulated futures still finishes after it. Higher percentile, more contingency, lower odds of overrun. Practitioners and government guidance use exactly this vocabulary: P50/P80 confidence levels are standard in cost and schedule risk work. We define ours transparently: confidence describes your inputs and ranges, not a guarantee of the future.
A layer, not a replacement
PERT & ROM estimating, Monte Carlo simulation, CPM critical path, EVM, risk scoring, portfolio oversight, Software/Marketing/Construction templates, and an MCP server for AI agents.
Run the example on your own schedule
Run the sizer free in Pocket PMO: 10,000-iteration Monte Carlo, PERT and critical path on your numbers, no signup.