A plan is a guess.
Give it some daylight.
You know what needs doing. Find out what fits —
and which small change gives your week room to breathe.
Hours aren’t the whole story.
Simulation estimates from your assumptions.
Not predictions or guarantees.
What if you changed one thing?
Same duration draws. One change. A fair comparison within this model.
A week you can actually see.
Likely-duration schedule. Shaded room is time you have left.
Be honest about the unknowns.
Give each task a best, likely and worst duration. Deadlines are end of day.
Best · likely · worstDeadlineOn time
Current → changed
Open the model. No black box.
What these numbers mean
A seeded Monte Carlo simulation samples independent triangular task durations from your three estimates. It schedules one student’s work, pauses at the daily hour limit, and respects release days and dependencies.
The percentage is the fraction of simulated weeks that met every deadline. It is conditional on the inputs and this scheduling policy, not an empirically calibrated prediction. A scope cut assumes proportional duration savings without reducing quality; decide whether that tradeoff is realistic.
What it leaves out
Shared bad days, fatigue, task-switching costs, interruptions and changing priorities are not modeled. A heuristic chooses the most urgent downstream deadline; it does not prove an optimal schedule. “Impossible” warnings below are limited to structural constraints.
Everything stays in this browser. No account, AI inference, calendar access or network API. Export a replay to reproduce every result.