How forecasting works
Traditional project tools give you one date. Chronomap gives you a range. Here's how that works.
The problem with single-point estimates
You ask a developer: "how long will this take?" They say "about a week." Your project plan now says 5 days. But what does that really mean?
Maybe it means "if nothing goes wrong." Maybe it means "the median case." Maybe it means "I'm being optimistic so you don't worry."
The result: plans built on single-point estimates are wrong by default. Not because anyone lied — but because a single number can't represent something uncertain.
Three numbers instead of one
Chronomap asks for three estimates per task:
These three numbers define a shape — a probability distribution of how long the task might take. The shape tells you not just "when" but "how confident."
The shape is skewed right — the pessimistic case stretches further than the optimistic one. This is normal. Bad surprises are bigger than good ones.
What happens with dependencies
A single uncertain task is easy to reason about. But projects have chains: task B can't start until task A finishes.
If A takes 3-12 days and B takes 2-8 days, when does B finish? Not "5 + 5 = 10 days." The uncertainty compounds.
Brightness represents likelihood — the brighter the region, the more probable. A has a tight bright core near day 3-5 that fades toward day 12. B inherits all of A's uncertainty plus its own, so its bright region is broader and its tail stretches further.
Chronomap propagates these distributions through your entire dependency graph automatically. You enter estimates; it computes the forecast.
Now here's the payoff: if A actually finishes on Day 4, its uncertainty vanishes. It's no longer a distribution - it's a fact. And that fact tightens B's forecast automatically:
Compare B's bar to the previous diagram. Before, it stretched from day 5 all the way to day 20 — carrying both A's uncertainty and its own. Now that A is a known quantity, B only carries its own uncertainty: day 6 to day 12. The forecast tightened by itself.
You didn't re-estimate anything. You didn't drag a bar. A completed, and the downstream forecast updated automatically. This is what happens across the entire project as work progresses — the forecast starts wide and converges toward certainty.
People and contention
Dependencies aren't the only thing that affects timing. People do too.
If one person is assigned to two independent tasks, they can't work on both full-time. Chronomap models this: each task gets a share of that person's time, stretching its calendar duration accordingly.
With two people, both tasks finish in ~3 days. With one person splitting time between them, both stretch to ~6 days. Adding tasks to someone's plate pushes out everything they're working on.
This is why adding more tasks to someone's plate doesn't just delay new work — it pushes out everything they're already doing.
Decisions and branches
Not everything is a straight line. Sometimes work branches based on a decision that hasn't been made yet.
Only one path will happen — but both lead to integration testing. The question is: when does testing start?
A simple Gantt chart doesn't represent this neatly, especially with many different decisions. Chronomap handles it naturally:
- Each branch gets a probability weight. Two equally likely options? 50/50. Three options? A third each.
- The forecast blends them. "Integration testing" doesn't wait for both paths — it waits for whichever one happens. So its forecast is the weighted average of the two possible start dates, not the worst case.
- People aren't double-booked. If the same person would do the work on either path, their time isn't split — because only one path will actually run.
The blended forecast is bimodal — it has two humps, one from each possible path. The shape honestly represents "it might be 2 weeks or it might be 8 weeks, we don't know yet." Once the decision is made, the forecast snaps to the chosen path and the other hump disappears.
Why this matters
Honest commitments
"We'll ship between April 10 and April 28" is more useful to a stakeholder than "we might ship by April 15."
No manual re-planning
Change an estimate, add a task, reassign someone — the forecast updates instantly. No dragging bars around on a gantt chart.
Clarity increases with time
As tasks complete and decisions are made, the forecast tightens automatically. Early in the project the range is wide. Near the end, it converges.
Add certainty
For any deliverable, see clear recommendations about how to reduce its uncertainty and accelerate its delivery - with additional resources, down-scoping, or decision making.
In summary
- You estimate each task with three numbers: optimistic, likely, pessimistic, and define which depend on which.
- Chronomap builds a probability distribution for each task's duration.
- It propagates those distributions through your dependency graph.
- It accounts for people splitting time across competing work.
- It handles unmade decisions by weighting each possible path.
- The output is a forecast with confidence levels (P10/P50/P90) — not a single date.
- Chronomap analytics help you understand how accurate your estimate durations are in retrospect.