A fast route is not automatically a useful plan. If a small forecast or performance error sends the answer to the other side of the course, the important result is that split—not the last decimal of the baseline ETA.
Compute the route normally, then open Route stability in the result panel. This analysis is available for routes computed in the current tab because it reuses the forecast package already on the device.
Run the matrix
Press Sensitivity. The course, departure, constraints and forecast stay fixed while it tries nearby wind-speed and boat-speed changes.
The original route is the baseline. Each other cell reports its elapsed-time change and can be selected to overlay that track on the chart. The stability readout also reports how far the alternative track diverges from the baseline.
These are quick checks. They ask whether the route's shape moves, not for full routes in their own right.
Read the pattern
If every track stays in the same lane, the strategic answer is stable even when the ETA changes. That is stronger evidence than one precise finish time.
If the tracks split around a shore, current boundary, or weather feature, inspect the models and the animated field at the split. The route has identified a decision the crew needs to preserve.
If only boat-performance changes move the route, verify the polar and the percentage the crew can actually hold. If only wind changes move it, the forecast timing or strength is doing the work.
What this is not
The matrix applies simple scale changes to one forecast. It is not an ensemble and does not assign probabilities. An experimental GEFS band may also appear on enabled deployments; that measures ensemble wind spread along the deterministic route, but still does not compute an ensemble route.
Grouping alternative routes into a few distinct choices is a possible future feature. It is currently in research only.