Use a moderate number of sample paths in static ensemble path displays

Increase the number of visible sample paths until a single path overlap has less influence on location judgments, but stop before the display looks exhaustive. For example, displays with 17–33 visible paths produced less overlap bias than a 9-path display, while a denser 65-path display performed worse than 17–33 because viewers were more likely to read it as showing all possible paths.

A sparse ensemble path display makes each visible line look more meaningful than it should. Adding more sample paths makes the display read more like a distribution and less like a set of individually decisive routes, but pushing density too high can make the display seem exhaustive.

Mechanism: Moderate path counts shift attention from any one visible line toward the overall distribution of paths, which reduces the tendency to overestimate risk when a location happens to sit on a displayed line.

Evidence: In static hurricane ensemble maps, the difference between judgments for points on a displayed line versus equally distant points off a displayed line was smaller with 17 and 33 paths than with 9 paths. A 65-path display still reduced the bias relative to 9 paths, but performed worse than 17 and 33 paths, and more viewers treated it as showing all possible paths (Padilla et al., 2020).

Notes: The best-performing count in these stimuli was about 30 paths, but the paper notes that the practical maximum can vary with spread, line thickness, and other display parameters.

• User Goal: Judge how much risk or damage a specific location faces.
• Task: Read uncertainty from a static path ensemble rather than from an animation.
• Data: Geospatial path uncertainty shown as multiple sampled lines.
• Chart Setting: A single-image forecast display where viewers can see individual paths.
• Audience: Novice or non-expert viewers with little prior instruction.
• Success Criterion: Smaller judgment differences between equal-distance locations that differ only by touching a displayed line.

Break it when: The added path count makes the display look like the full set of possible outcomes or makes gaps in the distribution newly salient. Why: Very dense displays can restore misinterpretation by making viewers think the map shows all possible paths.

Sacrifice: Individual paths become less separable as the display gets denser.
Risk: If density is pushed too far, viewers may infer that the display is exhaustive rather than sampled.
Mitigation: Increase path count from sparse to moderate, then stop short of a display that reads as the full set of outcomes.

• Mistake: Leaving only a sparse handful of visible sample paths. Why it fails: Viewers give too much weight to whether one displayed line touches the location.
• Mistake: Continuing to add paths until the map looks full. Why it fails: Viewers become more likely to think the display shows all possible paths.

Failure Sign: A location on a displayed line is judged much riskier than an equally distant location off the line.
Quick Check: Create matched probe views where the target location stays at the same distance from the bundle center and only line overlap changes.
Stronger Test: Compare the on-line minus off-line judgment gap across candidate path counts and keep the count with the smaller gap.

• Increase a very sparse path display to a moderate count of visible sample paths.
• Re-test the display with matched on-line and off-line probe locations after the change.
• If the display now looks exhaustive, remove paths until viewers no longer read it as showing all possible outcomes.