Avoid exact-point readouts from individual ensemble paths

Avoid using whether a single ensemble member crosses a point as the main readout for exact-location decisions. For example, do not ask viewers to decide which of two nearby locations is at greater risk from an ensemble-path map when one location sits on a path and the other is slightly nearer the center, because the crossed location attracts responses.

A visible path crossing is a salient local feature, so readers can treat one sampled path as if it were a decisive route rather than one draw from a distribution. That makes exact-point judgments shift toward crossed locations even when another nearby location better matches the overall distribution.

Mechanism: Individual path crossings compete with distribution-based reading and bias exact point comparisons toward the crossed point.

Evidence: In point-based damage judgments from ensemble hurricane forecasts, novice viewers almost always chose the nearer location when it was on a path, but they shifted strongly toward a farther location when that farther location was the one crossed by a path; the effect remained when an equal-damage option was added (Padilla et al., 2017).

Notes: The paper suggests ensemble displays are better suited to pattern, area, and gist judgments than to point-based judgments.

• User Goal: Make a location-specific decision about one point versus another.
• Task: Exact point comparison.
• Data: Geospatial uncertainty shown as multiple possible paths.
• Chart Setting: A static ensemble display where individual paths visibly cross some candidate locations.
• Audience: Novice readers.
• Success Criterion: Path overlap does not distort which location is judged more at risk.

Break it when: The task is to judge overall patterns, areas, or spread rather than exact points. Why: The paper argues that ensemble displays are better suited to those non-point tasks.

Sacrifice: You lose an apparently direct point cue from visible path crossings. Risk: If you ignore this rule, readers can treat sampled paths as exhaustive routes or as full-storm tracks. Mitigation: Use the ensemble display for area or pattern interpretation instead of point comparison.

Mistake: Letting viewers infer that a point on one visible path is more likely to be hit than a nearby point off the path. Why it fails: Readers overweight the single visible path even when the off-path point is closer to the center of the distribution.

Failure Sign: A farther point is chosen much more often once it happens to lie on a single path. Quick Check: Create paired-location cases where one farther point lies on a path and one nearer point does not, then compare the choices. Stronger Test: Repeat the paired-location check with an equal-outcome option and see whether path overlap still shifts choices away from the nearer point.

• Remove exact-point comparison tasks from ensemble-path displays.
• Reframe the judgment around areas, spread, or overall pattern instead of a single crossed point.
• If exact-point decisions are required, do not rely on the ensemble-path display alone.