Scoring model

How Vibe Scores are calculated

VeloraMaps turns a location into a 0 to 100 Vibe Score by combining stable map context, nearby activity, road-noise proxies, amenities, nature, bounded weather context, and light calibration from user feedback. The model is not trying to declare a place objectively good or bad. It estimates how suitable a specific point feels for common intents such as staying, focusing, socialising, or relaxing, then shows the underlying pillar scores and confidence so the result can be questioned.

Last updated: May 6, 2026

Data sources by pillar

Each score starts with the exact coordinate selected on the map, then evaluates nearby signals around that point. Most map-derived data is cached server-side by rounded coordinate so repeat checks are fast and do not overload public map infrastructure.

Quiet

Road and activity quietness

Source: OpenStreetMap road geometry through the VeloraMaps OSM cache, plus time-of-day adjustment and optional user calibration. Refresh: OSM-derived signals are cached for roughly 12 to 24 hours; time adjustment recalculates on every visit.

Green

Parks, gardens, woodland, and natural context

Source: OpenStreetMap landuse, leisure, and natural features through Overpass or cached OSM signals. If OSM is temporarily unavailable, VeloraMaps can estimate from place category and name, labelled as approximate. Refresh: cached OSM data refreshes about daily.

Social

Nearby activity and social amenities

Source: OpenStreetMap points of interest such as cafes, restaurants, shops, transport nodes, and social venues. Refresh: cached OSM signal results refresh on a 12 to 24 hour cycle, with live fallback when available.

Work-friendly

Focus conditions and practical support

Source: a blend of Quiet, amenities, bounded weather context from Open-Meteo, and nearby services from OSM. Refresh: weather updates live from Open-Meteo; map signals use the OSM cache; user calibration applies immediately on the device.

From inputs to final score

The score is built in stages. Raw signals become sub-scores, stable place signals are weighted for the selected intent, live weather and time are applied as bounded right-now modifiers, then the final score is rounded for display. Confidence is shown separately as a label and likely range rather than being hidden inside the score.

Input signals
  • Road density and proximity
  • Parks, nature, and landuse
  • Amenities and activity points
  • Weather and local time context
Score breakdown
  • Quiet
  • Green
  • Social
  • Work-friendly
Final score
  • Weighted by mode
  • Shown with confidence and likely range
  • Explained with source labels
  • Shown as 0 to 100

Score ranges

0-39 Low

Low fit

A low score usually means one or more core conditions are clearly working against the selected intent. Example: a busy road junction with little greenery may be low for Relax because traffic, noise, and limited nature all push against the goal.

40-69 Medium

Mixed fit

A medium score means useful qualities exist, but with trade-offs. Example: a city centre street might be strong for Social because it has food and transport, but only medium for Stay because high activity can become tiring.

70-100 High

Strong fit

A high score means the local signals line up well for the selected use case. Example: Principe Real in Lisbon can score high because it combines gardens, cafes, walkability, and enough calm to feel balanced.

Limitations

Vibe Scores are estimates, not guarantees. They do not measure personal safety, rent, crowd mood, air quality at street level, accessibility for a specific disability, opening hours, temporary roadworks, events, queues, or whether a venue is personally enjoyable. They also depend on public map coverage: some cities have richer OpenStreetMap data than others, and private spaces may be missing.

What the score does capture

It captures nearby structure: road intensity, mapped green or natural areas, amenity density, activity proxies, live weather, and how those signals compare with the selected intent.

What the score does not capture

It does not replace local judgement. A high score should be treated as a useful shortlist signal, not a final decision about where to live, work, stay, or meet.