People are using Strava heatmaps to pick "safe, wealthy" suburbs. Cute trick — except joggers run through rough areas too. So we blended the activity-heat signal with what the heatmap can't show: actual crime statistics and census incomes, per suburb. Drag the sliders to set what you care about; the ranking re-orders live.
The viral "Strava heatmap = good suburb" trick has a real insight buried in it — places where people run, ride and walk at night are usually places people feel safe. But heat lines alone mostly show where fit people live, and they happily run straight through high-crime pockets. So we treat activity as one signal of five, and back it with official data.
Density of public GPS traces (OpenStreetMap, ODbL) plus gyms, ovals, pools and sports facilities per km². The glowing map layer is this data, rendered raw.
Offences per 1,000 residents from South Australia Police, full financial year, per suburb. Note: SAPOL withhold sexual offences at suburb level for privacy, so this counts offences against the person and against property — it is not directly comparable with the Melbourne or Sydney totals.
Median household income vs median rent (ABS Census 2021). High score = your rent buys more suburb than it should.
OpenStreetMap park and café density per km², blended 50/50 with the same counts normalised per 1,000 residents so both crowded inner suburbs and sparser leafy ones can score well.
A composite of fitness-infrastructure density (gyms/ovals/pools per capita), proximity + attendance at the nearest parkrun event, % of the suburb within 400m of a park, and café density per capita. Off by default — drag its slider up to weight it into the ranking.
Every suburb gets a 0–100 percentile per signal. Your sliders weight the signals; the headline number is the weighted blend. No paid placements, ever.
Ten rankings computed from the same data — each one shows its method.
299 suburb profiles — scores, rents, crime rate, parks and nearest parkrun.