What STATS19 is
The Department for Transport (DfT) publishes a yearly file of every collision attended by police in Great Britain. It's the cleanest road-safety dataset in the country: lat/lng, severity (fatal / serious / slight), casualty counts, road type, weather, light conditions.
We ingest the last 3 years of it into our property scoring. For every property we measure collisions within 1500m of the centroid and weight by KSI (killed-or-seriously-injured) — fatal × 5, serious × 2, slight × 1.
What 'normal' looks like
A typical English suburban outcode: 3–8 KSI per year in a 1500m radius. Anything in that range is normal urban traffic.
A central-city outcode: 15–40 KSI per year. Higher because you're surrounded by main roads and pedestrian volume is dense.
A rural village: under 1 KSI per year, but a single A-road collision can dominate the local figure.
What's a red flag
What matters for family buyers
Two things STATS19 is uniquely good at telling you:
→ Family rankings (with road-safety integrated): /rankings/family
→ Pull a property report — every report lists fatal/serious/slight counts in a 1.5km radius for the last 3 years
Caveats
STATS19 is collision data only — it doesn't capture near-misses, vehicle speeds, or air quality from traffic. It doesn't capture cyclist incidents that weren't attended by police. So it's a floor on road risk, not a ceiling. Use it alongside a site visit at school-run time.