Placetrics
Methodology · scoring

Liveability score

Single 0–100 number across seven factors. Each factor passes through a non-linear response curve before weighting, a synergy bonus rewards areas that are both affordable and safe, a flood deduction penalises high-risk zones, and hard penalty gates cap scores in the most dangerous or unaffordable LSOAs.

Formula

The maths

Step 1 — curved_i = f_i(component_i)   [non-linear per factor, see table]
Step 2 — base = Σ (curved_i × w_i) / Σ w_i   [normalised for missing data]
Step 3 — bonus = up to +5 pts if rent_afford > 65 AND crime > 65
Step 4 — flood_deduct = −4 (zone 2) or −8 (zone 3)
Step 5 — score = clamp(0, 100, base + bonus − flood_deduct)
          then: cap at 55 if crime_index < 20
                cap at 50 if rental_afford_idx < 15
Weights

Why each input gets the weight it does

InputWeightRationale
Rental affordability28%The primary constraint for households choosing where to live. Uses a convex curve (t^1.5): an area at the 10th affordability percentile scores only 3/100 on this component — far less than the linear 10 — because extreme unaffordability is functionally a dealbreaker, not just a mild negative.
Safety (crime)22%Applied through a cliff curve: below the 30th percentile nationally the slope is 0.75× (steeper penalty), above it the curve recovers toward linear. An area at the 10th crime percentile caps the total liveability score at 55 via the penalty gate, regardless of how well other components score.
Transport access18%Uses a concave curve (t^0.60): being anywhere near a station is worth more than marginal improvement at the top. A neighbourhood in the 10th transport percentile still gets 25/100 on this component — basic access anywhere has real value.
Healthcare access12%Composite of GP proximity, GP density, GP patient-load ratio, dentist proximity and hospital/urgent-care distance. Near-linear curve (t^0.90). Became a materially stronger signal post-COVID, particularly for areas with overstretched GP practices.
Air quality8%Mean of inverted PM2.5 and NO2 national percentile ranks. Mild concave curve (t^0.85). Clean air has health consequences at all ages, but the marginal benefit of already-clean areas is lower than the harm from polluted ones.
Schools quality7%Weighted Ofsted mix within 2 km (Outstanding→1.0, Good→0.75, RI→0.30, Inadequate→0), re-percentiled nationally. Linear response. Kept as a general-liveability signal, not inflated to avoid penalising childless households who don't weight it.
EPC efficiency5%Running cost signal. EPC band A–G maps to 100–0 via moderate concave curve (t^0.80). The 2022 energy crisis raised the salience of energy bills as part of total housing cost, particularly for lower-income households.
Data sources

What feeds this score

Limitations

What this score doesn't capture

  • The weight percentages are editorial, not regression-derived. Non-linear curves mean the effective influence of each factor varies with an area's actual score — an extremely unaffordable area loses more from the affordability component than the 28% headline weight implies.
  • The synergy bonus rewards areas that are simultaneously affordable and safe. These are genuinely rare at the local-area level, which is why the bonus exists: the algorithm should recognise them as special rather than treating affordable+safe as just the sum of two averages.
  • Penalty gates (caps at 55 or 50) are intentional hard limits. A 57/100 liveability score in a nationally dangerous area would mislead users. The cap is displayed on the page wherever it triggers.
  • Healthcare_index is not re-percentiled after composition — it uses raw sub-scores. This means its distribution is less uniform than the percentile-ranked components, and it may pull the weighted average slightly toward the centre.

Last methodology change: 29 April 2026. See /changelog for prior revisions, or contact hello@placetrics.co.uk with corrections.