How Icioola computes its ratings
Your priorities are ranked, then converted into weights by the SR method, which blends the reciprocal-rank and linear-rank rules: the criterion ranked i out of n gets 1/i + (n + 1 − i)/n, then all are scaled to 100%.
With four criteria that gives 44, 27, 18 and 11%: the first criterion weighs markedly more than the last, without you having to drag any slider.
A weak point weighs more than a strong one
Criteria are combined by geometric mean, not arithmetic. A weak point costs far more than a strong point brings: with equal weights, a town scoring 10 and 90 gets 30, where a plain average would give it 50. Its strengths only partly offset the hole. To rule a town out altogether, the filters below are what count.
Budget filters; commute filters, and ranks by public transport
If a town exceeds your budget or your maximum commute, it is excluded, not penalised. We tell you how many towns were excluded, and why. In the 22 local editions, where commutes are computed door to door on published timetables, the commute also becomes one of your ranked criteria; everywhere else it is computed by car and remains a filter only.
Each measure on its actual value, across all of France
Since 9 October 2026, the public score rates each measure on its actual value, on one scale for all of France, rather than on a rank. A rank blew up tiny gaps in the middle and flattened real gaps at the extremes. Now a rate twice the French median costs 20 points wherever it sits on the scale. The six families weigh what people in France say matters (OpinionWay survey of October 2025, 1,014 people): setting 24%, safety 22.2%, health 15%, transport 14.2%, everyday access 13.8%, schools 10.8%.
Where the thresholds sit
The grade reads off the score out of 100. The thresholds were set once, on the 2026 towns, so that most sit in the middle (10% A, 40% C, 10% E), then frozen, like the Nutri-Score or energy labels: a town changes letter when its own figures change, never because others moved.
- A — 57 or more (2,255 towns)
- B — 50 to below 57 (5,296)
- C — 40 to below 50 (10,509)
- D — 32 to below 40 (5,436)
- E — below 32 (2,600)
Distribution measured across 26,096 ranked towns. It will move from year to year; the thresholds will not.
The score out of 100 is not a school mark
It brings the six families together by a geometric mean, where a weak spot weighs more than a strong one. Its French median sits at 44.8, and observed values run from 14.3 to 75.4. A town at 47 is therefore above half the country, even if it looks far from 100.
Measured markers: 10% of towns fall below 32, a quarter below 38.1, three quarters below 51.1, and 10% exceed 56.3.
What never enters the calculation
The share of social housing, places of worship and the mayor's political affiliation are excluded from scoring, by explicit choice. An automated test verifies at every build that none of this data can slip into the engine.
When data is missing
A criterion without data receives a neutral score — neither bonus nor penalty — and the real coverage of each criterion is published on the French Data & licences page.
Try it with your address · How commute times are calculated · Data & licences · Where France stands · Performing arts schools · Resilience ranking · Buying to let · Version française