← Back to validators

Score methodology

What the score means

The diversity score (0–100) measures how well a validator contributes to network decentralization. It combines geographic diversity (country and city) with infrastructure diversity (hosting provider). Less saturated locations and providers receive higher scores.

Input fields used

Logarithmic penalty

Each field is penalized based on how many other validators share the same value. The penalty uses the natural logarithm so that going from 1→2 validators has a small impact, while 10→100 has diminishing additional impact.

Penalty = log(max(1, count)). For example: count 1 → 0, count 2 → 0.69, count 10 → 2.30, count 100 → 4.61.

Formulas

geoPenalty     = log(countryCount) × 14 + log(cityCount) × 6
providerPenalty = log(providerCount) × 18

geoScore     = clamp(100 - geoPenalty, 0, 100)
providerScore = clamp(100 - providerPenalty, 0, 100)

total        = clamp(geoScore × 0.55 + providerScore × 0.45, 0, 100)

Worked example

Fictional validator "Acme Staking" in Germany, Berlin, provider AWS. Assume: 20 validators in Germany, 5 in Berlin, 15 on AWS.

geoPenalty = log(20)×14 + log(5)×6 = 2.996×14 + 1.609×6 ≈ 41.93 + 9.65 ≈ 51.6
providerPenalty = log(15)×18 ≈ 2.708×18 ≈ 48.7

geoScore = 100 - 51.6 ≈ 48
providerScore = 100 - 48.7 ≈ 51

total = 48×0.55 + 51×0.45 ≈ 26.4 + 23.0 ≈ 49 → badge: saturated

Badge interpretation

Missing metadata handling

Validators are classified into two categories: