We want to be upfront about this: ipscoper's "IP risk/purity score" is currently a heuristic estimate, not the output of a licensed fraud-detection database (like MaxMind minFraud or IPQualityScore). We'd rather explain exactly how it works than hand you a number that looks precise but isn't transparent about its limits.
Current scoring logic
The score is based on pattern-matching against an IP's ASN and ISP name:
- If the ISP/ASN name matches datacenter or cloud-provider patterns (e.g., containing "hosting," "cloud," "data center") → classified as a datacenter IP, assigned a higher baseline risk value
- If it matches typical residential broadband carrier naming patterns → classified as residential, assigned a lower baseline risk value
- Cases that don't clearly match either pattern receive a mid-range value
This is a name-pattern-matching rule system. It does not draw on historical behavior data, blacklist hit records, or comparison against dedicated proxy/VPN node databases.
What this score can and can't tell you
What it can tell you: a rough directional read — does this IP look more like a datacenter environment, or more like a typical home network.
What it can't tell you:
- Whether this specific IP has a documented history of fraud, spam, or attack activity
- Whether it matches a known proxy/VPN/Tor node database
- A precise numeric "probability of risk" — our score sits in a handful of fixed tiers, not a continuous probabilistic model
If you need more rigorous risk data
For high-stakes scenarios involving payments or account security, we'd recommend layering a licensed commercial risk API — such as MaxMind minFraud or IPQualityScore — on top of our estimate for a second opinion.