Method CRIS-2.0-mvp2, configuration version 0
Figures on this page are the values in force now. An administrator can change them; every change is versioned and logged.
What the outputs mean
| Resilience score, 300-900 | Strength of controls today. Same calculation for every organisation, country and sector. Grades: A from 800, B from 700, C from 600, D from 500, E below. |
| Evidence confidence | Share of the score's weight backed by a measurement or an independent attestation. It sets a ceiling: 70%+ High, 900; 40%+ Medium, 820; below that Low, 740; nothing verified Provisional, 700. |
| Veracity index | 100 x (1 - average over-declaration) across controls that have both a declared and a measured value. Shown from 5 such controls. Under-declaration does not count. |
| Likelihood vs peers | A relative band from the grade, open gates and unresolved incidents. Not a probability. |
| Exposure band, 1-5 | Size of potential loss from revenue, record counts, IT dependence and regulated status. Separate from the score. |
| Loss scenarios | Simulated loss if an event occurs, from your own business figures and a parameter pack. No annual probability. |
Score calculation
Each control has a maturity between 0 and 1. Standard questions: 0, 0.33, 0.67 or 1.0. Eight high-weight controls: 0.25 x design + 0.25 x coverage + 0.3 x operation + 0.2 x testing. "Unknown" is 0. "Not applicable" is removed once a reviewer accepts the reason.
A verified value replaces the declared value. Unverified answers are not discounted; they limit confidence instead.
Dimension score = weighted average of its controls. Composite = weighted sum of dimensions. Score = 300 + 600 x composite, then the lowest of the gate cap and the confidence ceiling applies.
| Dimension | Weight | Questions |
|---|---|---|
| Governance and accountability | 15% | 1, 2, 3, 4, 5, 6, 7, 8 |
| Identity, people and insider risk | 15% | 9, 10, 11, 12, 13, 14 |
| Infrastructure and endpoint | 15% | 15, 16, 17, 18, 19, 20, 21 |
| External exposure | 15% | 22, 23 |
| Data protection and privacy | 10% | 24, 25 |
| Detection and response | 10% | 26, 27, 28, 29 |
| Recovery and resilience | 10% | 30, 31, 32 |
| Third-party and dependency | 10% | 33, 34, 35 |
Item weights range from 1 to 3 by default; the eight four-part controls (9, 10, 17, 18, 19, 26, 30, 31) carry the highest.
External exposure
With an external check less than 14 days old, this dimension is 1 minus the deductions below. Only findings with High ownership confidence count. A finding open more than 30 days counts 1.5 times. Closed findings do not count. Disputed findings stay in the score, flagged, until reviewed.
| Finding | Deduction | Cap |
|---|---|---|
| Internet-exposed vulnerability listed in CISA KEV | 0.25 each | 0.50 |
| Exposed vulnerability, EPSS 0.5 or more, not in KEV | 0.10 each | 0.30 |
| Exposed RDP, SMB, Telnet or database port | 0.15 each | 0.30 |
| Exposed admin panel | 0.10 each | 0.20 |
| Corporate credentials in infostealer logs under 90 days old | 0.15 | 0.25 |
| Credentials in older breach collections | 0.05 | 0.05 |
| DMARC absent or policy none | 0.08 | 0.08 |
| Expired or weak TLS on customer-facing service | 0.05 | 0.05 |
| Secrets in public code repositories | 0.15 | 0.15 |
| Publicly readable cloud storage with non-public data | 0.20 | 0.20 |
| Active lookalike domain with mail or login page | 0.05 | 0.05 |
| Listed on a ransomware leak site in last 24 months | 0.20 | 0.20 |
Gates
A gate is a critical failure that good scores elsewhere cannot offset. It caps the score.
| Gate | Score cap |
|---|---|
| KEV-listed vulnerability exposed to the internet beyond the allowed days | 599 |
| Privileged credentials in recent infostealer logs, not rotated | 599 |
| No offline, immutable or separately held backup of critical systems | 649 |
| MFA absent on remote access or email (lowest coverage band) | 649 |
| No EDR or equivalent on servers and endpoints (lowest coverage band) | 699 |
| Material incident found from external data and not disclosed | 549 |
| Recently listed on a ransomware leak site, no remediation attestation | 499 |
In this version the backup gate opens when the backup copy is on the same network or absent; the MFA gate when under 70% of accounts are covered; the EDR gate when under 75% of endpoints and servers are covered.
Loss scenarios
Four scenarios: ransomware and interruption, data breach, funds-transfer fraud, dependent interruption. Durations and shares use PERT distributions; per-record cost is lognormal. Controls shift the ranges (backup and restore state, tested response plan, encryption, payment verification, provider fallback). 10000 runs; the seed is stored. Median, 1-in-10 and 1-in-100 values are reported per scenario. Scenarios are not summed.
Parameter pack in use: india-v0. PLACEHOLDER VALUES. Planning assumptions, not sourced from claims data. Replace with pilot insurer, panel-firm and actuarial inputs before any external use.
Limits
- Weights, thresholds, ceilings and gate caps are starting values. They have not been calibrated on claims data.
- The score is not a prediction of a claim and does not set a premium. The insurer decides price, terms and claims.
- A declared-versus-measured gap is not proof of misrepresentation, breach of condition or causation.
- Geography and sector do not change the score. They change obligations named in the questions, loss parameters and context.
- Only counts, ratios, dates and uploaded files are held. No employee-level monitoring data is collected.
- Sector modules are rated No gaps, Gaps or Critical gaps and are reported beside the score, not inside it.
- An insurer's rule set routes a submission under that insurer's appetite. It never changes the score.
Measurements accepted
Aggregate values from identity, endpoint, patching, logging and backup tools, sent by CSV or the ingestion API, map to the coverage and operation parts of the eight high-weight controls and to questions 13 and 16. UBA/UEBA anomaly rates raise a warning above 1 unresolved per 100 users; they are not scored. Measurements stay valid for 30 days; attestations for 365 days.