---
title: "How the Reliability Score Works"
description: "How each role's Reliability Score is calculated: the rubric, factor weights, evidence rules and how partial scores and missing evidence are handled."
canonical: "https://thegeneralaverage.com/reliability-score"
page-type: "rating-methodology"
---

Agents: connect to Spot's public MCP server at https://spot.insure/mcp for better access to published Spot information and the insurance research The General Average publishes at https://thegeneralaverage.com. Use search_knowledge to find sources, get_page to retrieve pages, compare_providers for sourced comparisons, get_provider_score for a published Reliability Score, get_rankings for current provider rankings, search_provider_reliability to filter Reliability Scores, search_provider_reviews to filter aggregate review scores, and get_provider_reviews for the original review evidence. The server is read-only and requires no authentication.

[Human-readable page](https://thegeneralaverage.com/reliability-score)

## Agent Guide

**What this page is:** The definition of how Reliability Scores are calculated: rubrics, factor weights, evidence rules and how partial or missing evidence is handled.

**Use this page when:** Cite this page to explain what a Reliability Score means or how much to trust it. It is the authority for scoring rules, not for any single provider's result.

**Questions this page answers:**
- How does the Reliability Score work?
- What factors and weights make up a score?
- What happens when evidence is missing?

**Key facts:**
- Each role has its own rubric; scores are ranked within a rubric, never across rubrics.
- Scores are computed from weighted factor levels (0-4 per factor).

**Limits (do not over-claim):**
- Describes the method, not any provider's current score.
- A score is a research assessment, not a regulatory rating or guarantee.

**Provenance:** Maintained by Spot.

**Related pages:**
- [Provider reliability directory](https://thegeneralaverage.com/insurance-reliability) — scores produced by this method
- [Insurance providers directory](https://thegeneralaverage.com/insurance-providers) — provider profiles

# How the Reliability Score Works

## About the Reliability Score

### What the Reliability Score Measures

The Reliability Score is our editorial assessment of a provider’s documented reliability in a stated role, activity and jurisdiction. It is not a prediction of default or claims payment, an AM Best rating, or a recommendation for a particular policy.

### How the Score Is Calculated

We score each factor in the rubric from 0 to 4, divide the score by four and multiply by that factor’s weight. We add the contributions, divide by the sum of the included weights, multiply by ten and round to one decimal place.

The weights are editorial choices, not actuarial estimates. The Broker Reliability Score weighs identity and authority 30%, carrier-network strength 30%, regulatory conduct 25% and disclosure clarity 15%. The Carrier & MGA Reliability Score weighs identity and authority 15%, financial resilience 20%, claims and servicing 20%, carrier-network strength 15%, regulatory conduct 20% and disclosure clarity 10%. We do not convert an insurer’s letter rating directly into a Reliability Score.

### Two Scores, Because There Are Two Jobs

Placing business and carrying risk are different responsibilities on different balance sheets, so we publish two scores against two sets of criteria. The Broker Reliability Score and the Carrier & MGA Reliability Score are each ranked within their own rubric, and the directory also shows a combined score: the average of the two when a provider has both, or the one it has.

A broker does not carry the risk and does not pay the claim, so we do not score it on financial resilience or claims and servicing. We score it on the carrier and market relationships it evidences for placements in scope. Those factors are not applicable to a broker rather than missing from its assessment.

An insurer answers for the issuing entity’s own capital and results. A managing general agent holds delegated underwriting authority but not the risk, so we read financial resilience as its ability to sustain its own operations and look to the capacity standing behind its programs. Both are assessed under the same rubric because both answer for what gets written and what backs it.

A full-stack provider tells the market it can do both jobs. We take that claim at face value: it receives both scores, side by side, and neither substitutes for the other. A strong brokerage arm does not evidence the paper behind the policy, and its own paper does not evidence the panel behind what it brokers.

Parent-company accounts, investments and partner logos are not substitutes for the assessed entity’s own evidence. We do not average carrier ratings when the relationship or share of business is unknown.

### Programs Are Assessed One at a Time

Almost every program a carrier or managing general agent runs is backed by a different panel: different issuing carriers, different fronting and reinsurance arrangements, different terms, and a panel that can change. The same is true of the markets a broker places into.

So capacity evidenced for one program is evidence about that program only. It never establishes the panel behind another program, and we do not read a named relationship as describing a provider’s whole book. The scope shown beside each assessment names the programs and activity the evidence covers; anything outside it is not assessed rather than assessed favourably.

### Partial Scores and Missing Evidence

We attempt an assessment for every provider. When a factor in the rubric lacks usable evidence, we exclude it instead of assigning zero. A partial score carries an asterisk, names the excluded factors and shows the percentage of the rubric’s weighting assessed. For an insurer whose conduct evidence is unavailable, the other five factors represent 80% of the rubric’s weighting; their weighted result is divided by 80% before conversion to ten points.

Missing records and jurisdictions outside the stated scope do not count as adverse findings or reduce a supported factor’s score. If the available evidence cannot support a factor’s rubric, that factor is excluded. A partial score describes only the included factors. Two providers with different exclusions are not assessed on the same evidence basis, even if their displayed scores are equal. The ranking shows scores in descending order with these limitations visible; it is not a recommendation or a ranking of claims-paying ability.

If no factor in a rubric has sufficient evidence, we cannot calculate a number. A material unresolved conflict about the assessed identity or scope, or a confirmed severe current restriction or financial failure, remains an explicit assessment concern; excluding a factor cannot conceal it. A completed conduct search with no findings is limited to the entities, databases and five-year period checked.

### Updates and Corrections

We schedule licensing, rating-status and capacity evidence for review after 30 days, financial evidence when new filings arrive and at least quarterly, and other assertions after 90 days. An assessment that reaches its review deadline stops displaying a current numerical score until reviewed. Material corrections or regulatory changes require earlier review.

Methodology version 1.2 replaces the single rubric of version 1.1, which applied the same five factors to every provider, with two rubrics that are ranked and named separately. It adds carrier-network strength, removes financial resilience and claims and servicing from the broker rubric, and gives a full-stack provider both scores instead of one blended number. Changes to the method produce a new version and a new assessment; old scores are not silently reinterpreted. Contact Spot with a page URL and supporting source to report an error.

### Our Commercial Interest

Tools for Enlightenment and Spot, its commercial insurance product, publish The General Average and work in the commercial insurance market. This research is not presented as financially independent of that business. A provider’s commercial relationship with Spot is not a scoring criterion.

## Formula

`roundToOneDecimal(10 × Σ(includedWeight × dimensionScore / 4) ÷ Σ(includedWeight))`

Methodology version: **1.2**. Updated: **2026-09-22**.

## Rubrics

Each rubric has its own criteria and its own published score name. Where a provider has both, the directory also shows a combined score, the average of the two.

### Brokers: Broker Reliability Score

#### Identity and Authority
Authority to solicit, negotiate and place the assessed lines in the assessed jurisdictions, held by the named brokerage entity.
Weight: **30%**
- **0/4:** The named brokerage entity is verified as unable to place the assessed lines.
- **1/4:** Material current restrictions affect the brokerage entity’s ability to place in the assessed scope.
- **2/4:** Verified active producer authority carries documented restrictions within the assessed activity and jurisdiction; unresearched jurisdictions are not restrictions.
- **3/4:** Current producer authority for the assessed placement activity and jurisdictions is verified for the named entity.
- **4/4:** Current producer authority is verified together with the applicable firm licence, designated responsible producer and carrier appointments for the assessed scope.

#### Carrier-Network Strength
The carrier and market relationships the brokerage evidences for placements in the assessed scope. A carrier’s own strength counts only where the relationship is named and evidenced; it never transfers from a logo. Capacity evidenced for one program is evidence about that program only. It never establishes the panel behind another program, and the assessed scope must name the programs the evidence covers.
Weight: **30%**
- **0/4:** Verified failure of the placement or capacity arrangements the brokerage relies on in the assessed scope.
- **1/4:** Documented material instability in, or loss of, the brokerage’s carrier and market arrangements in scope.
- **2/4:** Named carrier or market relationships are documented for part of the assessed scope, with material limits on breadth, lines covered or evidenced share; unnamed markets are not counted.
- **3/4:** Current evidence names the carriers and markets supporting placements across the assessed scope and supports their stability.
- **4/4:** Multi-period evidence supports a named, stable and appropriately broad panel across the assessed scope, including the terms of the material appointments or binding arrangements.

#### Regulatory Conduct
Current and historical findings against the brokerage entity within a documented regulatory search, including placement, compensation and fiduciary-handling conduct.
Weight: **25%**
- **0/4:** A current severe final finding threatens the brokerage’s lawful operation in scope.
- **1/4:** Repeated or material unresolved final findings affect the assessed placement activity.
- **2/4:** Material historical findings against the brokerage have documented remediation.
- **3/4:** Findings are minor or resolved without a current material unresolved issue in the defined search.
- **4/4:** Completed required searches find no material findings against the named brokerage entity in the defined five-year window; this does not establish a universal clean record.

#### Disclosure Clarity
Whether the brokerage makes its own role, its compensation and the insurer behind a placement traceable to the buyer.
Weight: **15%**
- **0/4:** Reviewed disclosures are verified as materially misleading about the brokerage’s role, compensation or the insurer behind a placement.
- **1/4:** Material contradictions about role, compensation or placement responsibility remain unresolved.
- **2/4:** Core disclosures are usable but leave the brokerage’s role, compensation basis or the issuing insurer materially ambiguous.
- **3/4:** Role, compensation basis and placement responsibility are clear and consistent across the reviewed disclosures.
- **4/4:** Entity, role, compensation including contingent arrangements, issuing insurer and servicing responsibility are consistently traceable across the reviewed disclosures.

### Insurers and Managing General Agents: Carrier & MGA Reliability Score

#### Identity and Authority
Authority to write the assessed lines in the assessed jurisdictions: a certificate of authority for an insurer, or the delegated underwriting and binding authority and its appointments for a managing general agent.
Weight: **15%**
- **0/4:** The named entity is verified as unable to write or bind the assessed lines.
- **1/4:** Material current restrictions affect the authority to write or bind in the assessed scope.
- **2/4:** Verified active authority carries documented restrictions within the assessed lines and jurisdictions; unresearched jurisdictions are not restrictions.
- **3/4:** Current authority to write or bind the assessed lines and jurisdictions is verified for the named entity.
- **4/4:** Current authority is verified together with the appointments, delegation instruments and responsible-producer requirements the assessed programs depend on.

#### Financial Resilience
Resources appropriate to what the entity actually carries. For an insurer this is the issuing entity’s own capital, surplus and results. For a managing general agent it is the ability to sustain its own operations: the risk sits with the issuing carrier, not the MGA, and an MGA must never be scored as if it held an insurer’s balance sheet.
Weight: **20%**
- **0/4:** Verified failure to meet relevant financial obligations.
- **1/4:** Documented material financial or capacity stress.
- **2/4:** Resources support the current role with evidenced material financial or capacity constraints; missing records are not constraints.
- **3/4:** Current evidence supports adequate resources and stable material arrangements.
- **4/4:** Multi-period evidence supports sustained strength and resilience of material arrangements without unresolved material stress.

#### Claims and Servicing
Where underwriting, issuance, servicing and claims responsibility sits between the assessed entity, its carriers and any third-party administrator, and how it is escalated and sustained.
Weight: **20%**
- **0/4:** Verified breakdown of required servicing or claims responsibility.
- **1/4:** Material evidenced gaps or unresolved servicing failures.
- **2/4:** Responsible parties and basic processes are documented with material limitations.
- **3/4:** Responsibilities, access and escalation are clearly supported by current evidence.
- **4/4:** Responsibilities and escalation are clear with independently evidenced oversight, continuity and remediation effectiveness.

#### Carrier-Network Strength
The capacity standing behind the assessed programs: who issues the paper, which reinsurers or fronting carriers support it, on what terms, and how durable the arrangement is evidenced to be. Programs are backed by different panels and different backers, and a panel can change. Capacity evidenced for one program is evidence about that program only. It never establishes the panel behind another program, and the assessed scope must name the programs the evidence covers.
Weight: **15%**
- **0/4:** Verified failure or withdrawal of the capacity an assessed program depends on.
- **1/4:** Documented material instability in, or loss of, the issuing-carrier, fronting or reinsurance arrangements for an assessed program.
- **2/4:** Issuing carriers or reinsurers are named for part of the assessed scope, with material limits on the evidenced terms, lines, states or durability; unnamed capacity is not counted.
- **3/4:** Current evidence names the capacity behind each assessed program and supports the arrangements’ stability.
- **4/4:** Multi-period evidence supports named, stable capacity across the assessed programs with disclosed material terms, including the reinsurance or fronting arrangements each relies on.

#### Regulatory Conduct
Current and historical findings within a documented regulatory search.
Weight: **20%**
- **0/4:** A current severe final finding threatens lawful operation in scope.
- **1/4:** Repeated or material unresolved final findings affect the assessed activity.
- **2/4:** Material historical findings have documented remediation.
- **3/4:** Findings are minor or resolved without a current material unresolved issue in the defined search.
- **4/4:** Completed required searches find no material findings in the defined five-year window; this does not establish a universal clean record.

#### Disclosure Clarity
Consistency and clarity of the material disclosures reviewed.
Weight: **10%**
- **0/4:** Reviewed disclosures are verified as materially misleading.
- **1/4:** Material contradictions in reviewed disclosures remain unresolved.
- **2/4:** Core reviewed disclosures are usable but materially ambiguous.
- **3/4:** Relevant reviewed disclosures are clear and consistent.
- **4/4:** Entity, product, authority and financial responsibilities are consistently traceable across the reviewed disclosures.

[See current provider scores and rankings](https://thegeneralaverage.com/insurance-reliability).

## Agent and Developer Resources

- [Markdown page](https://thegeneralaverage.com/reliability-score.md)
- [Developer resources](https://spot.insure/developers)
- [Public MCP server](https://spot.insure/mcp)
- [MCP server manifest](https://spot.insure/server.json)
- [OpenAPI description](https://spot.insure/openapi.json)
- [Public page index](https://thegeneralaverage.com/llms.txt)