APPLIED ANALYSIS SERIES — AAS-013

THE DIRECTIVE™ — APPLIED ANALYSIS SERIES — AAS-013

The Predictability Paradox™: When Foreseeable Harm Produces Institutional Surprise

A SAFECHAIN™ Analysis Opening the Consequence Tier of The Directive™ Applied Analysis Series: Foreseeability, Safeguarding Response, and Preventable Harm

Reference: SAFECHAIN/AAS/2026/013

Author: Samantha Avril-Andreassen FRSA

Organisation: SAFECHAINN Ltd (Company No. 12038453)

Abstract

AAS-010 to AAS-012 form the Accountability / Governance Synthesis Tier of this series: AAS-010 examined why recognised problems recur (The Implementation Paradox™, Paper 23); AAS-011 examined why responsibility for those recurrences is frequently not locatable (The Accountability Gap™, Paper 22); AAS-012 examined why their persistence may not register as failure within an institution's own evaluation of itself (Institutional Capture™, Paper 32, applied via a five-part taxonomy from Paper 6).

AAS-013 opens a second tier — the Consequence Tier — examining what these governance conditions produce when they intersect with a specific safeguarding decision: foreseeable harm that occurs despite, or because of, conditions AAS-010 to AAS-012 each described. Using Paper 24 (The Predictability Paradox™) as its primary framework — why foreseeable harm continues to produce institutional surprise — supported by Paper 23, Paper 26 (The Continuity Deficit™), Paper 22, and Paper 33, this paper examines AAS-012's five-part taxonomy of capture mechanisms as its primary series input, asking what each capture type produces when applied not to an institution's general performance, but to a specific, foreseeable instance of harm.

AAS-013 marks the beginning of the Consequence Tier of The Directive™ Applied Analysis Series. Whereas AAS-010 to AAS-012 examined implementation, accountability, and institutional self-protection, AAS-013 examines the consequence of those governance conditions: foreseeable harm that systems had sufficient information to address but did not effectively prevent.

Keywords: The Predictability Paradox™, Foreseeability, Institutional Surprise, Safeguarding, Preventable Harm, Governance, SAFECHAIN™, The Directive™

A Note on Framework Verification, Tier Structure, and Terminology

Three corrections, made during this paper's drafting, are recorded here. First, an earlier framework proposed for this paper cited 'Paper 24 — Preventable Harm Architecture™' as primary. Checked against the canonical Application Index, Paper 24 is The Predictability Paradox™ — concerned with why foreseeable harm continues to produce institutional surprise. This is the third instance of a proposed framework citing a paper number under an invented name (cf. Paper 6/32 in AAS-006, AAS-007, AAS-012). Paper 24's actual concern — institutional surprise at foreseeable harm — is, if anything, a more precise fit for this paper's thesis than the proposed name, and this paper proceeds on that basis.

Second, on tier structure: AAS-012's Register entry described AAS-010 to AAS-012 as forming a 'trio.' AAS-013, rather than extending that trio to a fourth instalment covering the same ground, opens a second tier — the Consequence Tier (AAS-013 to AAS-015) — building on the first tier's conclusions as inputs rather than re-examining AAS-001 to AAS-009's original findings a further time. Section 2 sets out both tiers' relationship explicitly. AAS-001, AAS-002 and AAS-007 appear in this paper only as supporting examples, not as this paper's primary evidence base; that role is taken by AAS-012's five-part taxonomy.

Third, on terminology: this paper uses, in plain language and without capitalisation or trademark, what this paper describes as risk conversion — the process by which information becomes action, concern becomes intervention, and knowledge becomes protection. This is the fifth instance of a descriptive term introduced during drafting that could be mistaken for a new framework (cf. AAS-004's Principles, AAS-009's multiplication effect, AAS-011's accountability mapping, AAS-012's capture types — though the latter were formally applied via Paper 6). A Methodology addendum addressing this recurring pattern, flagged as an open item since AAS-004 and recorded again in AAS-011's Register note (3g), is now overdue across five papers and is recorded here as a standing item for the next available session focused on architecture maintenance rather than AAS drafting.

1. Introduction: Two Tiers

AAS-010 to AAS-012 asked, in sequence: why do recognised problems recur (implementation); why is responsibility for that recurrence not locatable (accountability); and why might that recurrence not register as a problem within an institution's own terms (capture). Each paper built on the one before it, and each was explicit that its findings were structural rather than evidence of any individual's failure.

AAS-013 to AAS-015 — the Consequence Tier — ask a different kind of question: given the conditions AAS-010 to AAS-012 described, what happens in a specific case? Not 'why does this pattern recur in general' but 'what does this pattern produce when the thing that recurs is a single, specific, foreseeable harm to a single, specific person.' This paper is the first of the three, and its primary task is to connect AAS-012's five capture types — described there in general terms — to the specific question of foreseeability: for each capture type, what does it look like when the 'outcome' an institution is not registering as a problem is a risk to an identifiable person, rather than a general performance measure?

2. The Two Tiers, Set Out

The table below sets out both tiers of synthesis paper, for reference. The first tier is established; the second begins with this paper.

Paper Framework (primary) Question

Tier 1: Accountability / Governance Synthesis (AAS-010 to AAS-012)

AAS-010 The Implementation Paradox™ (Paper 23, primary) Why recognised problems recur: correctly-scoped responses, unscoped relationships between them

AAS-011 The Accountability Gap™ (Paper 22, primary) Why responsibility for recurring gaps is frequently not locatable, even where the gap is describable

AAS-012 Institutional Capture™ (Paper 32, primary) Why the persistence of these gaps may not register as failure within an institution's own evaluation of itself

Tier 2: Consequence Tier (AAS-013 to AAS-015)

AAS-013 The Predictability Paradox™ (Paper 24, primary) What do AAS-012's five capture types produce when the unregistered outcome is foreseeable harm to an identifiable person?

AAS-014 and AAS-015 are reserved as the remainder of the Consequence Tier and are not developed in this paper.

3. The Predictability Paradox™, and Why 'Surprise' Is the Right Word

Paper 24, The Predictability Paradox™, examines why foreseeable harm continues to produce institutional surprise. This is a precise description of a specific moment: the point, after harm has occurred, at which a review is commissioned, and that review finds — as such reviews very often do — that the information needed to anticipate the harm was already present in the system beforehand. The 'surprise' Paper 24 names is not a claim about any individual's emotional reaction. It is a description of the gap between what a system, considered as a whole, could have known in advance, and what the same system's response to the harm treats as newly discovered.

This paper's argument is that AAS-012's five capture types describe, in general terms, exactly the mechanism by which this gap is produced. Section 4 takes each of three capture types from AAS-012 — Procedural, Resource, and Compliance Capture — and traces what each produces when the 'measure' that does not register a problem is, specifically, a measure that would have flagged a foreseeable harm in advance.

4. From Capture Type to Predictability Gap

AAS-012 applied its five capture types to four general findings (AAS-001, AAS-007, AAS-010, AAS-011). This paper revisits three of those applications — Procedural, Resource, and Compliance Capture — specifically through the lens of foreseeability: in each case, what information was available before the relevant harm, and why might that information not have been registered, by the measure in question, as something requiring a response?

Capture Type (AAS-012) Source What the measure does not register

Procedural Capture AAS-001 — C1A form A form's completion is measured; the coercive control field's absence is not, so the information gap it produces is not visible as a gap to the system measuring completion

Resource Capture AAS-007 — strangulation statistics Recording within 'Assault with injury' satisfies existing reporting requirements; a new category would require a resourcing decision the existing requirement does not prompt

Compliance Capture AAS-010 — Express FR Pilot Success is measured against PD36ZH's own scope; cumulative burden outside that scope (AAS-009) is not within what 'success' was defined to measure

In each of these three cases, the information that would establish foreseeability — that coercive control was alleged (AAS-001), that strangulation had occurred (AAS-007), that procedural burden extended beyond a pilot's scope (AAS-010, via AAS-009) — existed within the system. In each case, AAS-012 identified a measure (form completion; existing reporting categories; a pilot's own success criteria) that did not, and structurally could not, register that information as relevant to itself. This paper's contribution is to note that, were any of these three situations to result in harm to a specific person, a subsequent review would very likely find — as Section 3 describes — that the information was 'there all along,' producing exactly the institutional surprise Paper 24 names. The surprise is genuine, in the sense that the measure genuinely did not flag it; the information's prior existence is also genuine. Both can be true simultaneously, which is what makes Paper 24 a paradox rather than a simple failure.

5. Risk, Response, and What This Series Has Already Shown

AAS-002 of this series provided an early example of exactly the pattern Section 3 describes, in a context this paper does not need to re-derive. AAS-002 examined the Magistrates' Association's first Domestic Abuse and Stalking Conference, and Professor Jane Monckton Smith's eight-stage homicide timeline — an explicit, published framework for recognising, before the fact, the pattern of escalation that the Alice Ruggles case followed, in Clive Ruggles' own words, 'frighteningly well.' AAS-002 asked what ensures that a recognition framework of this kind, once it exists and is disseminated, is applied consistently — and connected this, via Paper 23, to the Implementation Paradox™ that AAS-010 would later make its primary framework.

This paper's relationship to AAS-002 is as a supporting example, not as primary evidence — that role belongs to AAS-012's capture types, per Section 4. But AAS-002 establishes something Section 3 and Section 4 do not, on their own, establish: that a framework for recognising foreseeability in advance — the eight-stage timeline — already exists, is published, and is disseminated, for at least one category of preventable harm. The Predictability Paradox™, in the case AAS-002 examined, is therefore not a paradox of absent information or absent frameworks; both exist. It is, as Section 3 argues, a paradox of whether the measures by which a system evaluates its own performance — Section 4's three capture types — register that framework's output as something requiring a response, in any specific case, before the case becomes a review.

AAS-001's C1A example, also revisited in Section 4, is the same pattern from a different angle: not a recognition framework's output (AAS-002), but the disclosure that would need to reach such a framework in the first place. Read together, AAS-001 and AAS-002 describe two ends of a chain — information entering the system (AAS-001), and a framework for interpreting it once it has (AAS-002) — and Section 4's capture-type analysis describes why the measures connecting the two ends of that chain may not treat the chain's completeness as something they are responsible for.

6. The Continuity Deficit™, the Responsibility Paradox™, and a Single Case

Paper 26, The Continuity Deficit™ — the most-cited framework in this series, appearing in all of AAS-001 to AAS-009 — describes information or status failing to survive between stages. AAS-010 and AAS-011 each, in different ways, examined what this means at the level of a pattern across many cases. This paper's contribution, opening the Consequence Tier, is to note what Paper 26 means for a single case: each of Section 4's three rows describes a point at which information relevant to one specific person's safeguarding — not 'safeguarding information in general' — failed to survive into the measure that might have acted on it.

Paper 33, The Responsibility Paradox™ — used in eight of AAS-001 to AAS-012 — describes why, for any one of Section 4's three rows, no single actor is positioned to be responsible for the row as a whole. This paper's addition is that, where the outcome of a row is a specific instance of foreseeable harm rather than a general pattern, the Responsibility Paradox™ does not resolve merely because the stakes have become concrete and personal. The diffusion of responsibility AAS-011 and AAS-012 described as a property of complex systems in general remains a property of the system at the moment a specific, identifiable person's safeguarding depends on that system functioning differently than Section 4 describes.

7. What Might Follow

Consistent with the Consequence Tier's purpose — examining what AAS-010 to AAS-012's conditions produce, rather than proposing to resolve those conditions directly, which AAS-010 to AAS-012 each already addressed in their own terms — the suggestions below concern Section 4's three specific capture-to-predictability connections.

• Whether, for any of Section 4's three rows, a simple addition to the relevant measure — not changing what the measure measures, but adding a single linked question ('does this completion/category/scope-boundary correspond to information that, read together with other information in the system, would indicate foreseeable risk to a specific person?') — could be piloted, without altering the measure's primary function

• Whether AAS-002's eight-stage timeline, as a published recognition framework, could be explicitly cross-referenced from the point at which Section 4's Row 1 measure (C1A form completion) is recorded — not as a new requirement, but as a prompt connecting two things (a form, a framework) that AAS-001 and AAS-002 each examined separately and which Section 5 has now connected

• Whether AAS-014, the next paper in this tier, could examine a specific documented case — a domestic homicide review or safeguarding adult review where the 'information was there all along' finding (Section 3) is a matter of public record — applying Section 4's three-row structure to that case's actual sequence of events, as AAS-001 and AAS-007's worked chains did for their respective subjects

8. Conclusion: The Paradox Is the Connection, Not the Information

Section 3 named the Predictability Paradox™ precisely: the gap is not between information existing and not existing, or between frameworks existing and not existing — AAS-001 and AAS-002 each show that, in the cases this series has examined, both exist. The paradox is between information and frameworks existing, on one hand, and the measures by which institutions evaluate their own performance not treating the connection between them as something those measures are responsible for, on the other.

AAS-012 named this as a structural property of how institutions evaluate themselves, without attributing it to any individual's intent. This paper has applied that property to the specific case of foreseeable harm, and found the same structure: not an absence, but a disconnection, between what a system knows, in the aggregate, and what any specific measure within that system is built to act on. The Consequence Tier's remaining papers — AAS-014 and AAS-015 — are reserved to examine what this disconnection produces, in a specific documented case (AAS-014) and in aggregate (AAS-015), respectively.

Reading This Alongside the Architecture

This paper forms part of The Directive™ Applied Analysis Series and should be read alongside:

• Paper 24 — The Predictability Paradox™ (primary; first application in this series)

• Paper 22 — The Accountability Paradox™

• Paper 23 — The Implementation Paradox™

• Paper 26 — The Continuity Deficit™

• Paper 33 — The Responsibility Paradox™

This paper's primary series input is AAS-012 (Institutional Capture™, and its five-part taxonomy via Paper 6), with AAS-001, AAS-002 and AAS-007 as supporting examples per Section 5.

SAFECHAIN™ welcomes discussion with the institutions named across AAS-001 to AAS-012 on the questions raised in Section 7, particularly the third proposal regarding AAS-014.

References

AAS-001 — Two Reports, One Chain: Reading the Domestic Abuse Commissioner's Findings Alongside Scratching the Surface.

AAS-002 — From Recognition to Practice: A Governance Reading of the Magistrates' Association's First Domestic Abuse and Stalking Conference.

AAS-007 — Non-Fatal Strangulation and the Failure of Risk Transmission.

AAS-009 — The Administrative Weaponisation of Procedure.

AAS-010 — The Implementation Paradox™: When Knowing Is Not the Same as Doing.

AAS-011 — The Accountability Gap™: How Responsibility Becomes Difficult to Locate in Complex Systems.

AAS-012 — Institutional Capture™: Applying Paper 32 — The Power Paradox™.

© 2026 Samantha Avril-Andreassen. All rights reserved. SAFECHAINN Ltd (Company No. 12038453).

Version 1.0

Reference: SAFECHAIN/AAS/2026/013

Copyright & Intellectual Property Notice

© 2026 Samantha Avril-Andreassen. All rights reserved.

SAFECHAIN™, SAFECHAINN Ltd, The Directive™, Participation Integrity™, Passport of Erasure™, Shadow Ledger™, Coercive Debt Lifecycle™, Legacy Harm Architecture™, Institutional Failure Taxonomy™, Vulnerability Index™, Safeguarding Intelligence Model™, Seal of Integrity™, MØPIT™, SIP™, CPIT™, REBUILD™, COMPASS™, and all associated frameworks, methodologies, models, diagrams, terminology, research architecture, governance structures, assessment tools, training systems, and implementation mechanisms are proprietary intellectual property authored by Samantha Avril-Andreassen.

No part of this publication may be reproduced, copied, adapted, distributed, implemented, commercialised, taught, incorporated into training programmes, accreditation schemes, policy frameworks, software systems, artificial intelligence models, governance products, consultancy services, or derivative works without the prior written permission of the author.

The existence of this publication does not grant any licence to implement, replicate, modify, commercialise, or operationalise any SAFECHAIN™ intellectual property.

All rights reserved.