How to use the model

User guide

Everything you need to operate the AIOOJ model effectively — from opening the interface to reading each output panel and using the results in settlement negotiations

Contents
Foundation of the model

Default inputs and their rationale

Read this first. Every output the model produces — win probability, expected recovery, settlement floor, corridor — is a mathematical consequence of nine probability inputs (P1 through P9) plus four valuation parameters. The default values loaded when you open the model are not arbitrary. Each one is a calibrated legal assessment anchored to specific evidence, case law, and empirical data from this proceeding. Understanding what those defaults represent — and why — is essential before adjusting any slider. Any variation from the defaults is a deliberate departure from the calibrated position. It should be made only when you have a specific legal reason to do so.

For the complete evidential and legal rationale behind each of the nine default values — what evidence justifies each number, what it means, and when it should change — read the P1 to P9 Defaults: Rationale and Justification page.
The nine probability inputs
Input Default What it measures Why this number — the specific evidentiary and legal basis
P1 / L1
Korea rate reduction		 90% Probability that the Korea rate reduction claim succeeds at trial as a standalone cause of action The 80% rate reduction from 1.875% to 0.375% was implemented without any written amendment — a direct violation of EA Clause 20 and the GPR-TIMAP Double Modification Lock (s.9.6). Aegon advanced five successive justifications across 19 years, each abandoned: the Marketing Agreement authority (abandoned Oct 2024, CC-74); a fictitious LGI-TIMAP agreement (cited twice by Allens, never produced); unilateral MoC authority (defeated by Clause 20); the Hallal acknowledgment (Aug 2024 admission that Marketing Agreement related only to Australia — destroying every Korea justification simultaneously); and the April 2025 settlement offer restoring 1.875% (a tacit admission that the correct rate was always 1.875%). Nine consecutive written agreements across two countries and two calculation architectures all converge on 1.875%. Sequential collapse of five justifications over 19 years establishes institutional knowledge of wrongdoing. 90% reflects the overwhelming documentary record, subject to trial unpredictability.
P2 / L2
Deed / summary judgment		 89% Probability of obtaining summary judgment or judgment on the Deed of Guarantee against TLIC as successor guarantor The Deed demand was made 25 November 2025, compliance deadline 2 December 2025. TLIC (NAIC 69078, Iowa-domiciled) is the confirmed successor guarantor through the JCPLIG → JCPLIC → TLIC chain. The $3,319,108 corrective payment of February 2024 constitutes a s.54 Limitation Act acknowledgment (D3.70, 99.9% certainty). Allens drafted the Deed, the EA, and the Marketing Agreement as a single integrated suite — self-drafting and unified drafter estoppel apply at maximum force. The EA is physically annexed to the Deed as Schedule 1 (Recital B), fixing guarantee scope within the four corners of the executed instrument. Hard-capped at 92% reflecting Spencer v Commonwealth summary judgment caution and corporate succession chain verification pending (CP items with Corrs).
P3 / L3
EA breach methodology		 82% Probability that the Employment Agreement commission underpayment methodology claim succeeds — the core merits of the EA stream Aegon substituted formula-driven estimates for contractually mandated actual reinsurance premiums from July 2005. CC-6 and CC-7 confirm adjustments to actuals ceased from late 2007. CC-8 and CC-9 confirm Aegon cannot retrieve the basis for its estimation rates. The February 2024 corrective payment of $3,319,108 is a binding admission that prior calculations were wrong. The "Two Sets of Books" forensic finding (TAB_38_3, strategic reserve) shows the correct commission ran in the Lumley Ingenium system for 18 years while manual reports paid less — a "Comm to KH" column proving the correct figure was always known. The contract construction authorities (Mount Bruce Mining [2015] HCA 37; Electricity Generation Corporation v Woodside Energy [2014] HCA 7) strongly support a plain-meaning reading of "actual reinsurance premiums received." 82% reflects the 60/40 calibration blend: 60% case-specific legal analysis, 40% empirical dataset signal (Contract/Construction real anchors weighted mean).
P4 / L4
s.55 postponement		 80% Probability that the s.55 Limitation Act 1969 (NSW) concealment extension applies, postponing limitation for the full historical claim period Aegon actively concealed the commission underpayment through a combination of: (a) systematic refusal to produce actual premium records despite the contractual obligation to do so; (b) illegal NDA conditions attached to information requests — conditioning disclosure on Harrison signing non-disclosure agreements, which courts treat as concealment; (c) shifting and irreconcilable explanations across 19 years (the five-position collapse); (d) story-shifting that required Harrison to discover the truth through adversarial forensic analysis rather than voluntary disclosure; and (e) the formal Clause 6.7 invocation of 2024 (preceded by written requests from 2021) that produced continued refusal. The Briginshaw standard applies to concealment findings but the documentary record is unusually strong. Hard-capped at 82% reflecting judicial caution in applying s.55. Reduced from 0.82 to 0.80 per 100-case dataset: Limitation/Timing weighted mean = 0.624 across all cases, applying the 60/40 blend.
P5 / L5
Demand accrual / s.63 rebuttal		 85% Probability that the Deed cause of action accrued on the November 2025 demand (not earlier), defeating any s.63 Limitation Act extinguishment argument The three-limb s.63 rebuttal: (1) Clause 2.1 of the Deed is a primary indemnity obligation, not a guarantee — it survives independently under Sunbird Plaza Pty Ltd v Maloney (1988) 166 CLR 245; (2) the Wardley Australia Ltd v Western Australia (1992) 175 CLR 514 sequencing argument — the Deed cause of action accrued December 2025 when demand was made, carrying its own 12-year limitation period to December 2037; (3) the s.54 acknowledgment (February 2024 corrective payment, D3.70 admission) separately resets any limitation clock under Route D. Ankar v National Westminster (1987) anchors the guarantee construction. The Deed ultimate probability (L2 OR L5) is approximately 98% at defaults — either route succeeding is sufficient.
P6 / L6
Consulting Agreement stream		 81% Probability that the Consulting Agreement commission underpayment claim succeeds — CA principal: $16,730,644 The CA stream applies the same methodology breach analysis as the EA stream (P3), but under a different instrument. The same estimation substitution methodology and the same corrective payment admission apply. The CA stream is the weakest of the three in terms of independent corroboration — it relies more heavily on extrapolation from the EA and Korea evidence. Calibrated at 81% per the 60/40 blend with the Contract/Construction empirical anchors. Marginally lower than P3 reflecting the reduced volume of CA-specific documentary evidence.
P7 / L7
Component B		 71% Probability that the Component B generalisation of the breach across additional programme cohorts succeeds The most inferential of the nine inputs. Component B extends the core methodology breach argument to additional premium cohorts and programme categories not yet individually documented with specific corrective payment evidence. The legal basis (same contractual obligation, same estimation methodology, same Clause 6.7 refusal to produce records) is sound, but the evidentiary specificity is lower than P3 and P6. Set at 71% — deliberately conservative — reflecting that without production of the underlying premium records (currently withheld under the Clause 6.7 demand), Component B depends on inference from the established breach pattern. An IDO (interrogatory / discovery order) would be expected to increase P7 significantly — up to approximately 0.80.
P8 / L8
Route C / fresh fraud		 78% Probability that Route C (s.249K Crimes Act 1900 (NSW) fresh fraud, commencing 2021) provides an independent limitation extension for the Contract stream Route C is an independent limitation extension for the Contract stream's limitation layer (alongside L4 s.55 and L5 demand accrual). The fresh fraud basis: Aegon's 2021-2025 conduct — renewed written denials of liability, repeated refusals to produce premium records, the illegal NDA conditions, and continued payment at the wrong rate after written demand — constitutes fresh fraud within the limitation period that separately extends time. The six documented fraudulent statements in the Two Sets of Books evidence (D.2.67, D.2.83b false "actual data"; D.2.57, D.2.76b, D.2.85 false denials) are available as evidentiary support. v9.1 fix: L8 was absent from the Boolean win function gates in v9 — corrected so L8 now registers its true marginal contribution in the Shapley decomposition.
P9
Conduct / costs factor		 78% Probability of an indemnity costs order reflecting the defendant's litigation conduct P9 operates exclusively on the costs-adjusted exposure calculation — it does not affect win probability or the core EV. The conduct pattern supporting an indemnity costs order: systematic document withholding, illegal NDA conditions, five irreconcilable positions over 19 years, continued denial after the August 2024 Hallal admission, and the corrective payment made without corresponding liability concession. Research range for indemnity costs applications on comparable conduct: 65–80%. Set at 78% reflecting the strength of the conduct record. Deliberately excluded from SAI and EFI uplifts to prevent double-counting with the conduct facts already embedded in those indices.
The four valuation parameters
Parameter Default Basis and rationale
Discount rate 7% p.a. A conservative commercial discount rate reflecting: the time value of money across the expected litigation period; the opportunity cost of capital tied up in the proceeding; and the risk that even a successful judgment takes time to enforce against an Iowa-domiciled entity. 7% is deliberately conservative — lower than typical litigation funding discount rates (10–15%) to reflect the high win probability and near-certain floor recovery. Move this higher if TLIC enforcement delay risk increases.
Years to trial 1.5 years Derived from FCA empirical data (2023–24 and 2024–25 annual reports; 5-year longitudinal study across 19,483 cases) and NSW Supreme Court Provisional Statistics 2024. The three-branch model: Summary Judgment 30% / 7 months; Negotiated resolution 55% / 9 months; Full trial 15% / 22 months. Weighted mean = approximately 10 months. The 1.5-year base is the midpoint of a realistic range — the branch-weighted PV calculation applies the correct weights across all three branches rather than assuming a single timeline.
DFCR
(defendant capacity risk)		 5% TLIC (NAIC 69078) is an Iowa-domiciled insurer with AM Best rating and Iowa Insurance Division regulatory oversight. At 5%, DFCR is low but not zero — reflecting residual uncertainty about enforcement of an Australian judgment against a US-regulated entity. The Q4 2026 US redomiciliation EGM introduces additional corporate uncertainty. If Aegon's redomiciliation proceeds with material restructuring of the TLIC guarantee chain, increase DFCR. Monitor Iowa Insurance Division filings and AM Best rating actions as leading indicators.
Costs estimate $4.5m A conservative estimate of the plaintiff's total legal costs across the full proceeding. Used exclusively in the costs-adjusted total exposure calculation (the defendant's true economic exposure including an indemnity costs order) and the settlement corridor upper bound. Does not affect win probability, EV, or settlement floor. Adjust upward as actual costs become clearer following Corrs engagement for the full filing program.
The calibration methodology behind every default

No default value was set by intuition. Each was derived through a documented four-step calibration process:

  1. Case-specific legal assessment (60% weight): The starting point for each input is an assessment of the specific facts, evidence, and legal arguments in this proceeding — the corrective payment, the Hallal admission, the five-position collapse, the Clause 20 breach, the Deed succession chain. This analysis drives 60% of each base value.
  2. Empirical dataset calibration (40% weight): 14 verified real anchor cases from official HCA/NSWCA sources were scored across five dimensions (claim construction, accounting/withholding, contract interpretation, limitation/accrual, early disposition). The weighted mean for each dimension anchors 40% of the corresponding input. Purely synthetic comparator rows carry zero weight in the calibration — only verified cases from official sources count.
  3. SAI and EFI uplift (applied after blend): The Structural Advantage Index (SAI = 91.25, compressed to 78.9) and Environmental Factors Index (EFI, half-weight) apply small uplifts to each input. Transmission weights are conservative: SAI × 0.001, EFI × 0.0005 per input. Eight deliberate score reductions were applied to prevent artificial inflation.
  4. Hard caps: P2 Deed is capped at 0.92 (Spencer v Commonwealth summary judgment caution), P4 s.55 at 0.82 (Briginshaw standard caution), and P3 EA at 0.88. No input can exceed its judicial caution ceiling regardless of how strong the underlying facts are.
The most important rule about default values: The defaults reflect the best available calibration at March 2026. They should be changed when counsel provides a materially different assessment of a specific legal issue, when new evidence emerges (particularly if actual premium records are produced under a court order), or when the defendant's position shifts in a material way. They should never be changed simply to produce a more favourable output. The model's value is entirely dependent on the integrity of its inputs. If the inputs are inflated, every output is wrong — and any settlement position built on those outputs is built on sand.
First steps

Getting started

1
Open the model

Navigate to aiooj.com (or open the downloaded index.html file in any modern browser — Chrome, Safari, Firefox, or Edge all work). No login, no installation, no internet connection required after the page loads.

The model loads with all inputs at their base values — the calibrated defaults derived from the case law benchmark set, the 100-case empirical dataset, and the specific facts of this proceeding. These are the starting point for every session.

2
Understand what updates automatically

Every output panel — all eight KPI cards, the claim path list, the settlement band chart, the P2×P4 heat map, the Shapley ranking, the stress floor comparison, and the settlement corridor — updates instantly whenever any slider is moved. There is no submit button. There is no recalculate. The model recomputes in real time on every input change.

The heat map base-case cell (outlined in gold) tracks your current P2 and P4 values as you move the sliders — it always shows you where you are on the stress grid.

3
Reset at any time

Click Reset in the top-right corner to return all inputs to their calibrated base values. There is no undo — Reset is immediate and complete. If you want to preserve a scenario before resetting, use Export summary first to capture the current state.

Interface

Interface layout

Input panel
P2 — Deed / SJ success89%
P4 — s.55 postponement80%
Discount rate7%
Overall win probability
99.1%
Expected recovery
$91.2m
Settlement floor P10
$83.4m
Branch-weighted PV
$86.1m

The sidebar (left) contains all inputs. The main area (right) contains all outputs. Move any slider and the entire right panel updates immediately.

Input panel

Using the sliders

The sidebar is divided into four groups: core legal probabilities (P1–P9), valuation parameters, and scenario weights.

Core legal probabilities
P1 through P9
Each slider represents one discrete legal issue. The value shown is the probability that issue succeeds. Base values are the calibrated defaults — move them only when you have a specific reason to deviate based on new legal analysis or counsel input. The three load-bearing variables are P2 (Deed), P4 (s.55), and P3 (EA methodology).
Valuation parameters
Discount rate, years to trial, DFCR, costs
The discount rate (default 7%) and years to trial (default 1.5) drive the present value calculations. DFCR is the defendant financial capacity risk — 5% at base. The costs estimate drives the indemnity costs overlay. These parameters affect PV outputs but not win probability.
Scenario weights
LS-1 through LS-4
The four scenario weights determine the distribution of recovery values when the plaintiff wins: LS-1 full claim (25%), LS-2 Deed-led full (50%), LS-3 truncated Deed at $75m (15%), LS-4 contract-only at $47m (10%). Adjusting these changes expected recovery without changing win probability.
Precision note
All sliders are integers
Probability sliders move in 1% increments. The valuation sliders (discount rate, DFCR) also move in 1% increments. Years to trial moves in 0.1 year increments. Costs estimate moves in $0.5m increments. The displayed value always shows the current setting.
When to change the base values: The base values represent the best available calibration at March 2026. Change them when: (a) counsel provides a materially different assessment of a specific legal issue; (b) new evidence emerges (e.g. premium records produced under a court order); (c) the defendant's position shifts in a way that affects a specific probability; or (d) you are running a deliberate stress scenario. Do not change them to make the outputs look better.
Dashboard metrics

Reading the KPI panel

KPIWhat it means and how to use it
Overall win probabilityThe probability of any material recovery above $0.01m. This will be near-certain at base values — this is correct and structural, not inflated. See the Foundations page for the full explanation. The practically important question is the recovery distribution, not this number.
Expected recovery (EV)Probability-weighted average recovery across all outcomes including nil recovery. This is the decision-theoretically correct number for a risk-neutral decision-maker. A rational defendant cannot offer materially less than this without specific contrary arguments not captured in the model.
Settlement resistance floor (P10)The portfolio-weighted beta P10 — the level below which only 10% of winning-scenario outcomes fall. This is the plaintiff's methodologically defensible walk-away threshold. Accepting below P10 means accepting a worse result than 90% of realistic trial outcomes in winning scenarios. Do not accept less than this in mediation.
Branch-weighted PVPresent value across the three disposition branches: Summary Judgment (30% weight, 7 months), Negotiated (55%, 9 months), Full Trial (15%, 22 months). This is higher than a simple 1.5-year point-estimate PV because the SJ branch pulls the distribution forward. Use this number in time-value settlement arguments.
Costs-adjusted total exposureExpected recovery plus the probability-weighted expected indemnity costs order (P9 × costs estimate). This is the defendant's true total economic exposure. Present this number to the defendant in negotiations — not EV alone.
DFCR-adjusted EVExpected recovery discounted by the Defendant Financial Capacity Risk. At base (5%), the discount is modest. If Aegon's Q4 2026 redomiciliation raises corporate uncertainty, increase DFCR and watch this number fall — that reduction represents real collection risk to be priced into any settlement.
Limitation double-failure PP(s.55 AND demand accrual both fail simultaneously). At base values this is very small. If this number rises materially above 5%, it signals that the limitation position needs reinforcement — the full historical claim period is at risk.
Near-full recovery P (>$55m)The probability that any recovery exceeds $55m. This is driven almost entirely by the Deed ultimate probability. It tells the defendant what probability they face of paying close to the full claim — this is the number that creates settlement urgency.
Claim paths

Claim path probabilities

Seven paths are shown with colour-coded indicators: green (>80%) means strong and should be treated as a near-certain feature in settlement discussions; amber (65–80%) is favourable but carries material uncertainty; red (<65%) is contested and should be treated as upside rather than base case.

Reading the path list together: The Deed ultimate probability (L2 OR L5) is always higher than either L2 or L5 individually — this is correct, because two independent routes means either one failing does not defeat the stream. The Overall win probability is always higher than any single path — the same OR-gate logic. These relationships should be internally consistent: if they are not, check whether the sliders have been moved to inconsistent positions.
Recovery distribution

Settlement band distribution

Five recovery bands show the probability mass distribution. The near-full band (>$55m) should contain the vast majority of probability at base values — this is driven by the Deed ultimate probability.

Near-full >$55m
The dominant outcome at base values. Driven by Deed ultimate (L2 OR L5). If this band drops below 85%, check P2 and P5 — one of the Deed routes has weakened significantly.
High $35–$55m
Truncated Deed scenario. Should be a small band at base values — Deed failure at full quantum but some recovery. Grows if you reduce P2 and P5 together.
Mid $15–$35m
Contract-only recovery without Deed. Grows if you stress Deed significantly while keeping Contract strong. Should remain modest at base values.
Korea <$15m
Korea floor scenario — all other streams fail. Should be a very small probability. If this grows materially, both Deed and Contract streams have been stressed below realistic levels.
Nil
All three streams fail simultaneously. At base values this is less than 0.007% — mathematically negligible. Grows only if you stress all inputs simultaneously to unrealistically low levels.
Bar length
Bars are relative to the highest-probability band, not to 100%. This makes the distribution shape visible. Check the percentage figure on the right of each bar for the actual probability.
Stress analysis

P2×P4 stress heat map

The heat map simultaneously varies P2 (Deed/SJ success, rows) and P4 (s.55 concealment, columns) across a 7×7 grid. This is the genuine stress diagonal for this case — the scenario the defendant's legal team will target: simultaneous Deed underperformance and limitation failure.

How to read it: The EV $m tab shows a genuine gradient across the grid — P2 controls row variation (Deed stream strength) and P4 controls column variation (whether the full historical quantum survives via s.55 limitation extension). Find the gold-outlined cell — that is your current P2/P4 position. The cells to the lower-left show what happens if both weaken simultaneously. Even at the most adverse corner (P2=60%, P4=55%), material recovery persists because Korea (L1=90%) and demand accrual (L5=85%) continue to operate as independent floor anchors. This is the quantitative demonstration of the structural asymmetry argument.
Driver ranking

Shapley driver ranking

The Shapley decomposition ranks all eight L-variables by their marginal contribution to overall win probability. Each bar shows how much overall win probability would decrease if that variable were removed entirely — the larger the bar, the more load-bearing that variable is.

Expected ranking at base values: L2 (Deed/SJ) typically ranks first — its removal collapses the primary path. L4 (s.55) typically ranks second — it is the enabling condition for the full historical quantum. L3 (EA methodology) and L5 (demand accrual) typically rank third and fourth. L7 (Component B) and L8 (Route C) rank lowest as the most speculative variables. If the ranking looks different, check whether any input has been moved to an extreme value.
v9.1 fix: L8 (Route C) previously showed zero contribution due to a formula defect — it was absent from the Boolean win function gates. This has been corrected. L8 is now wired as an independent OR gate in the Contract stream's limitation layer and shows a genuine marginal contribution. Any reading of zero for L8 in a prior version should be disregarded.
Adversarial scenarios

Stress floor panel

The stress floor panel shows five reference points in a single view, ordered from the most optimistic (base EV) to the most conservative (Korea minimum).

Reference pointDescription and use
Base EVExpected recovery at your current slider inputs. The anchor number — all other stress figures are shown relative to this.
Quantum P10 floorThe settlement resistance minimum — the level below which only 10% of winning-scenario outcomes fall. This is the plaintiff's walk-away threshold. The bar length relative to base EV shows how much downside cushion the P10 floor provides.
Scenario A — stressed, OR gatesAll eight inputs moved to adverse stress values simultaneously (P2=0.70, P3=0.70, P4=0.60, P5=0.65, P6=0.65, P7=0.50, P1=0.85, P8=0.65), while preserving OR gates. This reflects legal reality — the three recovery streams remain independent. Use this to answer "what if everything goes wrong at once but the legal structure holds?"
Scenario B — AND gateCorrelated collapse: all three major paths must succeed simultaneously (AND gate). This is the extreme lower bound and the most conservative possible assumption. In practice the legal structure is OR-gated, so this scenario overstates the downside. Use it to show the defendant the absolute floor — even under AND-gate correlated collapse, the plaintiff retains meaningful recovery.
Korea minimumKorea alone at stress probability (0.85 × $13.7m). The irreducible floor — the recovery the plaintiff receives even if both Deed and Contract streams fail completely. This number should anchor any settlement discussion as the minimum the defendant should expect to pay.
Negotiation framework

Settlement corridor

The settlement corridor shows the rational negotiation range — from the plaintiff's PV floor to the defendant's costs-adjusted PV ceiling — overlaid on the full $100.7m claim as a visual reference.

How to use the corridor in negotiations: The lower bound (green) is the plaintiff's rational minimum — accepting less than PV of EV means accepting worse than the discounted expected outcome. The upper bound (gold) is the defendant's rational ceiling — above this, the defendant is better off at trial on a net-present-value basis. A rational settlement lives somewhere between these two numbers. The gap between them is the negotiation space.
The Q4 2026 EGM deadline: Aegon's planned US redomiciliation EGM creates a hard settlement horizon. Undisclosed material litigation of this magnitude triggers SEC 10-K and GAAP disclosure obligations as the EGM approaches. The defendant's incentive to settle increases significantly in the months before Q4 2026. All litigation and regulatory escalation steps should be timed to maximise pressure at this deadline.
Absolute worst case

Joint failure analysis panel

The joint failure analysis panel sits below the settlement corridor. It answers the question every funder and every mediator asks: "What is your absolute worst case, and how likely is it?"

The answer: ~$35m at a probability of 0.027% (approximately 1 in 3,700). For the worst case to materialise, six independently established legal positions — each supported by mapped High Court and Court of Appeal authority — must all fail simultaneously. The panel documents each event, its generous failure probability (10–20%), and the step-by-step joint probability calculation.
Why this panel exists

The worst case analysis does not threaten the plaintiff — it protects the settlement position. Without a documented worst case, the defendant can assert that "everything might fail" and use uncertainty as leverage. With a documented worst case, that assertion has a precise answer: 0.027% probability, $35m floor. The panel converts every defendant argument from a threat into a quantified improbability with a documented floor.

The individual failure probabilities have been set generously at 10–20% — substantially higher than the case-specific analysis alone would justify. This is deliberate. A worst-case estimate that a sceptical counsel or funder might challenge as optimistic is useless. A worst-case estimate derived from generous inputs that produces a floor of $35m is unassailable.

The 12-year floor quantum

The floor is not zero. The Deed Limb 1 primary indemnity — "shall indemnify Harrison against any loss, expense or liability" — cannot fail on these facts. The worst case is therefore: Deed liability confirmed, all limitation arguments fail, quantum limited to the 12-year merits period. On that basis: EA principal ~$28-32m + CA ~$8-10m + Korea Component A ~$13.7m + interest ~$12-15m = approximately $55-65m (midpoint $35m).

$35m is 34.7% of the full $100.7m claim. A defendant who argues total failure on every limitation and fraud argument — a 0.027% probability event — still faces a judgment of approximately $35m. No rational settlement offer can be below $35m.

How to use this in mediation: When the defendant argues that limitation arguments are uncertain and the fraud evidence is disputed, open this panel and walk through the six events. Ask the defendant to identify which of the six they are confident of winning. The arithmetic is transparent — every event they concede reduces the joint failure probability by the factor of their concession. By the time they have conceded four of the six, the remaining joint probability is so small that the floor becomes the practical outcome.

For the full legal analysis behind each of the six events and the complete derivation of the $35m floor quantum, see the Absolute worst case section of the P1–P9 Rationale page.

Documentation

Exporting a summary

The Export summary button (top right) opens a formatted print page showing all current inputs and all key output metrics.

Analytical use

Scenario analysis techniques

A
Counsel assessment scenario

Before a counsel briefing, ask counsel to assess P2 (Deed), P4 (s.55), and P3 (EA methodology) independently. Enter their assessments into the sliders. Export the resulting summary. This documents counsel's probability inputs alongside the model's outputs — creating a defensible record of the probability basis for settlement positioning.

B
IDO (discovery order) scenario

If a production order for the Aegon premium records is obtained, increase P3 by approximately +8pp (records confirm EA methodology), P6 by +6pp (same benefit flows to CA stream), P7 by +8pp (Component B generalisation becomes directly testable), and reduce P4 by approximately -6pp (concealment argument partially weakened — though prior refusal remains evidential). The net effect of IDO is strongly positive. Export both pre-IDO and post-IDO scenarios to document the uplift.

C
Defendant's worst-case scenario

To build the defendant's negotiating position for mediation preparation, stress all inputs toward their Low boundary simultaneously. The Stress Floor panel already shows Scenario A (stressed, OR gates) and Scenario B (AND gate). Use Scenario B as the floor the defendant is arguing from — then demonstrate using the heat map and path list that even Scenario B supports material recovery, narrowing the negotiating space significantly.

D
Timing sensitivity analysis

Vary the years to trial and discount rate sliders to show the time-value cost of delay. At 7% discount rate, each additional year of delay reduces the PV of a $91m EV by approximately $6m. Use this in negotiations to demonstrate that the defendant's "wait and see" strategy has a quantifiable cost — and that settlement now is more rational than settlement in 18 months at the same nominal value.

Important caveats

What not to do

Reference

Quick reference

Key actions
Move any slider
All outputs update instantly. No submit required.
Reset button
Returns all inputs to calibrated base values. Immediate and complete — export first if you want to preserve the current scenario.
Export summary
Opens a print-formatted page with all current inputs and outputs. Print to PDF to save. Includes timestamp and confidentiality notice.
EV $m / Win % toggle
Switches the P2×P4 heat map between expected value and win probability views. Located above the heat map table.
Gold-outlined cell
Your current P2/P4 position on the heat map. Tracks in real time as you move those sliders.
The three numbers that matter most in mediation
Settlement floor P10
The plaintiff's walk-away threshold. Do not accept below this. Located in the KPI panel (third card).
Costs-adjusted exposure
The defendant's true total economic exposure. Present this, not EV alone. Located in the KPI panel (fifth card).
Rational corridor
PV floor to costs-adjusted PV ceiling — the range where both parties have rational incentive to settle. Located in the corridor panel at the bottom of the page.
What each colour means
Green indicator
Probability above 80% — strong, should be treated as near-certain in settlement discussions.
Amber indicator
Probability 65–80% — favourable but carries material uncertainty. Present with stated range.
Red indicator
Probability below 65% — contested. Treat as upside, not base case in settlement discussions.
Gold outline
In the heat map: your current base case position. Tracks P2 and P4 inputs in real time.
AIOOJ User Guide — Model v9.1 — March 2026 — Confidential — Not legal advice