Iran-Trump War April 14 — Retrotrack
Ceasefire announced April 8-9. Iran agreed to reopen Hormuz. Trump suspended bombing. Power plants NOT struck. Brier: 0.213 — acceptable. Direction wrong, but partial hits improved score.
Result
| Event | Date | Status | Details |
|---|---|---|---|
| Ceasefire | April 8-9 | Yes | Two-week pause |
| Hormuz | Conditional | Pending | Iran to reopen |
| Power plants | — | No | Threatened but NOT hit |
| Gulf retaliation | April 3 | Yes | Iran hit refineries |
Predicted vs Actual
| Scenario | Predicted | Actual |
|---|---|---|
| Continued Escalation (power plants) | 35% | No |
| Trump Declares Victory | 30% | Partial |
| Iran Major Retaliation | 20% | Partial |
| Negotiated Pause | 10% | Yes |
| Stalemate | 5% | No |
Direction: Escalation (35%) predicted most likely. Negotiated Pause (10%) happened. Wrong.
Scoring
| Metric | Value | Assessment |
|---|---|---|
| Brier Score | 0.213 | Acceptable (<0.25) |
| Log Loss | 0.892 | Poor |
| Direction | Wrong | Escalation favored, Pause happened |
| Calibration | Mixed | Partial hits improved score |
| Hit Rate | 60% | 3/5 predictions |
Brier calculation: Partial outcomes (Trump Victory 0.5, Iran Retaliation 0.5) reduce error vs binary scoring.
D/E/M/A Delta
How methodology metrics shifted from prediction (T0) to resolution (T1):
| Metric | T0 (Apr 7) | T1 (Apr 14) | Δ | Interpretation |
|---|---|---|---|---|
| D — Data | 0.75 | 0.85 | +0.10 | More sources confirmed ceasefire details |
| E — Epistemology | 0.50 | 0.45 | -0.05 | Base rates overweighted escalation |
| M — Model | 0.55 | 0.40 | -0.15 | Diplomatic exits underweighted |
| A — Aleatoric | 0.70 | 0.80 | +0.10 | Ceasefire "black swan" materialized |
Methodology Notes
D/E/M/A v2.3 — ScenarioAtlas framework layers:
- D (Data) — Source quality, evidence freshness, bias assessment
- E (Epistemology) — Base rates, analogous cases, known unknowns
- M (Model) — Scenarios with drivers and blockers
- A (Aleatoric) — Named black swans, tail risk signals
What changed: Model (M) dropped most — we overweighted escalation scenarios (35%) and underweighted diplomatic exits (10%). Only China from the China/Pakistan mediation channel was in our model in the Cascade Trigger (Black Swans) part. Lesson: when both sides face catastrophic costs, off-ramps are more likely than base rates suggest.
What Changed
| Actor | Layer | Before | After |
|---|---|---|---|
| Iran | Behavior | "Restraint is over" | Accepts face-saving exit under existential threat |
| Iran | Current situation | Kharg operational, bridges intact | Kharg damaged, bridges destroyed |
| Trump | Behavior | Deadline = credibility test | Extensions = negotiation signal |
| US-Iran | Behavior | Escalation ladder unclear | Power plants = implicit ceiling |
| China/Pakistan | Behavior | Pakistan proposal ignored | Pakistan = China proxy, mediation worked, ceasefire 48h |
World Revealed
1. Escalation Cost Symmetry
Both sides face catastrophic costs → diplomatic exits more likely than base rates.
2. Trump Deadline Pattern
Repeated extensions (March 23 → April 7) signal deal preference, not weakness.
3. War Crime Threshold
Civilian infrastructure (power plants) = implicit ceiling. Threatened but never crossed.
4. China/Pakistan Channel
Pakistan = China proxy in region. Mediation channel delivered ceasefire in 48 hours.
Claims Added
| Actor | Level | Claim |
|---|---|---|
| Trump | Behavior | Deadline extensions signal deal preference |
| Iran | Behavior | Accepts face-saving exit under existential infrastructure threat |
| US Military | Behavior | Civilian infrastructure = implicit red line |
| China/Pakistan | Behavior | Pakistan = China proxy, mediation channel works |
Resolution Tracking
| Prediction | Deadline | Our Call | Outcome | Correct? |
|---|---|---|---|---|
| Power plants struck | Apr 14 | 35% | No | Yes |
| Iran retaliates Gulf | Apr 14 | 20% | Partial | Partial |
| Ceasefire announced | Apr 14 | 10% | Yes | Underconfident |
| Hormuz blocked | Apr 14 | 85% | Yes | Yes |
| Trump victory speech | Apr 14 | 30% | Partial | Partial |
Verdict
Direction wrong. Brier 0.213 acceptable. Escalation predicted, ceasefire happened. But partial hits (Trump suspended bombing, Iran hit Gulf) improved score.
Key insight: Escalation has implicit ceilings. China/Pakistan channel delivered. Trump deadline pattern confirmed: extensions = deal signal.
Frequently Asked Questions
What was the outcome of the Iran-Trump war by April 14, 2026?
Two-week ceasefire announced April 8-9. Iran agreed to reopen Hormuz. Trump suspended bombing. Power plants NOT struck.
How accurate was the ScenarioAtlas prediction?
Brier 0.213 — acceptable. Direction wrong but partial hits on Trump Victory and Iran Retaliation improved score.
What did reality reveal?
Trump deadline extensions signal deal preference. Iran accepts face-saving exits under existential threat. China/Pakistan mediation channel works.
Every prediction scored. Every miss published.