Methodological exercise by the St. Gallen Endowment's technology team. Not an institutional assessment. Each probability answers: how likely is this scenario to be the dominant crisis trajectory four weeks from the assessment date? They express relative confidence, not precise forecasts. Treat them as a ranking with approximate magnitudes.
Eight mutually exclusive scenarios used by the monitoring system. Scenarios are mutually exclusive at any given assessment point but can transition into one another over time. Probabilities shown are from the latest assessment.
4-week outlook
A diplomatic settlement halts military operations before either side achieves decisive military objectives. Requires a credible Iranian interlocutor, willingness to negotiate on both sides, a mediator (Oman, Turkey, or China), and a face-saving formula for both leaderships. Historical analogues: the Kosovo air campaign ending through Russian-mediated diplomacy.
4-week outlook
Trump declares that US military objectives have been achieved and withdraws American forces. He frames whatever destruction has been inflicted as a decisive victory. Israel continues operating alone but with sharply reduced capability. The war does not end; it transforms into a different war. This scenario requires no Iranian cooperation, no mediator, no mutual agreement — just a single decision by a single person based on domestic political calculus.
4-week outlook
Iran capitulates under military pressure within the first two weeks. The regime accepts US/Israeli terms: permanent end to enrichment, ballistic missile limits, and cessation of proxy support. This requires both a collapse of Iranian military capability and a political decision to surrender, which historically almost never happens to regimes that retain any capacity to fight.
4-week outlook
A sustained US/Israeli air campaign systematically degrades Iran's military infrastructure over 3 to 5 weeks, reducing Iranian retaliatory capacity. Success requires Iranian fire rate to decline measurably week over week and no major escalation beyond what air power can address.
4-week outlook
Neither side achieves decisive results. Iran sustains retaliatory fires at a reduced but persistent rate. The Strait of Hormuz remains contested. Oil prices stay elevated. The conflict becomes a war of economic and military attrition, ending when domestic political costs force one or both sides to seek an exit. Historical analogues: the Iran-Iraq Tanker War (1984–1988).
4-week outlook
The conflict expands to include active military engagement by actors beyond the initial belligerents. Qualifying events include Houthi activation in the Red Sea, Gulf state transition to offensive operations against Iran, Hezbollah opening a sustained second front, or third-party military intervention.
4-week outlook
Internal divisions within the Iranian state, exacerbated by leadership decapitation, lead to institutional fragmentation. The IRGC, civilian government, and clerical establishment pursue contradictory strategies.
4-week outlook
Iran, under existential threat, attempts to rapidly assemble and potentially demonstrate a nuclear device using pre-positioned fissile material that survived both campaigns. Lowest probability, highest consequence. This scenario models Iranian nuclear decision-making under existential pressure, not a US nuclear strike. The risk pathway runs through desperation, pre-delegation, and miscommunication, not through deliberate US escalation.
Scenario probabilities for nuclear outcomes are inherently unreliable and should not be used for operational planning. This analysis identifies structural risk factors, not predictive estimates.
The monitoring system employs 14 specialised analytical agents, each grounded in a specific theoretical framework from international relations, security studies, and decision science. Every agent assesses the same evidence independently — no agent sees another's output before producing its own assessment. Disagreements are surfaced, not suppressed.
These agents apply analytical disciplines to the raw intelligence, independent of regional expertise. They compete rather than coordinate: the scenario planner separates predetermined forces from genuine uncertainty; the forecaster enforces base-rate discipline; the tail-risk auditor guards against probability compression; the red team constructs the strongest possible counter-case.
Predetermined elements vs. genuine uncertainties
Separates forces already in motion from genuine uncertainties. Tests whether each scenario is internally consistent and structurally distinct. Detects when predetermined elements shift, rendering scenarios implausible. Filters noise from signal by clarifying what is actually uncertain versus what is already determined.
Coercive bargaining, commitments, and focal points
Maps the bargaining structure: each side's best alternative to agreement, credible threats, and audience costs. Identifies natural coordination boundaries where settlement could occur. Tests whether either side can credibly commit to a deal. Distinguishes between the capability to threaten and the credibility of the threat.
Base rates, Bayesian reasoning, and calibration
Applies historical base rates before considering case-specific evidence. Detects scope insensitivity — when a single dramatic event receives disproportionate probability weight. Requires explicit prior-to-posterior reasoning chains. Maps what observable evidence would force significant probability shifts.
Fat tails, fragility, and extreme events
Audits whether the probability distribution itself is wrong, not just the point estimates. Identifies fragile elements sensitive to small perturbations. Detects narrative fallacy — finding coherent stories in noisy data. Guards against overconfidence in probability compression of extreme scenarios.
Counter-case construction and disconfirmation
Constructs the strongest possible case against the consensus scenario. Applies Analysis of Competing Hypotheses, prioritising evidence that is inconsistent with a scenario over evidence that merely confirms it. Performs pre-mortem analysis: if the consensus is catastrophically wrong in two weeks, what will we wish we had noticed?
Structured evidence-hypothesis matrix
Maintains a matrix of eight scenarios against all new evidence items. Counts inconsistencies per scenario — the scenario with fewest inconsistencies is better supported. Performs sensitivity analysis: remove one evidence item and check if the ranking changes. Identifies what diagnostic evidence is missing.
How decision-makers misread intentions
Diagnoses which actors are operating under spiral-model versus deterrence-model assumptions. Identifies motivated biases — what each side wants to see. Flags signal confusion: deliberate communications misread as unintentional indicators of intent. Tests whether the same historical analogy is dominating multiple actors' perceptions.
Escalation that happens without anyone intending it
Maps entanglement: dual-use systems, co-located nuclear and conventional assets, shared command-and-control infrastructure. Identifies civil-military control gaps and pre-delegated authority. Detects use-it-or-lose-it pressure on strategic assets. Models the fog-of-war risks where misidentification in fast-moving operations could trigger unplanned escalation.
Position on escalation ladder and threshold-crossing
Maps the current crisis position to specific rungs on Kahn's 44-rung escalation ladder, with six major thresholds. Assesses which side has escalation dominance at the current rung. Identifies de-escalation pathways, both bilateral and unilateral. Detects rung-skipping risk — events that could jump multiple levels at once.
These agents apply country- and sector-specific expertise combined with international relations theory. They prevent "generic IR" analysis by insisting on institutional detail: the IRGC is not a monolith, Gulf states are not a bloc, nuclear decisions are driven by domestic coalition type as much as external threat.
IRGC as institutional actor with its own interests
Disaggregates "Iran" into institutional components: IRGC, Supreme Leader, civilian government, and clerical establishment. Maps IRGC institutional interests including revolutionary ideology, economic rents from sanctions, and political dominance. Detects when IRGC commanders act beyond the Supreme Leader's authorized positions. Analyzes proxy network autonomy.
What military force can and cannot achieve
Assesses what the US and Israel can actually do militarily — not what policymakers wish, but what military force can deliver at what cost. Identifies binding constraints: sortie rates, munitions stocks, ISR coverage, basing access. Analyzes the day-after problem: what does the political landscape look like after a successful strike? Detects gaps between political objectives and military capability.
Gulf monarchies and regime security logic
Gulf monarchies prioritise regime survival over external threat balancing — explaining their persistent under-balancing against Iran. Treats Gulf states individually: Saudi Arabia, the UAE, Qatar, Oman, Bahrain, and Kuwait are not a bloc. Uses oil production decisions as alignment signals. Tracks the divergent trajectories of normalisation, mediation offers, and hedging behaviour.
Domestic political economy of nuclear decisions
Nuclear decisions are driven by ruling coalition type, not external security threats alone. Autarkic "backlash" coalitions pursue weapons; trade-dependent "outward-looking" coalitions forgo them. Assesses the sanctions paradox: sanctions strengthen backlash actors by destroying their outward-looking competitors. Tracks private-sector strength as a proxy for coalition type.
Nuclear posture types and crisis dynamics
Not all nuclear arsenals are alike. States adopt distinct postures — catalytic, assured retaliation, or asymmetric escalation — each creating different crisis dynamics. Detects posture shifts under crisis pressure, particularly the dangerous transition from catalytic to asymmetric escalation when conventional forces collapse. A small, untested, vulnerable arsenal cannot support assured retaliation.
After all 14 perspective agents have completed their independent assessments, a peripheral scanner examines intelligence that no agent cited. Using Ansoff's weak-signal detection framework, it classifies overlooked signals by strength, maintains a persistent watch list across cycles, and identifies cross-domain convergence patterns that the primary agents' frameworks may structurally miss.
This system produces three distinct output formats — regular scenario assessments, issue consultations, and strategic deliberations. Each is generated through a different process with its own structure, scoring method, and quality controls. Select one below to see how it works.
How Scenario Assessments Are Generated
Regular update-cycle assessments are the core product of the system. They track what is happening — updating scenario probabilities as new developments emerge. Each update follows a six-phase process. The orchestrator executes the entire workflow mechanically, enforcing quality gates at each transition.
The system searches for new developments against a predefined set of watch variables. Every piece of intelligence must include a verifiable source URL and publication date. High-impact claims require the system to fetch and read the actual source page — search snippets are insufficient. A minimum of three distinct new developments is required to proceed.
All 14 perspective agents receive the raw intelligence and current probabilities. Each agent is spawned in complete isolation — agent N does not see agent N-1's output. This independence is enforced architecturally, not by instruction. Each agent produces a structured assessment identifying scenario adjustments with named causal mechanisms.
After all perspective agents complete, a peripheral scanner examines intelligence that no agent cited. This catches signals that fall outside or between the primary agents' theoretical frameworks.
All 14 agents' scenario adjustments are equal-weighted and averaged. Adjustments are converted to percentage points using a predefined magnitude scale, applied to current probabilities, and renormalised so all scenarios sum to 93–95% — the remainder is an explicit unspecified tail-risk reserve. Structured disagreements (where agents diverge by more than 15 percentage points on any scenario) are flagged and explained in the briefing.
A structured briefing is generated with key developments, the updated probability distribution, watch-variable status updates, and a divergence summary where agents disagreed. This briefing is what appears on the Assessments page.
All outputs — raw intelligence, perspective assessments, aggregated probabilities, briefing, and updated charts — are version-controlled and pushed. The entire analytical trail is preserved for auditability.
Each perspective agent is spawned with only the raw intelligence and current probabilities. No agent receives another agent's output. Before aggregation, the system verifies that no assessment references another agent's reasoning. Contaminated outputs are discarded. This prevents herding and ensures genuine diversity of analytical judgement.
Each agent independently proposes adjustments to scenario probabilities. These adjustments are equal-weighted and averaged (simple mean), and the resulting shift is applied to the previous cycle's probabilities. There is no Bayesian updating or weighted scoring — it is a democratic average with mandatory adversarial challenge from the red team agent.
The aggregate tracks disagreement ranges (min/max across agents per scenario) to surface where consensus is weak. Structured disagreements — where agents diverge by more than 15 percentage points on any scenario — are flagged and explained in the briefing.
During the calibration phase (the first 10+ cycles with resolvable outcomes), all agents receive equal weight. After sufficient calibration data accumulates, the system will transition to track-record-weighted averaging using Brier scores — agents with better forecasting records receive higher influence.
The system enforces several constraints that distinguish it from unstructured analysis:
Every probability adjustment must cite at least one named causal mechanism from the relevant theoretical framework — Fearon on commitment problems, Schelling on focal points, Posen on inadvertent escalation, and so on. "Tensions increased" is not analysis; the system requires the causal process connecting an event to a scenario shift.
Scenario probabilities must sum to 93–95%, not 100%. The remainder is an explicit tail-risk reserve for outcomes the scenario framework does not cover. Every update follows a Bayesian chain: prior, evidence, likelihood ratio, posterior. Adjustments exceeding moderate magnitude require explicit justification.
Every piece of intelligence entering the system must include a verifiable source URL and publication date. Claims without sources are excluded. High-impact claims — those that could significantly move probabilities — require the system to fetch and read the actual source page, not merely rely on search snippets.
When perspective agents diverge significantly on any scenario, the briefing explicitly identifies which agents disagree, what analytical difference drives the divergence, and what evidence or mechanism each side is weighting. Disagreement is a feature of the system, not a failure.
How Issue Consultations Are Generated
In addition to regular update-cycle briefings, the system can be posed ad-hoc analytical questions — "consultations." These are bounded questions that arise during the conflict and require the full analytical panel to respond.
A specific analytical question is formulated — for example, "Why is the nuclear escalation scenario judged with double-digit probability?" or "What strategy is Iran pursuing?" The question must be bounded and answerable within the analytical framework.
All 14 perspective agents receive the question along with the current evidence base. Each agent answers through its specific analytical lens — the same independence enforcement as the update cycle applies. No agent sees another's response.
The 14 responses are synthesised into a single consultation document. The synthesis preserves disagreements and highlights where analytical lenses diverge on the question. Consultations produce scenario implications (directional, not quantified), feasibility assessments, failure-mode analysis, and adversary-response predictions.
How Strategic Deliberations Are Generated
While update-cycle assessments track what is happening and consultations answer bounded analytical questions, deliberations address a different kind of question: what should each actor do? A deliberation maps the full option space for multiple actors, scores every option on two independent dimensions, and identifies where the most likely course of action diverges from the most strategically valuable one. The analytical value lies in separating prediction from prescription.
All 14 perspective agents receive the same evidence base and the deliberation question. Each agent works in complete isolation, reconstructing each actor's decision logic from the evidence and then enumerating every option available to that actor. For each option, the agent assigns two independent scores: Likelihood (how probable is it that the actor takes this action?) and Strategic Value (how much does this action advance the actor's objectives?). An option can score high on strategic value but low on likelihood — meaning it is the right move but the actor is constrained from taking it.
A deterministic algorithm merges options that multiple agents identified under different labels into a single canonical list. Options identified by only one or two agents are flagged as "orphans" — potential blind spots or implausible ideas that require explicit attention in the next stage.
Each agent receives the full harmonised option universe alongside its own initial output. It must score every option, including those it did not originally identify. Three structural mechanisms prevent artificial convergence. First, any score change greater than 15 percentage points requires the agent to name the specific evidence that caused the revision. Second, any score left unchanged requires the agent to state the strongest contrary argument and explain why it was not persuasive. Third, every orphan option must receive an explicit endorsement or rejection with reasoning.
For each option, the system computes statistical summaries (mean, median, interquartile range) for both Likelihood and Strategic Value across all 14 agents. Options are ranked by strategic value. The system identifies "likely mistakes" — options that score high on likelihood but low on strategic value — and "missed opportunities" where the reverse is true. The synthesis translates these findings into a reader-facing analysis.
| Assessment | Consultation | Deliberation | |
|---|---|---|---|
| Question | What changed? | Bounded ("Is scenario X underpriced?") | Open-ended ("What are the next best moves?") |
| Trigger | Scheduled cycle (2×/day) | Ad-hoc | Ad-hoc |
| Rounds | 1 | 1 | 2 (with harmonisation between rounds) |
| Score dimensions | Probability shift per scenario | Directional effect on scenarios | Dual: Likelihood + Strategic Value (0–100) |
| Anti-herding | Architectural isolation | Not needed (single round) | 3-point structural enforcement |
| Probability impact | Direct (updates distribution) | Indirect (informs next cycle) | None (analytical exercise only) |
What the system reads before it thinks. Three update cycles per day, each running a structured intelligence-gathering pipeline before any analytical work begins.
Each cycle processes information through four layers of filtering. The volume narrows sharply at each stage.
| Layer | Per cycle | Per day | Per week |
|---|---|---|---|
| Search queries fired | ~38 | ~115 | ~800 |
| Search results scanned | 200–400 | 600–1,200 | 4,200–8,400 |
| Pages read in full | 7–15 | 20–40 | 140–300 |
| Items entering the intelligence file | 25–30 | 70–90 | ~580 |
The final layer — roughly 70–90 curated items per day — is what the 14 analytical perspectives actually assess. Everything above it is discarded as duplicate, irrelevant, undated, or already known from a previous cycle.
Eight domains queried every cycle. Zero-result domains are logged explicitly — absence of signal is itself informative.
CENTCOM operations, air defence, strike reports
Enrichment status, facility damage, radiation monitoring
Ceasefire talks, mediation, UN Security Council
Houthi activity, Iraq militias, Hezbollah, Gulf states
Hormuz shipping, tanker insurance, VLCC rates
IRGC dynamics, protests, succession politics
Presidential rhetoric, war powers, gasoline prices
Brent crude, sanctions, futures curves
The eight domains above are permanent. But the system does not stop there. A peripheral scanner runs once daily, reviews what was collected versus what was missed, and generates additional targeted queries to close surveillance gaps. These queries are added and retired as the conflict evolves — the search surface adapts to the war, not the other way around.
Currently 11 dynamic queries are active, tracking emerging signals that the permanent domains were not designed to catch: Kharg Island ground operation planning, Gulf desalination infrastructure threats, cyber operations against energy systems, and others.
The peripheral scanner classifies each gap using an Ansoff matrix (known vs. unknown sources and signals) and prioritises queries that would detect weak signals before they become headline events.
Nine specialist publications monitored directly every cycle, fetched by URL rather than discovered through search. Content fingerprinting detects when a source has published new material since the previous cycle.
| Source | Focus | Frequency |
|---|---|---|
| Institute for Science and International Security | Nuclear programme analysis | Event-driven |
| FDD Long War Journal | Proxy and militia operations | Daily |
| Windward Maritime AI | Hormuz transits, vessel rerouting | Daily |
| Kpler | Crude and LNG commodity flows | Weekly |
| Commodity Context | Oil market structural analysis | Weekly |
| Israel Alma Center | IDF operational reports, strike counts | Daily |
| ACLED | Conflict event data, casualty tracking | Periodic |
| ISW/CTP | Military situation reports | Daily |
| US EIA | Hormuz chokepoint baseline data | Periodic |
Over the last eight days (26 cycles), the pipeline drew on 189 unique source outlets. Source diversity is checked mechanically: no single outlet may account for more than 40% of citations, and at least one non-Western source must be cited every cycle.
62%
Western (US, UK, EU)
30%
Middle East & Asia-Pacific
7%
Specialist & military
Most-cited outlets: Al Jazeera, CNBC, Times of Israel, Bloomberg, CNN. Middle Eastern sources (Al Jazeera, Times of Israel, Jerusalem Post, Iran International, The National, Haaretz) collectively balance the Western wire services.
Every item must have a verifiable source URL in the original language of publication.
Every item must have a publication date. Items older than the previous cycle are classified as pre-existing context, not new developments, and are flagged separately to the analytical agents.
High-impact claims — those likely to move scenario probabilities — must be verified by reading the full source page, not just the search snippet.
A minimum of three genuinely new developments per cycle is required. If fewer are found, the cycle is flagged as a low-signal period and proceeds with available intelligence.
Key design choices that shape how this system works and what it measures.
We tested the scenario framework across multiple AI systems (Gemini, ChatGPT, DeepSeek) to check whether the ordering of scenarios was model-dependent. The broad ranking held. One notable variation: we configured DeepSeek as a Tehran-perspective monitor drawing on non-Western sources. It inverted the ordering of the two leading scenarios, rating regional widening above attrition stalemate from the start. This is consistent with the core lesson: how you frame the question shapes the answer. The governance layer matters more than the model.
Sentiment trackers record the volume of commentary: how much is being said. This monitor tracks the analytical signal: what is structurally changing. The difference is extensive margin versus intensive margin. A day with 10,000 articles repeating the same headline scores high on sentiment measures but contains no new information. A single IRGC statement changing the target set from military bases to energy infrastructure scores low on volume but reshapes the scenario balance. This system tracks the latter.
Each scenario probability answers a single question: what is the likelihood that this scenario is the dominant crisis trajectory four weeks from the assessment date? "Dominant" means the scenario's defining mechanism is most actively shaping the conflict at that point. This is not a prediction of final outcomes — it is a forward-looking assessment of which dynamic will be primary. Scenarios with shorter natural timelines (such as coercive submission) are assessed on whether they initiate within the window. Scenarios with longer timelines (such as attrition stalemate) are assessed on whether the pattern is entrenched. For tail-risk scenarios like nuclear sprint, the probability tracks the likelihood of the triggering decision, not weapon completion.