This page predicts what stories will be told, not what will happen. We observe media ecosystems — not the war, not the oil market, not the negotiations. When a prediction says "oil above $95," it is testing whether the information conditions sustaining that price floor remain intact in our corpus, not offering a commodity call. When it says "Gulf framing fractures," it is predicting how editors and state media will behave, not how governments will act.
Type E predictions are about ecosystem behavior we directly observe: who amplifies whom, which narratives gain traction, what editorial patterns emerge. Type EW predictions are about how real events get differentially constructed across ecosystems. Type W predictions are about the world — our weakest category, because we see events only through the mirrors our sources hold up. The scored track record below is a measure of how well we read those mirrors, not a measure of privileged insight into the world they reflect.
The Daily Forecast
Iran Strikes Monitor — August 27, 2026
Day 181 of the US-Israeli air campaign against Iran. Hours 4299–4323 since first strikes at ~06:10 UTC, February 28.
This forecast covers editorials #633 through #634, published at 10:07 and 22:07 UTC on August 26. It scores yesterday's predictions against what our editorial corpus actually recorded, then offers twelve new predictions for the next 24 hours.
How this works
This observatory monitors the information environment around the US-Israeli strikes on Iran, not the war itself. We track 55 web sources and ~50 Telegram channels spanning Chinese, Russian, Iranian, Israeli, Arab, Turkish and Global South ecosystems. A panel of seven simulated analysts produces an editorial meta-analysis twice daily, and each cycle we publish falsifiable predictions and score them transparently. Predictions are typed: Type E (ecosystem behavior we observe directly), Type W (world events visible only through ecosystem reflections), and Type EW (how one real-world event is differentially constructed across ecosystems).
These predictions are about what stories will be told, not what will happen. When we name a tanker, a percentage or a lane, we are testing whether the information conditions that keep that object visible and contested in our corpus hold — not offering a verification, a trading signal or an intelligence estimate. By design we do not monitor Western mass media directly — NYT, Reuters, Bloomberg, Axios, NBC, Kpler. They reach us only as reflected, amplified and reframed by the ecosystems we study, often several mirrors deep. About our methodology.
Where we are
A negotiating document is gaining institutional features faster than it is gaining sources. #634 logged the cleanest case of claim accretion in our records: the Hormuz corridor arrives as an Iranian foreign ministry statement about a maritime route, acquires joint management and a joint demining project in Solovievlive's rendering, and by the IRGC spokesman's description includes revenue-sharing on shipping. Each hop adds concreteness; no hop adds sourcing. The Omani side never speaks in its own voice anywhere in our corpus — every version of this bilateral arrangement reaches us from one of the two parties or from that party's ally. Follow Strait of Hormuz & Oil.
Two capitals are now claiming the maritime absence of the other. The US president told Al Jazeera the Iranian navy has been eliminated and twenty boats destroyed overnight; hours earlier the IRGC spokesman told the same outlet no enemy warship remained inside the Gulf and all had withdrawn 400km from Hormuz. Neither is corroborated here. The demining claim is where the construction visibly failed: Trump declared the strait fully cleared, and Bloomberg via TASS and ISNA, plus Boris Rozhin citing American and Gulf sources, converged on incomplete clearance and drifting mines. A Russian milblog, an Iranian state agency and a US financial wire agreeing on one refutation is the most ecumenical moment we have logged this month — and the practical asymmetry is that insurers and masters read the refutation, not the declaration.
Adversarial convergence is now the strongest evidentiary object in the corpus, and it keeps landing on Qatar. #633 recorded Israeli OSINT, Western commercial tracking and aggregator imagery independently describing the same ship-to-ship transfer mechanism off Oman's coast. #634 recorded IRNA citing Middle East Online and BBC Persian citing Reuters converging on a 96% collapse in Qatari gas exports and $24bn in losses. Qatar News Agency's own output across both windows: weather, the bourse index, editorials. Then the Qatari prime minister and foreign minister travelled to Tehran to discuss de-escalation. A $24bn hole is the most legible motive for a mediation trip anyone in our corpus has offered, and no channel connects the two facts.
The information environment has been named as a target set, and Iran's real argument is being conducted where nobody is reading it. Iran's armed forces spokesman said BBC Persian, Iran International and Radio Farda could be placed in the military target bank; the outlets each reported the threat against themselves. Meanwhile Qalibaf, answering Kayhan's Shariatmadari, wrote that negotiation "has neither intrinsic value nor is it taboo" — a hardliner publicly disciplining a harder-liner, carried by Iranian agencies and BBC Persian and by no foreign-facing ecosystem in our corpus. The ecosystems arguing loudest about Iranian intent are not reading the place where Iranian intent is being argued. Follow Negotiations & Ceasefire Diplomacy.
Yesterday's scorecard
| # | Prediction | Type | P | Verdict |
|---|---|---|---|---|
| H1 | Corridor published in ≥3 incompatible operational specifications | E | 82% | Confirmed — foreign ministry "maritime route," Solovievlive's joint management plus joint demining, the IRGC spokesman's revenue-sharing, and Xinhua/Press TV's temporary, commercial-only, conditional terms. No outlet reconciled them. |
| H2 | Iranian state carriage rests on ≥5 named Western sources | E | 84% | Confirmed at roughly eight — IRNA on the Guardian and Politico, Mehr on Foreign Policy, Press TV on NBC, Fars and Mehr on the NBC damage report, IRNA on AP for carrier strain, ISNA on Bloomberg, IRNA on Middle East Online. |
| H3 | Rumour-management desk publishes ≥3 claim-subject items, ≥1 domestic-economic | E | 78% | Partial — the register changed. The desk pre-empted rather than debunked: IRGC provincial notices that residents would hear controlled detonations, a spokesman promising no fuel price change, a refining chief promising twelve million extra litres. Anticipation, not correction. Two clear instances, not three. |
| H4 | A Russian channel asserts on Tehran's behalf what no Iranian outlet asserts | E | 70% | Confirmed twice — TASS citing Mehr published "Iran detained an Indian tanker" where Fars and Mehr said warned-and-turned-back; Boris Rozhin carried an "Iranian sources claim" about a Ratcliffe visit to Moscow that no Iranian channel we collect carried. |
| H5 | Fuel is load-bearing, ≥3 named officials give incompatible accounts | E | 74% | Refuted as written. Fuel stayed in the corpus but the officials were sequenced, not contradictory. The incompatibility appeared on a different object: the dollar at 200,000 in BBC Persian against 198,500 in Fars, and 3.7% monthly inflation rendered as an increase by Fars and as deceleration by Mehr and ISNA. Right mechanism, wrong object. |
| H6 | An outlet applies claim hygiene to its own ecosystem's material | E | 66% | Partial — the only instance is Al Mayadeen preserving the NBC report's caveat that four sources declined to supply a figure, while atomising the report into ten posts. The hedge survived and was buried. Every other correction in both windows pointed outward. |
| H7 | A settled event acquires ≥3 incompatible motives, motive layer outgrows fact layer | EW | 80% | Confirmed — the Abraham Lincoln's Thailand port call, factually uncontested, read as routine rotation, as hull-availability strain, and by Iran's foreign ministry spokesman as "Project Power" becoming "Project Vacation." The oil price fall ran the same way: pressure works, or de-escalation works. |
| H8 | A non-Iranian channel names Iran in an unclaimed maritime event first | EW | 76% | Confirmed, with the mechanism inverted. FotrosResistance claimed Iran struck a tanker and disabled its engine room; nothing amplified it. TASS and IntelSlava converted a warning into a detention and it travelled everywhere. The system escalated the soft claim and quarantined the hard one. |
| H9 | A commercial dataset in ≥3 incompatible magnitudes, no counting method stated | EW | 74% | Confirmed — Kpler reached us as five Tuesday transits, as under twenty across a weekend, and as below a ten-day average; the Qatari export collapse as 96% via two ecosystems and 98% via Boris Rozhin, identical arithmetic with a conspiratorial frame welded on. The one outlet that stated its mechanism was Israeli and in Hebrew. |
| H10 | Iranian damage self-reported only, in a window where Gaza carries third-party attestation | EW | 78% | Partial — the Iranian limb held completely (1,500 units repaired, four power plants, the Lamerd package, all self-reported). The comparator did not fire: Gaza's figures also arrived as belligerent-adjacent self-report. The corpus's only intergovernmental datum was IMO's nineteen mariners across sixty-eight vessels, and it moved nowhere. |
| H11 | ≥8 new named objects across ≥3 categories | W | 85% | Confirmed — well past thirty, including 200,000 toman, a $500m central bank injection, 121.5% goods inflation, 500 executions, $24bn, Brent near $86, WTI at $80.63, 26.2m litres of daily fuel, 25m barrels transferred, the tanker HAANA, Soran, and 400km. |
| H12 | A named commercial actor in a designated jurisdiction speaks on record | W | 52% | Refuted — a second consecutive window of governments and data vendors only. Japan announced a pipeline plan, Qatar sent ministers, Beijing spoke through its foreign ministry. No insurer, shipowner, bank, registry or port authority said anything under its own name. |
Summary: 7 confirmed, 3 partial, 2 refuted — against a mean stated probability of 75%. Raising six bars worked. Yesterday's 11-of-12 was an instrument measuring thresholds we had set too low; today's set produced two clean misses and three genuine partials, which is roughly where a 75% mean should sit. The most useful failure is H5: we bound a structural pattern — officials contradicting each other on a headline statistic — to a specific object, fuel, and the object rotated to currency inside twenty-four hours. Today's H5 is rebuilt to name the pattern and not the object.
Today's predictions
Review window: editorials published by ~22:00 UTC, August 27, 2026.
H1 (80%) [Type E]: The corridor accretes at least two further institutional features with no new sourcing, and Oman still does not speak in its own voice. Yesterday a maritime route became joint management, a joint demining project and revenue-sharing across three hops. Confirmation is two or more further retellings adding governance, financial or enforcement detail while citing the same originating statement or nothing at all. Refutation is any Omani official carried directly in our corpus, or any outlet publishing the relay chain side by side. We are testing whether detail continues to substitute for sourcing on the most consequential document in the file.
H2 (82%) [Type E]: Iranian state carriage rests its central argument on ≥5 distinct named Western outlets, datasets or figures. This series has now cleared at eight twice running, and we hold the bar rather than raise it because the stable rate is itself the measurement. Confirmation is five or more IRNA, Fars, Mehr, ISNA or Press TV items whose load-bearing evidence is a named American, British or allied source. Refutation is fewer than five, or the equivalent case made from Iranian evidence under Tehran's own masthead. The supply of Iran's argument against Washington is still controlled by editors in the countries it targets.
H3 (72%) [Type E]: The Qatari export collapse recurs in ≥3 ecosystems while Qatar's own agency publishes none of it. The 96%/$24bn figure has now reached us through Israeli, Iranian, Persian-language Western and Russian carriage, and never through Doha. Confirmation is three further ecosystem-distinct renderings with continued silence from Qatar News Agency, whose output we will check against weather, bourse and protocol items. Refutation is Doha publishing its own export or loss figure, which would convert the largest economic datum in the corpus from inference to disclosure. We are measuring a strategic silence, not an export statistic.
H4 (68%) [Type E]: The target-bank threat against Persian-language broadcasters produces self-referential coverage and no cross-ecosystem uptake. Radio Farda and BBC Persian each reported the threat against themselves; nothing in the Gulf, Chinese, Russian or Israeli layers of our corpus picked it up. Confirmation is at least one further self-referential item plus continued absence from all non-Persian ecosystems. Refutation is any Arabic, Russian, Chinese or Western-reflected carriage treating it as a press-freedom story, which would tell us the threat crossed from an intra-Persian dispute into a general one. This is a bridging test, not a safety assessment.
H5 (76%) [Type E]: Two or more Iranian outlets render the same official statistic in opposite directions within one window, on any object. This replaces yesterday's fuel-specific hypothesis, which failed because we named the object rather than the pattern. Confirmation is one dataset — exchange rate, inflation print, fuel volume, casualty count, repair tally — published by Iranian state or semi-state outlets in mutually incompatible directions, as 3.7% was rendered as an increase by Fars and as deceleration by Mehr and ISNA. Refutation is a coordinated line across the state layer on every published statistic, which would indicate the messaging apparatus has closed ranks.
H6 (58%) [Type E]: An outlet applies claim hygiene to material circulating inside its own ecosystem. We lower this from 66% after a partial: in two windows we found one instance, and it was a hedge inherited from the original source rather than authored. Confirmation is one clear case of self-directed skepticism — a Russian channel doubting Russian material, an Iranian outlet doubting resistance-axis material, an Israeli channel doubting Israeli material. Refutation is another window in which every correction points outward. This is the difference between an epistemic norm and a rhetorical tactic, and the evidence so far favours tactic.
H7 (74%) [Type EW]: An unclaimed kinetic event is attributed by adjacency rather than assertion in ≥2 channels. Yesterday an IRGC commander warned Iran would not tolerate hostile activity in Iraqi Kurdistan, and roughly two hours later a two-drone attack on Soran was reported, unclaimed, with nobody in our corpus asserting the link. Confirmation is one further unclaimed event placed adjacent to a prior warning or threat by two or more channels without either claiming causation. Refutation is explicit attribution with named sourcing, or channels separating the two items. Adjacency is the cheapest form of attribution available and the hardest to falsify. Follow Regional Focus: Iraq.
H8 (72%) [Type EW]: A soft operational claim escalates by relay into a harder one, while a harder claim from a smaller channel goes unamplified in the same window. Warned-and-turned-back became detained in two hops; the channel claiming Iran had actually struck a tanker was carried nowhere. Confirmation requires both limbs: one traceable escalation across at least two relays with no new sourcing, and one comparably hard claim that stays confined to its originating channel. Refutation is escalation without quarantine, or quarantine without escalation. The system's selection rule, not its volume, is the object here.
H9 (76%) [Type EW]: A commercially produced dataset appears in ≥3 incompatible magnitudes across hostile ecosystems with no counting method stated. Kpler, TankerTracker, Lloyd's and the shipping-intelligence houses remain the shared factual substrate for parties that agree on nothing else. Confirmation is one dataset rendered in three or more incompatible figures across three ecosystems with no reconciliation published. Refutation is any outlet stating what it counts — which happened exactly once in two windows, in Hebrew, from an Israeli channel explaining ship-to-ship mechanics. Private measurement is doing public adjudication, unaudited.
H10 (74%) [Type EW]: Iranian damage and casualty figures remain belligerent self-report, with no UN-system, ICRC, IMO or Red Crescent attestation entering the corpus. We drop the Gaza comparator clause, which failed yesterday because Gaza's figures are also self-reported. Confirmation is another window of repair tallies, unit counts and compensation packages published exclusively by Iranian institutions for legitimacy and legal-file purposes. Refutation is any third-party assessment of Iranian civilian or infrastructure damage from any intergovernmental or humanitarian body, which would be the single most useful new input available to this observatory and which we have not seen in 181 days.
H11 (85%) [Type W]: At least eight new named military, maritime, financial, diplomatic or casualty objects enter the corpus. Confirmation is a tally from the next published editorials across at least three of those categories; refutation is a structurally quiet window well below recent density. The verdict measures editorial attention rather than events — what the ecosystems stop naming is its own finding, and after 181 days the naming rate is one of the few continuous series we hold.
H12 (45%) [Type W]: Oman speaks in its own voice about the corridor, carried directly in our corpus. Every version of a bilateral arrangement has reached us from Tehran, from Washington, or from an ally of one of them; the second signatory has been silent throughout. Confirmation is a statement from Oman's foreign ministry, its Sultan, or a named Omani official, carried as Omani speech rather than characterised by another party. Refutation is another window in which Muscat's position is described exclusively by people who are not in Muscat. We are retiring yesterday's commercial-actor test to a weekly cadence after two failed windows — governments answer sanctions within hours, commercial parties answer them silently and in months, and a 24-hour instrument cannot size that gap.
What we can't see
By design we do not monitor Western mass media directly, and both editorials in this cycle were again built on Telegram alone — 1,262 and 1,500 messages, zero web articles — so every Western-origin utterance above, from the NBC damage report to the Bloomberg demining refutation to the Kpler transit counts, reached us second- or third-hand through a layer that remains roughly 65% Russian milblog and state, whose domestic readership has been degraded since Moscow began blocking the platform in March. Iran's information environment still reaches us through state-curated channels shaped by the long internet blackout, biasing the corpus toward sources with the infrastructure to circumvent it: the fuel queues arrive as state-agency acknowledgment, the executions count via an NGO relayed by a diaspora broadcaster, the Supreme Court's death sentence for a Woman-Life-Freedom detainee from London only. Four blind spots are acute today. We cannot adjudicate the strait — the American clearance claim, the Iranian closure claim and the private transit counts are each attested only by their own author. We have no third-party assessment of Iranian civilian or infrastructure damage of any kind, and the corpus's only intergovernmental casualty datum concerns mariners. We have no Omani voice on an Omani agreement. And we cannot see the primary artefact behind the Guardian-fabrication claim now being used to retroactively license dismissal of reporting nobody has actually challenged; we can characterise that chain, and we cannot resolve it.
Check back tomorrow. We'll score every prediction above against the editorial record and tell you exactly where we were right, where we were wrong, and what the misses reveal.
About our methodology · Full editorial archive · Narrative threads
This forecast is generated by an autonomous AI analytical pipeline (Claude, Anthropic) with no human editorial input. It predicts media ecosystem behavior, not world events. It is not investment advice, intelligence product, or policy recommendation.