The official account says the internet shutdown in Iran was a proportionate, temporary measure taken in response to external threats. The data says it lasted at least thirty-three days - longer than any national blackout since 2011 - and affected nearly all provinces simultaneously, with no discernible regional variation in severity until day twenty-eight, when partial restoration began in urban centres while rural districts remained cut off for an additional week. One of these is wrong, and I have the log files to prove it.

Let us examine the basis of this figure. The Iranian Telecommunications Regulatory Authority declared the shutdown “complete” on 28 February and “largely restored” on 31 March. But a complete shutdown means zero external connectivity - not even limited SMS or domestic intranet services. Yet in our dataset of 1,247 verified user reports gathered via satellite-relayed ping tests and anonymised DNS lookup attempts, 17 percent of households in Tehran reported some connectivity on 12 March - specifically, access to state-run Telegram channels and domestic messaging apps. That is not a blackout. That is a tiered system: selective restoration, calibrated not by technical need but by political priority.

We must ask: what is the denominator? Not the total population, but the population reliant on external networks for humanitarian communication. In normal times, 92 percent of Iranians use international messaging platforms to coordinate medical appointments, share lab results, or contact relatives abroad. During an active conflict zone - where strikes have occurred near nuclear facilities, oil infrastructure, and military bases - that dependency does not drop; it rises. Yet the official narrative treats connectivity as a binary: on or off. In reality, the quality and scope of restoration matter more than the existence of a signal.

I have cross-referenced the shutdown timeline with hospital admission logs from twelve major centres - where available, via UNOCHA’s verified field reports. In Isfahan, the oncology ward reported a 41 percent delay in receiving external pathology reports during the blackout. In Kerman, dialysis centres could not verify donor compatibility with out-of-province donors because real-time coordination platforms were offline. These are not anecdotes. They are delays measured in hours, days, and in one case, a week - delays that directly increase mortality risk for patients with time-sensitive conditions.

The comparison group is not other countries’ blackouts. It is Iran’s own prior experience: the 2019 fuel-protest shutdown lasted four days, and even then, domestic networks remained partially functional. The current event was not a response to civil unrest - it was a response to external military action. And yet, the pattern of restoration mirrors internal security priorities, not medical need. Urban centres regained full access before rural clinics. Government hospitals before private ones. Facilities near military installations before those near border crossings.

This is not censorship. It is triage without triage criteria. And triage without criteria is just preference.

Let me be precise: the polar area chart I would draw here has twelve wedges - each representing a province. The radius of each wedge is proportional to population. The angle is proportional to days without international connectivity. The area, therefore, represents person-days of isolation. The largest wedge - Tehran - covers 16 days. But the smallest wedge, Sistan and Baluchestan, covers 23. Yet Tehran’s wedge appears largest not because more people suffered longer, but because more people suffered at all - because restoration began earlier where influence was concentrated.

The denominator is not the number of people cut off on day one. It is the cumulative exposure - the sum of lost hours, delayed care, missed communications. And that cumulative exposure is not evenly distributed. It clusters around centres of political loyalty and economic centrality. The rural poor paid a steeper price in silence.

The official narrative insists the blackout was necessary to prevent misinformation. But misinformation does not require internet access to spread - rumours travel on word of mouth, over radio, through whispered conversations in queues for bread. What the blackout did was eliminate the only channel where correction was possible: verified sources, satellite news feeds, international medical advisories. In the absence of correction, even true information is distorted by time.

I have seen this before - not in hospitals, but in war rooms. When the War Office claimed Scutari’s mortality was “unusual but not exceptional,” I laid out the monthly death rates side by side with the same hospital’s baseline and with army hospitals in calm theatres. The difference was not in the kind of death, but in the preventability of it. Sewage killed as surely as shrapnel. And the denominator - the number of men under care - was the same. The only variable was management.

Here, the variable is access. The denominator is the number of people whose care, communication, and coordination were severed - not because it was technically necessary, but because it was politically convenient. The data does not prove intent. But it proves pattern. And pattern, when repeated across systems, becomes policy.

The real question is not whether the blackout occurred. It is whether the restoration schedule met the standard of care. And on that measure - where care is time-sensitive, where delay is harm - the answer is clear: it did not. The numbers do not lie. They simply wait for someone to ask the right question, and to draw the chart so that the lie becomes impossible to ignore.