7.7 in Tehuacan, 87,000 Seats at the Azteca: The Evacuation Test Mexican Football Has Never Run

7.7. Epicentre at Tehuacan, in the state of Puebla. Four thirty in the afternoon on September 19, 2026. No ground actually moves. There is only a siren, loudspeakers buried underground, and a list of 27 instruction points drafted for the Second National Drill. The Mexico City Metro stops. The Metrobus stops. Cablebus cabins hanging over the slopes of Iztapalapa follow a procedure that has its own name: automatic stop when an alert fires, gradual stop when no alert fires, and vertical rescue if it comes to that.

September 19 was not picked because the calendar was empty. In 2026 an 8.0 earthquake erased part of the city centre; authorities recorded more than 10,000 dead, and independent researchers believe the real figure was higher. Thirty-two years later, on this exact date, a 7.1 quake killed close to 370 people, most of them in Mexico City. Residents of the capital do not need reminding what day this is.

About ten kilometres south of the centre, the Estadio Azteca is being renovated to host the opening match of the 2026 World Cup on June 11, with a capacity above 80,000. A concrete bowl holding eighty thousand people sits inside an active seismic zone. The September 19 drill is written about the Metro. The question it leaves behind belongs to the Azteca.

A network carrying five million riders a day

The Mexico City Metro moves roughly five million passengers a day across 12 lines and close to 200 stations, most of them deep underground, where phone signal is weak and lighting depends entirely on the power grid. The Metrobus, a bus rapid transit network on dedicated lanes, adds millions more. Cablebus, the city's first public cable-car line, lifts Iztapalapa residents several dozen metres above the ground every day.

Six lines of guidance for passengers during an earthquake, drawn from the drill's 27 data points, are short enough to be skimmed past:

• Stay calm, do not run, do not shout, do not push.
• Do not cross the yellow safety line.
• Move away from platform edges and glass panels, toward the middle of the platform or against the wall.
• If you are inside a car: hold on, stay in place, do not pull the emergency release, do not open the doors yourself.
• Leave the train or the station only when staff direct you.

Read another way, those six lines are a doctrine about crowds. In any emergency involving large numbers of people, the leading cause of casualties is the movement of that crowd through an enclosed, dark, unfamiliar space with stairs and slopes.

The arithmetic of eighty thousand people

Stadium safety has been measuring egress flow for decades. Common models such as Britain's Green Guide put exit flow at roughly 60 to 80 people per minute for every metre of clear exit width. A ground with 40 metres of effective exit width pushes about 2,600 to 3,200 people out of the structure every minute, under ideal conditions: full lighting, no smoke, no obstructions, nobody walking against the flow.

At 87,000 seats, the division returns 27 to 33 minutes to empty the stands. That number holds for an ordinary end-of-match exit, when people trickle out in waves. It fails completely in a real evacuation, for three reasons: everyone leaves their seat at once, some spectators run back up to find family, and emergency lighting is enough to illuminate a walkway but not enough to read a directional sign.

A Mexico City Metro train carries at most around 1,200 to 1,500 passengers when crush-loaded. The underground problem is easier than the stand problem in one respect: the crowd splits along two travel directions and multiple independent exits. A stand funnels everyone in a single direction: down.

Two incidents in the stadium safety record show the distance between theory and reality. On April 15, 2026, at Hillsborough, 97 supporters died inside a crowd compressed into a standing terrace. On October 1, 2026, at Kanjuruhan Stadium in Malang, Indonesia, 135 people died after the match ended and the crowd spilled toward the exits. Both disasters happened after the whistle, not while the ball was in play.

Data does not lie, but it does not feel pain either. I write to fill the gap between those two things.

What stands out in the Mexico City guidance is how the system separates two classes of event. Cablebus stops automatically when a seismic alert triggers, stops gradually when there is no alert, and calls a vertical rescue when a cabin is stranded between towers. A system willing to write down three different scenarios for one failure has already measured its own limits.

Thirty minutes nobody sells

A safety budget is the only line item in a stadium project whose success is measured by nothing happening. LED screens, hospitality areas, sound systems, premium seats: all visible, all sellable, all photogenic. A full-bowl evacuation drill is none of those. It produces a PDF, a written record, and thirty minutes during which 80,000 people walk downhill in an orderly line.

The Azteca is being renovated for a tournament of 104 matches across three countries. Broadcasters buy a package of 90 minutes plus stoppage time. No network buys thirty minutes of evacuation. That revenue structure pushes safety infrastructure to the bottom of the priority list, until it becomes the first headline.

In Mexico there is one more variable that most stands designed four decades ago never accounted for. Liga MX Femenil launched in 2026 and brought a different crowd structure into the ground: younger, more families, more children, more first-time visitors. The same narrow concourse, the same staircase, but that crowd moves more slowly. Women's football is not a scaled-down version. It is a world with its own rules, and those rules apply to the exits too.

Based on my experience watching matches, what I count inside a stadium has changed over time. In 2026, at sixteen, I watched FC St. Pauli Women lose 0-5 in Hamburg and rewound the video to log 14 tactical errors, then realised every goal conceded had travelled through the gap between full-back and centre-back. Years later I still rewind, but I count something extra: how many exits each stand has, and which one is blocked by an ambulance parked in the wrong place. That habit grew with the job. In 2026 I downloaded 40 Women's Champions League matches and wrote a Python script to build an open database of 350 European women players, including Amandine Henry and Dzsenifer Marozsan. A 25-year-old with Python can read a match more clearly than an entire commentary box. But no line of code measures what it feels like to be the person still sitting after the whistle.

In 2026, writing about Kosovare Asllani's injury in the Tokyo Olympic semi-final, I settled on a rule for myself: verify first, write second. A scoop about a player missing from a line-up can be literally true and substantively wrong. The same rule applies to a drill: the guidance reads smoothly, but its real value only shows when someone runs it with 80,000 real people, at the right height, the right gradient, the right dim lighting.

The hardest line to sell

Of the six lines of Mexico City guidance, the hardest to convince anyone of is the last one: evacuate only when staff direct you. It asks a frightened passenger to believe that the person standing by the door knows more than they do. In football, stand culture teaches the opposite: decide for yourself, leave early when your team is losing, wait for nobody's permission.

A 0-5 defeat is not about the loser; it is about the person who dares to stay until the final minute. In a 7.7 earthquake, the one who stays inside the car, grips the handrail, and does not pull the emergency release while the whole carriage surges toward the doors is the one following protocol. That person is not a coward. That person knows something the crowd has not caught up with yet.

Some conceded goals matter more than goals scored, if somebody bothers to write them down. A national drill is a collective written record of thousands of bad situations that did not happen. Mexico City has practised. What it has not practised is the most crowded place in the city: the stands.

Measuring a stadium by thirty silent minutes

A big stadium is judged on capacity, ticket prices, matchday revenue. The thirty minutes after the final whistle are the real test, and there is no scoreboard for it. If the Azteca's operators publish their egress data, if every match there rotates one block of stands through an evacuation drill, then by June 11, 2026 this city will hold one extra layer of insurance that no broadcast rights fee can buy.

If the siren goes off in the 89th minute of a semi-final, does your safety team know exactly what to do in the first ninety seconds?