Assefa and the Two Kilometres That Cost Her the Marathon World Record in Berlin

The last two kilometres were longer than the forty that came before them.

I was in front of a screen in Hai Phong, close to four in the morning local time, logging every split by hand the way I have for nine years of watching athletics. At 35 kilometres the clock showed a number I had to read twice: Tigst Assefa was running about twenty seconds under world-record pace. Those twenty seconds did not belong to someone hanging onto another runner. They belonged to a woman setting her own tempo, alone, on a road so flat you had to reopen the map to believe it. Then, at 41 kilometres, her stride pitched forward. Both feet locked. She went down at the roadside, and the world record vanished in the time it took me to set my pen on the desk.

She finished in 2 hours 11 minutes 04 seconds. Still the winner in Berlin for the third time. Still the second-fastest woman in marathon history. Still without the world record.

Berlin is not an accident

Among the six biggest marathons on the planet — Tokyo, Boston, London, Berlin, Chicago, New York — Berlin is the course chosen most often whenever someone wants to break a record. The route runs through the inner city, at an altitude of roughly 34 metres above sea level, with almost no meaningful incline. In the marathon, wind assistance and altitude do not work the way they do on the track; the only variables that matter are surface and terrain. Berlin offers both at something close to ideal.

The race's history backs that up. In 2026, Eliud Kipchoge ran 2:01:39 here to set the world record. Four years later, on the same streets, he lowered it to 2:01:09. In 2026, Assefa herself set the women's world record at 2:11:53. No other marathon has produced so many world records. Berlin does not stage a race; it stages a speed test with spectators.

One more rule matters here: major city marathons permit pacemakers, which national and Olympic championship races forbid. That means a race in Berlin can be turned into a pure time trial, with a group of pacesetters running ahead to hold the main contender on schedule.

Assefa did not come to Berlin to contest a placing. She came to reclaim something she once owned. In October 2026, in Chicago, Ruth Chepngetich ran 2:09:56 and took the record away, the first time a woman had ever run a marathon under 2 hours 10 minutes. Before that, the women's record had stood at 2:14:04, set by Brigid Kosgei in Chicago in 2026, and before her at 2:15:25, set by Paula Radcliffe in London in 2026, a mark that survived sixteen years. Then came the super-shoe era, carbon plates and foam midsoles up to 40 millimetres thick, all of it legal, lifting the whole speed baseline a notch. Chepngetich was the first to touch the 2:09 range.

Since then, every major women's marathon has circled a single question: who will be the second? Assefa is the most logical candidate. She came up from the 800 metres, moved to the marathon late and moved fast. At the Paris 2026 Olympics she took silver. At 28 or 29 she is squarely in the event's prime, the phase where the aerobic base is deep enough and the psychological endurance mature enough. Berlin was a calculated choice: a race outside the championship system, where the clock matters more than the placing, and where a record can be set without spending strength on a final.

Split architecture: the signature of a race

I have spent years watching road races, and there is one thing I always check before trusting any surge: how the finishing time was built, rather than the time itself.

Assefa's race carried a clear signature. She went through halfway inside world-record schedule. At 35 kilometres she was still under it. Hold that rhythm and she finishes somewhere around 2:09:36 to 2:09:45. This was no near-miss. It was a planned sub-2:10 attempt. Assefa's problem was not speed; it was that her body stopped answering at the exact moment she needed it most.

To grasp the margin, look at the numbers. Chepngetich ran 2:09:56. Assefa ran 2:11:04. Sixty-eight seconds over 42.195 kilometres, which is under 1.7 seconds per kilometre. That is a gap a late cramp can erase entirely.

In the marathon, the ideal shape is an even split. When the second half is clearly slower than the first, it is called a positive split, and it usually points to one of two things: the athlete went out too fast, or the body collapsed late. Assefa's case is the second. She did not lose rhythm through poor maths. She lost it to cramp.

Cramping in the marathon is not a matter of luck. At maximal effort sustained for more than two hours, the body loses salt through sweat, glycogen stores drain, and the neuromuscular system begins sending faulty signals to muscles that are already tired. That is a fuelling and electrolyte problem far more than a talent problem. In other words, it is a fixable error. I stress that because it changes how the result should be read: if the cause is physiological and adjustable, then Assefa's true ceiling still sits somewhere around the record line, just short of being touched cleanly.

There is another variable that gets little mention. With no rival, all pacing discipline falls to the pace group. Assefa broke clear with her pacemakers, meaning she had a fully professional team holding tempo ahead of her. But when there is nobody to race, the only thing you race is the clock. And the clock never tires. That is the most punishing kind of pressure in athletics: chasing a target with no shape, no face, no feedback. With no roar to hold on to, I hear my own applause more clearly, and I suspect that in those last two kilometres, Assefa could hear only her own breathing.

So how should 2:11:04 be read? It is faster than the world record Assefa herself once held, 2:11:53. If the all-time list stands as it is, it ranks second, behind Chepngetich alone. On a morning when everything broke down in the final two kilometres, she still ran faster than any woman in history except one.

The contrarian read: the problem is not the last two kilometres

The popular telling after this race is a story of collapse. An athlete falls within sight of the record. That telling skips a simple fact: she fell after running the second-fastest time ever.

I disagree with the reading of this as a failure. Sport usually measures what was lost: no record, no complete moment. Change the yardstick and the story flips. An athlete came to Berlin to run under 2:10, ran exactly that for 40 kilometres, and failed only in the last two. The error sat in physiological tolerance, not in ability.

What deserves mention is that she refused to quit. After going down, she got up, kept moving, finished. She did not leave the course. That detail does not appear on a results sheet. In how people remember an athlete, it appears a great deal.

Another layer gets hidden by the performance arithmetic. This result came in a race Ethiopia nearly owned. The top five women were all Ethiopian. In the men's race the winner ran 2:02:50, and the next two places were also Ethiopian, with only a Tanzanian in third breaking the line. Luck does not produce that. It comes from a high-altitude development system where distance running is raised as a subject from a very young age.

On the same day, the men's 2:02:50 showed that the speed baseline in Berlin still sits at the very top of the world marathon system. That also means conditions this year were not unfavourable. Sunny, but not muggy. Dry roads. Light wind. Everything a record attempt needs was present, except one thing inside Assefa's body.

When one nation fills the top five, the biggest problem stops being beating foreign rivals and becomes beating your own compatriots. In Ethiopia, the sceptical gaze does not come from the stands. It comes from the person training beside you every morning. I have lived a milder version of that mechanism. In 2026, at sixteen, I sat in the commentary box for the U19 Vietnam versus U19 Guam match at Lach Tray stadium, and before kick-off a production director asked me flat out what a girl understood about tactics. I did not argue. I went home, downloaded twelve qualifying matches, and watched until three in the morning. That first headset was heavier than I expected, but my voice was heavier still. The pressure in Ethiopia is probably far greater, because there the person competing with you is not a prejudice; it is five of the fastest women on the planet carrying the same flag.

There is also a question analysts are tracking without an answer. The 2:09:56 record Assefa was chasing belongs to one specific athlete. The stability of that mark is a matter of rules, and it has not been confirmed in either direction. I draw no inference about Assefa from this; her performance is entirely consistent with her standing as a former record holder and an Olympic medallist. But I log it as a variable to watch, because if that mark ever changes, the entire all-time list gets reshuffled, and her 2:11:04 takes on a different meaning.

What is left after the finish line

Stars are not born in finals, but in the races nobody watches. In Berlin there was no rival for Assefa over the second half, only herself, the road, and a number. And sometimes the hardest thing is not running fastest, but knowing when your body is about to run dry.

I think about the mornings in Hai Phong, where I still run along the coastal road before the sky lightens. In Vietnam the distance-running movement is growing fast, and every weekend brings more people completing 42 kilometres for the first time. Most of them will never touch 2:11, and that does not make their effort smaller. The marathon is the sport where your biggest opponent never appears on the course. The women's world record is still waiting for someone to touch it cleanly. Assefa came very close. The distance between very close and touched, in the marathon, is measured in two kilometres.