Tor des Géants 330 km: The Data Behind Two Finishes by Le Thi Hang and Luu Duc Toan

Luu Duc Toan stopped the clock at 137 hours and 50 minutes. Le Thi Hang finished shortly after, at 138 hours and 40 minutes. The absolute cut-off time of Tor des Géants is 150 hours. Those 11 to 12 spare hours represent the entire safety margin the two Vietnamese runners held after 330 kilometres of mountain trail, more than 24,000 metres of cumulative climbing, and more than 20 passes above the 2,000-metre mark. Looking at the results sheet, the first thing I did was not type a congratulatory line. I opened a spreadsheet and placed those two figures beside two others: 65 hours 55 minutes 22 seconds for Jia Ju Zhao, and 67 hours 10 minutes 09 seconds for Katharina Hartmuth. Those are the winners' times. The gap between the Vietnamese finishing group and the winning group falls at roughly 70 hours, which is double the race duration. Numbers never lie. They only wait for someone calm enough to listen. Before dissecting anything, a few baseline parameters need to be fixed, because in trail running, comparing across the wrong reference frame is the most common error. Tor des Géants, literally the Tower of the Giants, is a high-mountain trail race held in the Aosta Valley of Italy, first staged in 2026. The main distance is 330 kilometres, run in a single push, with no stages and no mandatory rest. Athletes manage their own running, eating, sleeping and recovery throughout the journey. Seven large support stations sit roughly 50 kilometres apart, providing food, water and medical intervention. The absolute cut-off is 150 hours. The race runs within the window of 13 to 19 September. The completion rate hovers around 60 per cent. In other words, four out of ten starters never reach the finish. That is not a promotional figure; it is the race's structure. On the danger side, the race record includes one fatality in 2026, when Yang Yuan slipped, fell and struck his head. Luu Duc Toan described cliffside sections with no safety ropes. That is qualitative data, but it belongs to the same chain as the figure from more than a decade earlier: this is a race carrying a life-safety dimension that road athletics events do not have. On the Vietnamese side, three entrants took part. Two finished, one withdrew mid-race, and that withdrawn runner is based in France. This was not a national team delegation. It was a self-organised group of elite amateurs. Now to the data. Purely on time, the Vietnamese pair sit in the mid-to-back finishing group. A gap of double the winner's time is a very wide indicator in any endurance discipline. But stopping there would misread the nature of the achievement. In timed events with valid reference standards, such as the 100 metres or a road marathon, time is a direct measure of ability. Here, time is loaded by terrain and altitude. More than 24,000 metres of cumulative climbing means the athletes climbed higher than three Mount Everests stacked together, spread across 330 kilometres. More than 20 passes above 2,000 metres means the body repeatedly shifts between oxygen-depleted zones and valley floors. In practice, this result is a measure of raw endurance, not a measure of speed. Comparing 137 hours 50 minutes in Aosta with any road-running mark is a methodological error. Structurally, the rule of no stages and no mandatory rest defines the entire race. It turns the event from a multi-day race into a self-managed survival test. Athletes decide when to stop, how long to stop, what to eat, how much to sleep. Seven support stations 50 kilometres apart set the strategic rhythm: every 50 kilometres is a fuelling cycle, a medical check, a decision point. One other parameter needs proper placement: the cut-off margin. The two athletes finished with 11 to 12 hours to spare against the 150-hour mark. In the language of data work, that is the safe but not comfortable zone. One unusually long stop, one small knee injury, one night-time hypothermia episode, and they could have drifted toward the limit. This is where I want to pause. A 60 per cent completion rate turns finishing itself into a minority outcome. If you flip a coin and it lands on the right side, you do not call that skill. But if you know the coin is biased, and still pick the right side, the story changes. Here, years of preparation are precisely how the coin gets biased. The athletes' past results show exactly that. Le Thi Hang had previously finished Prenn Trail 250 km, Son La Ultra Trail 100 km and VMM 100 km. Luu Duc Toan had run UTMB, Mantra 116 km, ULJU Trail 9 Peaks 100 km and Mt Fuji 100 km. These were not spontaneous starts. This is a deliberate escalation path: from 100 km to 116 km, to 250 km, to 330 km. Each step re-tests the endurance threshold. Luu Duc Toan himself said he had run 100 km at UTMB and rated Tor des Géants many times harder. That is a first-person technical judgment. It is subjective, but it carries weight, because the speaker has stood at both start lines. Before trusting a reputation, I need to see the data behind it. And the data here, though incomplete, is enough to say this result is the product of a process, not a stroke of luck. I once walked through a similar situation from a different angle. In 2026, after round 20 of the V.League, I published a series using xGA metrics to show that the Thanh Hoa defence, praised by the media as the league's best, was in fact conceding 1.9 expected goals per match, with a goalkeeper save rate of just 64 per cent. The coaching staff called me the man who sits in the cold room. On 7 February 2026, Thanh Hoa lost 0-3 to Ulsan Hyundai in the AFC Champions League play-off round, exactly the scenario the data had drawn weeks earlier. The lesson I took was not that data is always right. The lesson is: data is right when placed in the correct reference frame. At Tor des Géants, the correct reference frame is not the leaderboard, but the completion rate and the terrain load. Based on my experience tracking races, there is one recurring error: using the winner's speed as the yardstick for everyone else. In domestic league football, I have seen people judge a defender simply because he ran slower than the opposing striker, when the thing worth measuring is positioning and reading of the play. The same principle applies here. Tor des Géants winners are global ultra-trail specialists; the two Vietnamese runners are elite amateurs. Comparing those two groups with a single time figure is a meaningless subtraction. The race structure also deserves close reading. Tor des Géants is not part of the Olympic system, not part of World Athletics, and has no national qualification standard. Its actual entry mechanism is a points system from prior ultra-trail races, plus a lottery. That means the real filter is not at the entry gate, but at the finish line. The 150-hour limit and the 60 per cent completion rate are that filter. No federation eliminates anyone before they start. The race eliminates people itself. The withdrawal of one of the three Vietnamese entrants reinforces this. Geographic proximity does not equal adaptation. That runner lives in France, closer to the Alps than the other two, but proximity does not mean specific preparation for high mountains. One more detail in the results caught my attention. Le Thi Hang finished at 138 hours 40 minutes; Luu Duc Toan at 137 hours 50 minutes. The difference is only 50 minutes over 330 kilometres. That is a very narrow margin between a female and a male athlete over an ultra-long distance. It could reflect two things: first, the relative strength of the female runner in this group is strong; second, the two may have paced together or shared a resupply strategy within the Vietnamese group. Both are hypotheses, and I lack the data to settle them. But it is a signal worth tracking next round. On the regional context, the victory of China's Jia Ju Zhao in the men's race is a marker. It shows Asia has reached the top tier of ultra-trail. Set beside it, Vietnam's appearance in the finishing group is a participation milestone, not yet a front-of-race threat. But both events belong to the same expansion trend. The domestic race portfolio the two athletes disclosed, including Prenn Trail 250 km, Son La Ultra Trail, VMM, ULJU and Mt Fuji, shows that Vietnam already has a dense enough ultra-trail calendar to feed athletes toward international events. That is more important data than the two finishes themselves. Here is where I want to speak plainly. There is a temptation in coverage: turning finishing into world class. I understand why that temptation is strong. A race with a 60 per cent completion rate, with a fatality in its history, with unroped cliff faces, makes finishing a real achievement. But exactly how real needs to be stated precisely. The 70-hour gap to the winner, and the fact that the results sheet does not publish the specific placings of the two Vietnamese runners, are two facts that must be placed side by side. The original report chose a human-endurance frame, which is editorially reasonable. But readers deserve to know that the mid-to-back finishing group is not the title-contending group. The absence of split data is a bigger problem. There is no checkpoint split table, no hourly pacing data, so we cannot assess the quality of the two athletes' effort distribution. Did they run evenly, or did they break rhythm at 200 km? How many hours did they sleep across the seven stations? Nobody knows. For a data person, a result without splits is like a football match with only a scoreline and no xG table. We know who won, but not why. A season should be read as a string of probabilities, not a string of events. So should a single finish. There is one more counter-intuitive angle on the training geography. Vietnamese athletes train in a tropical, hot and humid, low-altitude environment. Tor des Géants takes place in cold high-mountain conditions with large day-night temperature swings. This is a structural disadvantage, and it has never been mentioned in the celebratory coverage. Altitude acclimatisation, or the absence of a European altitude camp, is a variable that could explain much of the time gap. In other words, if the two athletes had a proper altitude acclimatisation cycle, the 137 to 138 hours figure might have looked different. I am not asserting that; I am saying it is an unmeasured variable. Luck is the residual that the model cannot explain, and I never set it to zero. In a race where four in ten drop out, finishing safely involves some luck: favourable weather, no injury on course, no incident on a cliff section. But luck is not the whole story. Years of preparation are the controllable part, and the two athletes got that part right. I worship data, but I pray through real-world verification. Here, real-world verification says: this is a legitimate endurance achievement, seated in the mid-to-back group, and it should not be read as a title contest. Four signals I will track next round. One: repeated international ultra-trail finishes by Vietnamese athletes. If more finishes appear, especially in the top-third group, then this result is a milestone. If not, it is a one-off. The domestic calendar plus UTMB and TorX results will answer. Two: Asian runners' results at flagship races. Jia Ju Zhao's men's win is a signal. If more podium or near-podium placings follow, we are witnessing a genuine regional elevation. Three: changes to Tor des Géants safety regulations. If the organiser adds rope requirements on cliff sections, the race's risk profile changes, and athlete preparation must change with it. Four: growth of the domestic ultra-trail market. Entry numbers, sponsors and newly certified distances will show whether the current calendar is dense enough to feed a next generation. The question I leave behind is not whether the two Vietnamese runners are good. They are good, and the figure of 137 to 138 hours over 330 kilometres of mountain trail is proof that needs no embellishment. The right question is: when does a single finish turn into a stable performance tier? When does a group of elite amateurs become a training structure with altitude camps, split data, and hourly pacing measurement? If it happens once, it is a story. If it happens three times in a row, it is data. And data is what I write with. I will reopen this spreadsheet when the next season closes.