Why Durability, Not FTP, Explains This Year's Tour de France

Tadej Pogačar rolled into Paris on July 27 with his fifth Tour de France title, tying the record shared by Eddy Merckx, Bernard Hinault, Miguel Indurain, and Jacques Anquetil. The margin was 6 minutes and 26 seconds over Remco Evenepoel, a rider who is not exactly known for folding under pressure. Isaac Del Toro was third, nearly ten minutes back. It was one of the more one-sided Tours in years.

The easy explanation is the one you'll hear on every broadcast: he's just stronger. Bigger engine, higher numbers, more raw power than anyone else in the race. That explanation isn't wrong, but it's incomplete in a way that matters if you're trying to learn anything from it. Fresh power gaps at altitude explain a stage. They don't explain a three-week margin that widens instead of shrinking. For that you need a different variable, one that gets talked about far less than FTP but predicts long-event outcomes just as well, maybe better: durability.

What Durability Actually Means

Durability is a specific, measurable thing, not a synonym for "toughness." It's the size of the gap between what you can produce fresh and what you can still produce after hours of accumulated fatigue. Two riders can have identical 20-minute power numbers on a rested Tuesday and finish a five-hour stage in completely different states, because one of them loses 6% of that number after the fatigue and the other loses 12%.

That 6% versus 12% isn't a made-up round number. A 2025 study in Frontiers in Sports and Active Living tested competitive amateur cyclists doing 20-minute maximal efforts fresh, then again after burning 1,000 kJ of work at moderate intensity to simulate race fatigue. The more successful riders in the group held their power to within about 6.5% of their fresh number. The less successful group dropped 12.5%, nearly double. Same fresh fitness pool, different fade under load. The researchers' training suggestion was specific too: put race-pace efforts late in already-long sessions, and rehearse fueling deliberately enough that glycogen isn't the thing that ends your day.

That's the mechanism a 6:26 margin points to. Nobody wins a Grand Tour by 6 minutes on day-one legs. They win it by being the rider whose watts on stage 18 look the most like their watts on stage 3, while everyone around them is negotiating with their own fatigue.

This Isn't Just a Grand Tour Problem

Here's where it's tempting to file this under "interesting pro cycling trivia" and move on, and here's why that would be a mistake. Durability shows up at every level, on every kind of long day, whether or not there's a number pinned to your jersey.

The gran fondo that gets hard in the last 30 miles. The Sunday group ride where you're fine for two hours and then quietly get shelled on the climb everyone else seems to float up. The marathon where the wheels come off at mile 20 despite a training block that looked solid on paper. None of those are FTP problems in the moment they happen. They're durability problems: the engine was never the limiter, the ability to keep using it after three hours of accumulated fatigue was.

This matters just as much if you have no race on the calendar at all. If your goal is simply riding more, feeling stronger doing it, and not getting wrecked by the long Saturday ride your friends do every week, durability is arguably the more relevant number of the two. Nobody asks a general-fitness rider for their FTP at the coffee stop. Whether they can still hold a conversation on the last climb is the thing that actually shapes whether the ride was fun or a survival exercise, and whether they show up again next week.

Training the Fade, Not Just the Peak

The training implication tracks what the study found and what shows up across endurance coaching more broadly: durability responds to a different stimulus than peak power does, and it responds slowly. You don't build it with another set of 5x5 threshold intervals on fresh legs. You build it by doing quality work after you're already tired.

A few patterns that actually move this number:

Long rides with intensity in the back half, not the front. Two to three hours easy, then tempo or sweet-spot work in the final 45-60 minutes, is a fundamentally different stimulus than the same intervals done first thing. The legs that have to produce watts on top of 2 hours of accumulated fatigue are the legs you're trying to train.

Fueling practice as a training variable, not an afterthought. A meaningful share of late-ride fade is glycogen, not muscle. If you've never practiced eating on the bike at the rate a long day actually demands, that's not a durability problem to fix with more intervals, it's a fueling problem to fix by rehearsing it on the rides where it counts.

Consistency over months, not intensity over weeks. This is the same principle behind why CTL is treated as a capacity metric rather than a goal to chase directly: durability is built by accumulated, well-recovered volume over a long runway, and rushing the ramp rate to get there faster tends to produce the opposite of the intended adaptation. There's no durability shortcut equivalent to a hard interval session; the aerobic base underneath it has to actually be there.

Testing the fade honestly. You don't need a lab to get a rough read. Note your power or pace, and how it's trending relative to heart rate and effort, in the final third of your longest rides. If pace or power is dropping while heart rate and RPE are climbing, that's the fade the research is describing, and it's telling you something specific about where to spend training time next.

That last one is exactly what the efficiency factor trend chart on Fitness Trends is built to surface: power per heartbeat across comparable aerobic sessions, tracked over time, with the hard and surgy rides deliberately excluded because they'd muddy a comparison that only means something between like-for-like efforts. And on any ride that has a real measured normalized power, the app already breaks out variability index alongside it, which is a decent proxy for how surgy versus steady an effort actually was, useful context when you're trying to tell durability fade apart from just having ridden a spikier route. If you're training toward a general fitness goal rather than an event, the app tracks consistency streaks instead of a finish-line countdown, which is really the same underlying bet: the accumulated, unglamorous weeks are what's actually buying the capacity, not any single session.

The Honest Limits Here

A word on how far to take this. Pogačar's margin is one data point from one race, shaped by team tactics, a course, weather, and a field that included a late withdrawal after a crash, not a controlled experiment. The 6.5%-versus-12.5% figure comes from 14 amateur cyclists in one study, a real and useful signal, not a law of physiology carved in stone, and durability research is still a young and active area. Nobody should read this and conclude their next block should be all back-loaded tempo work with no fresh-legs intensity; peak power still matters, durability is additive to it, not a replacement for it. And none of this is a substitute for actual coaching judgment about your own fatigue, injury history, or how your body responds, which is a conversation to have with a coach or a professional, not a blog post.

What the Tour did do well is put a very visible number on something that's usually invisible: the gap between what you can do fresh and what you can still do tired. That gap is trainable. It's just trained differently than the number most of us default to chasing.