La Marmotte 2026 — every split of every rider

Two clocks, and only one of them ranks

Every rider on this course finishes with two different times, and almost nobody explains the difference. The gross time is what their own computer showed: start mat to finish line, everything included. The net time subtracts the two neutralised sections — and it is the net time that the classification is built on.

The three cols, and the Alpe

Each climb is timed between two mats and ranked on its own. The timekeeper publishes no per-section classification for the Marmotte. Every segment leaderboard here is derived from its own mat reads — the times are official, the ranking is ours.

The two sections nobody races

Two sections of the course are neutralised: the Glandon descent (km 39 → 57) and the La Grave traverse (km 132 → 134). The time taken on them is credited back, so in the official — net — classification they last zero seconds. No leaderboard here crosses one, which is why the seven segments cover 157 of the 177 km: the missing 20 km are exactly those two sections.

What would it have taken? 🎯

Pick a place — overall or in your category — and this gives the time that produced it in 2026, plus what the riders around it actually did on each col.

Every timed segment

Seven spans covering 157 of the 177 km. The missing 20 km are not an oversight — they are exactly the two neutralised sections, which in net time last zero seconds and cannot be ranked or turned into a speed.

The timekeeper publishes its own course kilometres for eleven of the twelve mats ("39km Glandon"), so distances here are measured, not interpolated — they are integers, so read ±0.5 km into each. The one exception is the preview mat: its label says km 176, but it sits a median of 22.7 seconds before the line — about 50 metres — so it is placed at 176.95 and marked ≈.

Category podiums

Top 5 of each category — click through for the full ranking. The timekeeper publishes category codes (M18…M70, F18…F65) and no rulebook. The age bands shown here are read off the 500 ages it does publish, and they are brackets by year of birth — which is why the youngest M18 on record is 17 and the oldest 28. 15 classified riders carry no category at all (the timekeeper files them under XXX); they have no band page and are not in the grid below.

Spotted in the peloton 🌟

Riders already identified for l’Étape du Tour 2026 who also turned up here, matched on name and age band. That makes this list narrow on purpose: it is notable people we already knew about who also rode the Marmotte, not notable people in the Marmotte — nobody has done the press pass on this field.

† = the cross-race matcher rates this link short of certain. Identities were established for l’Étape from press and the official race report; only the “same person” link is new here, so a homonym is still possible. Spotted someone we missed? Tell me.

The Marmotton

The Marmotton is not a separate race. It is where the Marmotte puts you when you miss a time cut — chiefly the barrier at the foot of Alpe d’Huez. The ride that remains is the Marmotte without the final 21 bends, about 163–165 km instead of 177. The timing mat at km 163 is called Marmotton because that is where you become one. 341 riders were classified in it.

Is it a start-time problem?

The barrier is a clock time but the start is staggered by bib, so a rider who crosses the line at 8:30 has two hours less to reach the foot of the Alpe than one who left at 6:30. Tempting story — and the raw numbers look damning: the share downgraded climbs from 1.6% in the lowest bib band to 17.4% in the highest.

Most of that is not the start, though. Bibs are seeded, so a low number is a faster rider before anyone starts a clock. Among riders who were classified, the median finish goes from 7:31:49 in the first band to 10:46:10 in the last — a spread of more than three hours. Compare the raw bands and you are mostly comparing fast riders with slower ones.

So hold ability fixed. Every rider is timed at the Col du Glandon, km 39, measured from their own start — it reflects how they ride, not when they left. Grouping by that:

The effect is real, but it lives in the middle of the field. Ride to the Glandon inside two hours and the barrier never threatens you whatever your number. Take more than two and a half and you are at risk regardless — a quarter of the low bibs went too. It is the 2,968 riders in between, all climbing at the same speed, where the bib decides: 6.0% of those starting from the back were downgraded against 1.1% of the rest, ×5.6 — the same legs, two hours apart on the start line.

315 of the 341 Marmotton riders have a Glandon split to match on. The mechanism is a rider’s own account of an organiser warning him the foot of the Alpe shut at 18:15 — so “be at the Galibier by 17:15”, having started at 8:30 on bib 5,075. That is one rider’s telling, not a published rule, and a bib band is a crude proxy for a start time we do not have per rider. Nicolas, who knows the sport, points out the other half of it: low numbers also go to riders who earned a good placement.

Read these times as live tracking, not an official result: the timing provider publishes Live and Entries for the Marmotton but no Official Results table, unlike the Granfondo. A handful of finishers share a time with someone else, and some split chains disagree with their own finish — normal for a live feed, and the reason nothing here is ranked against the Granfondo. The downgrade rule itself is the organiser’s, relayed by a rider rather than published: treat it as well-sourced, not official.

The same climb, fresh and after 164 km

The Grimpée de l’Alpe runs on the Friday: the same 13 km up to Alpe d’Huez, on fresh legs. 75 riders did both — the hill climb on Friday, then the same climb on Saturday with the Glandon, the Télégraphe and the Galibier already in them. It is a controlled experiment nobody designed: same riders, same road, one variable.

The ten who held up best and the five who suffered most, of 75. Ratios below 1 mean they climbed it faster tired than fresh — which usually means Friday was ridden as a warm-up, not a race. Matching is by bib: the Grimpée adds 100,000 to a Marmotte rider’s number, so 83 of its 209 finishers are Marmotte riders and 75 of those have a timed Alpe split on the Saturday too.

Times we hold back 🏍️

2 segment times out of 33,018 are physically impossible for the span. They stay visible and are excluded from every ranking and stat here.

Times that are physically impossible for the span are kept visible but excluded from the rankings and the stats. On this race the rule catches two rows out of 33,018 — a 62 km/h average up 39 km to the Glandon is a mat read stamped with the wrong clock, not a human. It means “not credible”, not “cheated”.