A study published on 25 June 2026 in Sport Sciences for Health tracked the split times of 6,589 finishers of the 2025 Quebrantahuesos. Riders who finished inside six hours rode the opening 47.5 km more conservatively than everyone behind them, then pulled away on the highest climbs and in the final section of the 197 km route.
How the Quebrantahuesos study measured pacing
Jose Luis Sánchez-Jiménez, Jose Ignacio Priego-Quesada and Fran Oficial-Casado of the University of Valencia collected the public timing data for all 7,752 participants of the 2025 edition and kept the 6,589 riders who finished with a complete set of splits. That sample was 6,360 men and 229 women. The paper is open access, and the dataset is published alongside it.
The course runs 197.2 km with 3,190 m of climbing, starting and finishing in Sabiñánigo and crossing the Pyrenees into France and back. Seven timing points split it up, at Somport (47.5 km), Plateau de Benaiu (103.2 km), Lac de Fabregues (138.3 km), Portalet (148.0 km), Panticosa (162.8 km), Biescas (180.3 km) and the finish.
Riders were grouped by finishing time in 30 minute bands, from the 211 who finished in under six hours to the 2,140 who took eight and a half hours or more. For each section the researchers expressed a rider's speed as a percentage of that rider's own average speed for the day. So a fit rider and a slower one are compared on how they spread their effort, and absolute speed drops out of it. The gap between a rider's relatively fastest and relatively slowest section gave a single number for how uneven their day was.
How did the fastest riders pace the Quebrantahuesos?
| Finishing time | Riders | Pacing range (points) |
|---|---|---|
| Under 6:00 | 211 | 116.7 |
| 6:00 to 6:30 | 648 | 124.4 |
| 6:30 to 7:00 | 920 | 128.2 |
| 7:00 to 7:30 | 1,065 | 130.9 |
| 7:30 to 8:00 | 955 | 129.4 |
| 8:00 to 8:30 | 920 | 127.7 |
| 8:30 or more | 2,140 | 128.9 |
Pacing range is in percentage points of each rider's own average speed. The sub-six-hour group was the steadiest by a clear margin, while the neighbouring groups further down the field barely differed from each other.
On the run to Somport, the two fastest groups rode at a lower speed relative to their own day than every slower group, and those first 47.5 km took up a larger share of their total ride time. The effect sizes there were large, between 1.6 and 2.0.
On Lac de Fabregues and Portalet, the steepest sections and the highest point of the route, the fastest riders spent a smaller share of their day than slower riders did. The final section into Sabiñánigo separated the groups most of all.
Should you pace a sportive evenly?
The team also let an algorithm sort riders by the shape of their pacing alone, ignoring finishing time, and it found three distinct profiles.
The fastest profile had the slowest relative start of the three profiles, then the quickest speeds through Lac de Fabregues and Portalet. Riders with that profile finished on average 4,917 seconds, about 82 minutes, ahead of the profile with the most uneven pacing. They also finished 6,743 seconds ahead of the profile with the evenest pacing, the one whose speed varied least from section to section.
Our view at WattPlan is that a controlled first hour is the most reliable pacing decision an amateur can make at a mountain sportive, because that hour is where the adrenaline and the fast groups push riders into efforts they pay for later in the day. Our guide to race pacing covers the physics of why power belongs on the climbs.
How to pace your next sportive
- Find the climbs that decide your event. On the Quebrantahuesos they came in the second half, on the steepest gradients and at the highest altitude, while the first climb, Somport, came at 47.5 km, where the fastest riders were holding back. Mark the equivalent climbs on your own course profile before you plan anything else.
- Set a ceiling for the opening section based on the power you can hold for your whole expected ride time, which on a long sportive sits well below FTP. WattPlan's Route plan works this way for any ride over an hour. Upload the event GPX and it sets each segment's target from the power you can sustain for the full predicted duration, warns you when W' is running low, and exports the plan as a GPX with pacing waypoints for your bike computer. Route plan is part of Premium.
- Train the late climb. Put the hardest climbing effort of your long rides in their final third, so you practise climbing well on tired legs. Your Durability Score estimates what you can hold at three and five hours from the long rides you have already done, which tells you how much the late climbs will cost you. Durability Score is also part of Premium.
- Read your own splits afterwards. If your event publishes split times, divide your speed for each section by your average speed for the day. If your first section was your relatively fastest, the time to find next year is in the opening hour.
What the data can't tell you
Split times record speed, with no power or heart rate behind them. A fast opening section may reflect a big group and a tailwind rather than hard riding, and descending skill changes speed without any extra effort. The authors list the missing power data and descending ability among their limitations.
It is one edition of one event, on a course whose biggest climbs fall late. A sportive with a hard climb in the first 20 km could reward a different shape.
The study shows an association and stops short of cause. The fastest riders may start conservatively because they are fitter and can afford to, rather than going faster because they held back. Finishing time also explained only about 2% of the variation in how uneven riders' pacing was, so every time band held riders who paced well and riders who paced badly. The authors call their three profiles exploratory.
With 229 women among 6,589 finishers, the study says little about how women paced this course. The paper does note that women were most common in the evenest of the three profiles.