REF-899 · Sánchez-Jiménez et al., 2026 · Sport Sciences for Health, Vol. · Paper

Pacing strategies in Gran Fondo cycling: a large-scale analysis of the Quebrantahuesos race

B+moderate evidenceMeasured on recreational to trained athletesWritten before October 2026, being re-checked

Pacing in mass-participation gran fondos is poorly described. Split times from the 2025 Quebrantahuesos were collected from public web data for 6589 finishers, and pacing range, split-level relative speed and the share of race time in each segment were compared across finishing-time groups and clustered with k-means.

Pacing variability differed by performance level: the fastest riders (under 6:00 h) had the lowest pacing range, started more conservatively and pulled further ahead in later segments, especially at the finish, and spent proportionally less time on the decisive mid-race climbs.

The best-performing cluster was slower early and stronger on the major climbs; the highest-variability cluster slowed mid-race and lost time in the final segment. The authors conclude that performance went with lower pacing variability and segment-specific strategy.

Participants

n=6589 finishers of the Quebrantahuesos Gran Fondo, 2025 edition.

Protocol

Publicly available split-time data collected by web scraping.
Pacing metrics: pacing range, split-level relative speed, and percentage of total race time spent in each segment.
Differences tested between performance-based finishing-time groups; k-means clustering on relative speed, time distribution and pacing range.

Key Results

Pacing variability differed significantly across performance levels (p < 0.001); the fastest cyclists (< 6:00 h) had the lowest pacing range of all groups.
Faster cyclists started more conservatively and became progressively more differentiated from slower cyclists in later segments, particularly at the finish.
Faster cyclists spent proportionally less time in the decisive mid-race climbs and showed the greatest separation in the final segment.
Three pacing clusters: the best-performing cluster had lower relative speed early and superior performance on the major climbs; the highest-variability cluster had slower mid-race speeds and more time accumulated in the final segment, which the authors read as suboptimal effort distribution or premature fatigue.
The abstract gives no speeds or segment times.

Practical Application

In 6589 finishers of the 2025 Quebrantahuesos gran fondo, the fastest riders (under 6:00 h) paced most evenly, started relatively conservatively and pulled away on the major climbs and in the final segment, while the most variable pacers slowed mid-race and lost most time at the end. To "how should I pace a gran fondo?": hold back early relative to your own average, save it for the big climbs, and aim to still be moving well in the last segment.
The finding is an association in finishers only. Fitter riders can both go faster and pace more evenly; the study does not show that pacing evenly makes a given rider faster.
The authors' reading of the high-variability cluster is "suboptimal effort distribution or premature fatigue"; the data are split times, so fuelling and fatigue were not measured.
Not from this paper; the base's own practice: pace below fresh thresholds and fuel for the back half, as in SYNTH-50; the same fast-start penalty appears in recreational marathoners (REF-820).

Limitations

One edition of one course; the results may be course-specific.
Split times only: no power, heart rate, group riding, weather or fuelling data.
Finishers only; riders who did not finish are not included.
Observational; clusters describe patterns, not causes.
Not indexed in PubMed or Europe PMC; the abstract was taken from the Crossref record.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

large observational dataset, n=6589