REF-829 · Hoffman, 2014 · International Journal of Sports Physiology and Performance, Vol. · Paper · PubMed

Pacing by Winners of a 161-km Mountain Ultramarathon

Blimited evidenceMeasured on elite to world-class athletesWritten before October 2026, being re-checked

Segmental speed and the coefficient of variation (CV) in speed were analysed over 10 race segments for 24 editions of the 161-km Western States Endurance Run (1985 through 2013). Segment speeds did not differ between the eventual winners and the lead runners, and differed between 1st and 2nd finishers only in the second half.

Winners had a lower mean CV in speed (12%) than the other top-5 finishers (14-15%). Among the 10 fastest finish times, CV in speed was related to finish time (r = .80). For the top-5 runners, CV in speed was related to maximum ambient temperature (coefficient = .14) and calendar year (coefficient = -.086).

The author concludes that mountain trail running involves wide variations in speed, but the fastest times come when fluctuations are limited: winners stayed relatively close behind the leaders, took the lead in the middle half, and then slowed less than the other top runners late in the race.

Participants

The winners and other top-5 finishers of 24 editions of the Western States Endurance Run, 1985 through 2013. The abstract does not give sex or the number of individual runners.

Protocol

Segmental speed over 10 race segments; coefficient of variation (CV) in speed per runner.
Comparisons: winners vs lead runners; 1st vs 2nd finishers; winners vs other top-5 finishers.
Correlation of CV with finish time (10 fastest finish times); multiple linear regression of top-5 CV on maximum ambient temperature and calendar year.

Key Results

Mean CV in speed: winners 12% vs other top-5 finishers 14-15% (P < .01).
CV in speed related to finish time among the 10 fastest finish times (r = .80, P = .006).
Segment speeds: no difference between eventual winners and lead runners; 1st vs 2nd finishers differed only in the second half.
Top-5 CV in speed rose with maximum ambient temperature (coefficient = .14, P < .05) and fell across calendar years (coefficient = -.086, P = .034).

Practical Application

Across 24 editions of the 161-km Western States Endurance Run, winners varied their speed less than the other top-5 finishers (CV 12% vs 14-15%), and among the fastest finish times a steadier pace went with a faster time (r = .80); winners sat close behind the leaders early, moved ahead in the middle half and then slowed less late on. In a 161-km mountain race, the aim is not to be in front early but to fade least: hold back in the first part, keep speed losses small, and expect the race to be won in the second half.
Speed variation rose with heat. Inference, not tested here: on a hot day, pace more conservatively and expect larger swings; the terrain already forces variation, so avoid adding more by chasing.
Speed varies by terrain on any mountain course; a low CV here is relative to other elite runners on the same course. Not from this paper; the base's own practice: pace a mountain ultra by effort and by minimising stops (REF-664, SYNTH-61), not by a target speed.

Limitations

One race; course, heat and field depth are specific to it.
Top finishers only; the sex of runners and the number of individuals are not given in the abstract.
Correlational; pacing and finish time share the same race data.
Short report (three pages); written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

elite race-record series, one race; correlational