REF-127 · De Koning et al., 1999 · Journal of Science and Medicine in Sport, Vol. · Paper · PubMed

Determination of Optimal Pacing Strategy in Track Cycling with an Energy Flow Model

Blimited evidenceMeasured on a mixed groupWritten before October 2026, being re-checked

Protocol

Simulation study: the 1000 m time trial and the 4000 m pursuit in track cycling were simulated with a model based on the flow of energy in cycling (from the abstract)
Different strategies for distributing the available anaerobic energy were evaluated
Model predictions of split times and final times were compared with values achieved by cyclists during championships; no participants were tested

Key Findings (summary)

De Koning, Bobbert & Foster simulated the 1000 m time trial and the 4000 m individual pursuit with a model based on the flow of energy in cycling, evaluated different ways of distributing the available anaerobic energy, and compared predicted split and final times with values achieved by cyclists at championships.

For the 1000 m, the best result came from the cyclist with the highest anaerobic peak power using an all-out strategy. For the 4000 m, the fastest time came from an all-out start at a high initial power, followed by a constant anaerobic power output after 12 seconds — which yields an evenly paced race.

Even small variations in pacing strategy had substantial effects on performance, so individual athletes may gain by experimenting with small pacing variations to find what works for them under specific conditions.

Model Results

1000 m time trial: all-out strategy; the highest anaerobic peak power wins.
4000 m pursuit: all-out start at high initial power → constant anaerobic power output after 12 s → evenly paced race.
Small changes in pacing strategy produce substantial changes in finishing time.
Long-distance events were NOT modelled in this paper.

Practical Application

For a 1000 m track time trial the model's fastest strategy is all-out from the gun, and for the 4000 m pursuit it is an all-out start at high power followed, after about 12 seconds, by a constant anaerobic power output that produces an evenly paced race — with small deviations in pacing costing measurable time.
The transferable rule for events of roughly 1-5 minutes: spend the anaerobic reserve hard at the start to get up to speed, then hold an even effort rather than surging or fading; rehearse the exact split pattern, because small pacing errors are expensive.
Inference, not tested: the paper models only 1000 m and 4000 m; the base's default of even pacing below critical power for events longer than about 40 minutes is a different body of evidence, not this model.
Not from this paper: for variable-power pacing on hilly time-trial courses see REF-126 (Boswell 2012).

Limitations

Two track events only; a simulation model validated by comparison with championship times, not by an intervention; no aerodynamic or tactical factors beyond the energy-flow equations.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from A: modelling study; predictions checked against real championship times but no experimental manipulation

Content correction 2026-09-27
"Long distances (> 40 min): even pacing maintaining < CP" -> not in the paper, which models only the 1000 m and 4000 m events; relabelled as the base's own inference from other evidence.
Added the abstract's method (energy-flow simulation compared with championship times) and its message that small pacing variations have substantial effects. Level A -> B (modelling); Sport general -> cycling.
2026-09-28: added Protocol from the abstract (Europe PMC/PubMed PMID 10668763).