Variable Power Pacing for Rolling Terrain: Sports Science Evidence on Slope Effects and Downhill Gliding Optimization

1. Introduction and Cutting-Edge Research Background

In the competitive world of cycling time trials and triathlon bike legs, “pacing” has always been the critical variable determining victory or defeat. Traditional views have long advocated that “Constant Power Pacing” maximizes physiological efficiency, a concept rooted in classic exercise physiology theories regarding “physiological metabolic stability”—when exercise intensity is maintained at a fixed percentage of Functional Threshold Power (FTP), the body’s acid-base balance and muscle glycogen depletion rate remain relatively stable, which was considered the golden rule for achieving optimal performance. However, in recent years, with the proliferation of power meters and the accumulation of large-scale race data, this ironclad rule has begun to face serious challenges.

Looking back at the historical context, in 2010, a sports science team from Boston University first published simulation research on “variable power output” on undulating terrain, pointing out that on sections with dramatic gradient changes, a fixed-power strategy would result in severe competitive disadvantages. Subsequently, UCI-certified computer simulation software such as BestBikeSplit and Golden Cheetah emerged, enabling sports scientists to use “numerical integration” to calculate a rider’s speed, wind resistance, and gravitational interactions second by second, thereby confirming: on undulating terrain, “overpowering” on uphill sections and reducing power (coasting/reduced power) on downhill sections can significantly shorten total finishing time.

The latest scientific findings come from a 2022 meta-analysis published in the Journal of Science and Cycling. This study compiled power data from 14 UCI time trials and found that top professional riders on sections with a gradient exceeding 3% often produced average power 5-8% higher than the race’s overall average power; conversely, when the gradient dropped below -3%, power output plummeted to 60-75% of the overall average. This finding shattered the “constant pacing” myth, instead supporting a new paradigm of “Grade-Responsive Power Modulation.”

For Taiwanese cyclists, this issue is particularly relevant. Whether tackling the 87-kilometer sustained climb of Eastbound Wuling, the steep hairpin turns of Westbound Wuling, the rolling hills of Yangmingshan’s “Wind Sword” route, or the continuous alternation of short climbs and descents on the Hualien-Taitung (Huatung) circuit, constant power pacing often leaves riders feeling powerless on climbs and wasting precious energy through excessive output on descents. This article will use rigorous physical modeling and real-world data to construct a “variable power pacing method” suitable for Taiwan’s terrain.

2. Core Mechanisms of Exercise Physiology and Biomechanics

2.1 Mathematical Model of Cycling Dynamics

To understand the advantages of variable power pacing, we must first establish the complete mechanical equations governing the motion of a rider and bicycle on gradients. On flat sections, the rider’s power output ( P ) primarily overcomes three major resistances: rolling resistance, aerodynamic drag, and drivetrain mechanical friction. However, once entering a slope, the gravitational component becomes the dominant source of resistance.

Total cycling power ( P_{total} ) can be expressed as:

[
P_{total} = \left( m \cdot g \cdot \sin\theta + C_{rr} \cdot m \cdot g \cdot \cos\theta + \frac{1}{2} \cdot \rho \cdot C_d \cdot A \cdot v^2 \right) \cdot v + P_{drivetrain}
]

Where ( m ) is the combined mass of rider and bicycle (kg), ( g ) is gravitational acceleration (9.81 m/s²), ( \theta ) is the gradient angle, ( C_{rr} ) is the rolling resistance coefficient (typically 0.004-0.006 for road bikes), ( \rho ) is air density (approximately 1.225 kg/m³ at sea level), ( C_d \cdot A ) is the aerodynamic drag area (typically 0.25-0.35 m² for a time trial position), and ( v ) is forward velocity (m/s).

The key insight lies in the fact that the gravitational term ( m \cdot g \cdot \sin\theta \cdot v ) is proportional to the sine of the gradient angle. When the gradient increases from 2% to 6%, (\sin\theta) increases from 0.02 to 0.06, instantly tripling gravitational resistance. To maintain the same speed, power requirements rise dramatically. Conversely, on descents, (\sin\theta) becomes negative, and gravity acts as a propulsive force. If a rider continues to output high power at this point, speed will surge non-linearly, but aerodynamic drag (proportional to ( v^3 )) will also increase sharply, leading to severely diminishing “marginal returns” on power.

2.2 Non-linear Power-Speed Relationship and Time-Saving Effects

Calculus plays a crucial role here. We divide the course into ( n ) infinitesimal segments ( ds ), where the riding time for each segment is ( dt = ds / v(s) ), and total finishing time is:

[
T_{total} = \int_{0}^{S} \frac{1}{v(s)} , ds
]

Since speed ( v ) is a non-linear function of power ( P ) (especially in high-speed zones where aerodynamic drag causes speed to grow only with the cube root of power), this means that “outputting an extra 10 watts on a low-speed section (uphill)” and “outputting an extra 10 watts on a high-speed section (downhill)” yield vastly different time benefits. Through optimization using Lagrange multipliers, we can prove that to minimize total time ( T_{total} ), extra power should be allocated to the segments with the highest “speed sensitivity”—namely, uphill sections, because speeds are low there, and a small power increase yields significant time savings; on descents, where speeds are already high, extra power only yields minimal speed gains.

Specifically, on a 6% uphill section, if a rider is climbing at 15 km/h, a 5% power increase (approximately 15 watts) can boost speed by about 3-4%; however, on a descent at 55 km/h, the same 5% power increase might only improve speed by 0.8-1.2%. The former saves far more time than the latter—this is the physical foundation of “variable power pacing.”

2.3 Physiological Metabolic Adaptations

From an exercise physiology perspective, the “overpowering” on uphill sections temporarily increases the rate of muscle glycolysis and lactate production. However, since this output only lasts 3-8 minutes (depending on climb length), the body can compensate through the “hydrogen ion buffering system” and “phosphocreatine (PCr) resynthesis” mechanisms without causing severe acidosis. In contrast, the “reduced power coasting” on descents provides an opportunity for active recovery, allowing heart rate to drop into the aerobic zone (Zone 2), promoting the clearance of metabolic waste products and preserving neuromuscular recruitment capacity for the next climb. Research shows that this “high-low alternating” power rhythm actually enables riders to sustain a higher average power in the latter half of a race. Although overall average heart rate is slightly higher, the “power-to-heart rate ratio (P/W)” is more economical.

3. Key Parameter Field Testing and Comparative Analysis

3.1 Field Test Data Sources and Methods

To validate the theoretical model, we referenced power data from 40 trained amateur riders (average FTP 280W, body weight 70kg) recorded in international literature on a simulated undulating course (total length 40km, cumulative elevation gain 480m). Subjects completed the test using both “constant power pacing” (locked at 75% FTP = 210W throughout) and “variable power pacing” (85-90% FTP on climbs, 75% FTP on flats, 50-60% FTP on descents), with finishing times and physiological markers recorded.

3.2 Data Comparison Table

3.3 Advanced Comparison: Time Savings Across Different Gradient Combinations

The data above clearly shows that the more dramatic the gradient changes on a course, the greater the advantage of variable power pacing. Mountainous courses can yield time savings of up to 3.2%, which closely aligns with simulations from BestBikeSplit software. Notably, even with a slightly lower overall average power (203W vs 210W), variable pacing still produces faster finishing times, completely overturning the traditional notion that “average power determines performance.”

4. Periodized Training Plans and Power Tuning Guidelines

4.1 Establishing Your Personal Grade-Responsive Power Curve

Before practical application, riders must first complete a crucial task: establishing their personal “grade-power response curve.” First, perform a 20-minute FTP test on a trainer to obtain your baseline threshold power. Next, find an outdoor route containing four distinct gradients—2%, 4%, 6%, and 8%—and ride each for 5 minutes at a “hard but sustainable” pace, recording your maximal mean power (MMP) for each gradient. These four data points will form the foundation of your personal grade-specific power baseline.

4.2 Grade-Responsive Power Zone Settings

4.3 Specialized Training Plan (Eight-Week Periodized Program)

Phase 1 (Weeks 1-2): Gradient Adaptation Period

• Tuesday: Flat endurance ride, 2 hours, heart rate Zone 2, power 65-75% FTP.
• Thursday: Hill interval training (6 x 3 minutes, gradient 4-5%, power 90% FTP, 3 minutes rest).
• Saturday: Long hilly ride, 3 hours, simulating the “overpower uphill, reduce power downhill” rhythm.

Phase 2 (Weeks 3-5): Power Modulation Intensification Period

• Tuesday: Variable power tempo ride, outputting 85% FTP on 1-3% gentle climbs and reducing to 50% on descents, lasting 90 minutes.
• Thursday: Uphill overreach training (6 x 90 seconds, gradient 6-8%, power 105% FTP, 4 minutes rest).
• Saturday: Simulated course training, selecting a route with multiple gradient types (such as Yangmingshan’s Wind Sword), fully executing the variable power pacing strategy.

Phase 3 (Weeks 6-8): Simulation and Optimization Period

• Tuesday: Time trial simulation (40 minutes), using grade-responsive power settings throughout, targeting a pacing accuracy of ±3%.
• Thursday: Recovery ride, 1 hour, heart rate strictly controlled in Zone 1-2.
• Saturday: Full race simulation (distance matching your target event), incorporating nutrition strategy, recording power data throughout for post-race analysis.

5. Race Nutrition, Environmental Adaptation, and Race-Day Strategies

5.1 Quantified Energy Nutrition Strategy

The power fluctuations accompanying variable power pacing place unique demands on energy metabolism. The overreach output on climbs accelerates glycogen depletion, while the recovery periods on descents provide a golden window for fueling. Recommended carbohydrate intake strategy:

• 2 hours before the race: Consume 1.5-2.0 g/kg body weight of low-fiber carbohydrates (such as white toast with jam) to ensure adequate glycogen stores.
• Every hour during the race: Target 60-90 grams of carbohydrates (adjusted based on race intensity). During the gentle descent sections following climbs, consume gels or energy bars (15-25 grams every 15-20 minutes), as gastric blood flow is more abundant at this time, optimizing absorption efficiency.
• Electrolyte supplementation: Consume 500-700mg of sodium per hour, achievable through sports drinks or salt tablets, especially during long-distance challenges in high temperatures (such as the 300km Huatung circuit).

5.2 Hydration Strategy and Climate Adaptation

In Taiwan’s hot and humid summer environment (such as the July Wuling Challenge), the high-speed coasting on descents creates a strong wind-chill effect, causing riders to underestimate fluid loss. It is recommended to force 2-3 sips of water (approximately 50-80ml) at the start of each descent and use the brief respite on descents to check remaining bottle levels. During cold or rainy seasons on Yangmingshan, power on descents should be reduced to 40-50% FTP to maintain core temperature, paired with a windbreaker jacket to avoid muscle stiffness from hypothermia.

5.3 Practical Application on Taiwan’s Classic Courses

• Eastbound Wuling (87km, 2800m elevation gain): The first 50 kilometers feature gentle gradients; maintain a steady output of 75-80% FTP to conserve energy. After Wushe, the gradient steepens; upon entering 6-8% sections, increase power to 90-95% FTP, and deliberately reduce to 50% for 30 seconds of recovery on each short descent following the switchback corners.
• Yangmingshan Wind Sword (rolling hills): This route features frequent and abrupt gradient changes; a “wave-style” pacing approach is recommended—88-92% FTP on climbs, 45-60% FTP on descents. Avoid excessive output on descents even with tailwinds.
• Huatung Circuit: Wind resistance significantly impacts the coastal sections. On tailwind sections, reduce power to 65% FTP to conserve energy; on headwind sections, increase to 85% FTP, and make good use of terrain shielding effects.

6. Common Operational Pitfalls and Scientific Myth-Busting

Myth 1: “Constant power is the iron rule; fluctuations will cause premature fatigue”

This is the biggest misconception. Research shows that as long as the frequency and amplitude of power fluctuations are kept within a reasonable range (uphill overreach not exceeding 105% FTP, duration not exceeding 10 minutes), the body’s buffering systems can fully cope. What truly causes premature fatigue is sustained output above FTP for extended periods (exceeding 20 minutes), not brief power surges.

Myth 2: “You should pedal hard on descents, using gravity to create higher speeds”

This view ignores the cubic relationship between aerodynamic drag and speed. When speed exceeds 50 km/h, an extra 50 watts only yields approximately 1-1.5 km/h in speed, translating to mere seconds of time saved, but the metabolic energy expended is substantial, leading to severe power decay on subsequent climbs. The correct approach is to maintain a light gear and spin smoothly on descents (approximately 50-60% FTP), preserving muscle temperature and neural activation, rather than sprinting at full effort.

Myth 3: “Variable power pacing is only suitable for climbing specialists”

The opposite is actually true. Heavier riders (high-power types) possess a natural power advantage on climbs but are disadvantaged on descents due to their larger aerodynamic drag area. Through variable power pacing, heavier riders can convert their uphill overreach output into potential energy, trading it for speed through coasting on descents, thereby narrowing the gap with lighter riders. Lighter riders, on the other hand, should carefully control their overreach amplitude on climbs (not exceeding 5%) to avoid premature exhaustion from insufficient muscular strength.

Myth 4: “Higher power numbers are always better; average power is the only metric that matters”

On undulating courses, the meaning of average power (AP) and normalized power (NP) differs significantly. Riders using variable power pacing typically have a lower AP than constant pacers, but their NP may be higher, as the NP algorithm weights high-power segments more heavily. This does not mean variable pacing is inferior; rather, it demonstrates that NP better reflects actual metabolic load. Riders should use NP rather than AP as their fatigue management metric, targeting an NP range of 85-90% FTP.

7. Expert FAQ

Q1: How do I determine if my grade-responsive power settings are correct?

A: The most effective method is to conduct an “A/B test.” Choose a fixed 20-30 kilometer route containing multiple gradient types, ride it once using constant power and once using variable power (with at least 48 hours between rides to ensure full recovery), and record finishing time and average power. If the variable power ride produces a faster finishing time, even with slightly lower average power, your settings are effective. Additionally, you can use BestBikeSplit software, inputting your personal data (FTP, weight, drag coefficient) for simulation; the software will automatically calculate the optimal power distribution curve as a race-day reference.

Q2: When overreaching on steep climbs (>8%), will it cause muscle cramps or injury?

A: Muscle cramps are typically associated with electrolyte imbalance, dehydration, or excessive fatigue, rather than power output alone. When performing overreach at 95-105% FTP on steep climbs, ensure: 1. Adequate sodium and magnesium supplementation before the race; 2. Maintaining a steady breathing rhythm during the climb (one deep breath every 2-4 pedal strokes); 3. If the gradient exceeds 10%, consider reducing the overreach amplitude to 90-95% FTP and increasing cadence (90-100 rpm) to distribute muscular load. If you feel abnormal muscle tightness, immediately reduce power to 70% FTP and consume electrolytes.

Q3: Does this pacing method apply to Ironman 226km long-distance events?

A: Yes, but parameters need adjustment. In the Ironman bike leg (180km), since a full marathon follows, power settings should be more conservative. It is recommended to reduce the uphill overreach amplitude to 3-5% (rather than 5-8%), and maintain power at 60-70% FTP on descents (rather than 40-50%) to ensure “freshness” of the leg muscles. Research indicates that NP during the bike leg of an Ironman should be controlled at 70-75% FTP, with the range of variable power fluctuations narrowed to within ±15% to avoid excessive glycogen depletion.

Q4: How can I execute variable power pacing without a power meter?

A: If you only have a heart rate monitor, you can use the “heart rate lag effect” for approximation. On climbs, elevate heart rate to 92-97% of threshold heart rate (approximately RPE 8), but note that heart rate response has a 30-60 second delay, so you should begin accelerating 10-15 seconds before the base of the climb. On descents, allow heart rate to naturally drop to 75-80% of threshold. If you don’t even have a heart rate monitor, use “breathing rhythm” as your guide: maintain an intensity on climbs where you “can speak single words but not full sentences,” and recover on descents to a level where you “can converse easily.” A more advanced technique is using “cadence” as a reference—maintain 75-85 rpm on climbs and increase to 95-105 rpm on descents, regulating effort through gear selection.

Q5: Does variable power pacing increase the burden on the cardiovascular system?

A: Based on physiological data, riders using variable power pacing typically have a slightly higher average heart rate than constant pacers (approximately 2-4 bpm), but the frequency and duration of maximum heart rate episodes do not significantly increase. The key lies in the active recovery on descents, which allows heart rate to drop rapidly, creating a “sawtooth” heart rate curve. This pattern actually trains the heart’s “chronotropic response,” improving stroke volume. For athletes without a history of cardiovascular disease, this is a safe and effective training stimulus. However, if you have a history of hypertension or arrhythmia, consult a physician first and train under coach supervision.

Conclusion: The variable power pacing method for undulating terrain represents the deep integration of sports science and practical racing experience. It breaks the traditional dogma of “constant pacing,” using the mathematical language of calculus to reveal the optimal path of power distribution. For Taiwan, an island of mountains and slopes, mastering this technique is equivalent to holding the key to personal best performances. Whether challenging the summit of Wuling or conquering the long roads of the Huatung circuit, let power become your strategic weapon, not a blind numerical constraint.