Every cyclist who has ever stared at a Garmin screen and wondered why holding 250 watts produces 34 km/h one day and 31 km/h the next is asking the right question. The answer is physics — specifically, the cubic relationship between power and speed that makes cycling one of the most unintuitive sports from a numbers perspective.
This relationship is not academic. It explains why you cannot buy speed with watts alone, why a 10 per cent FTP increase does not produce a 10 per cent speed increase, and why the rider in an aero tuck on a worse bike often beats the rider with better legs sitting bolt upright. It also explains why the smartest riders I have spoken to across 1,400-plus podcast episodes think about drag reduction with the same seriousness they give to training. Dan Bigham, former UCI Hour Record holder and now Head of Engineering at Red Bull-Bora-Hansgrohe, told me that most amateurs would gain more speed by spending an hour adjusting their position than by spending a month training harder. He is not wrong.
Run your own numbers through the power-to-speed calculator alongside this article. Seeing the maths applied to your weight, your position, and your power makes the concepts stick in a way that reading alone does not.
The Cubic Relationship: Why Doubling Power Does Not Double Speed
On flat ground at constant speed, a cyclist's power output is consumed by three forces: aerodynamic drag, rolling resistance, and drivetrain friction. Of these, aerodynamic drag dominates — and it follows a brutal mathematical law.
The power required to overcome aerodynamic drag is:
P_drag = 0.5 × ρ × CdA × v³
Where ρ is air density (roughly 1.225 kg/m³ at sea level), CdA is your aerodynamic drag area in square metres, and v is your velocity in metres per second. That exponent — the v cubed — is the part that ruins everything for the optimistic cyclist.
Here is what it means in practice. Take a 75 kg rider with a CdA of 0.32 m² (a decent position on the drops) and a combined rolling resistance and drivetrain loss that requires about 25 watts at 30 km/h. At 30 km/h (8.33 m/s), the drag power alone is roughly 95 watts. Add the rolling and drivetrain losses and total power sits around 120 watts.
Now double the speed to 60 km/h. The drag power cubes — it does not double, it does not quadruple, it goes up by a factor of eight. The drag component alone jumps to roughly 760 watts. Total power required is around 810 watts. That is a number only a track sprinter can produce, and only for about fifteen seconds.
The practical consequence: doubling your power output from 200 to 400 watts on flat ground increases your speed by roughly 26 per cent, not 100 per cent. From about 32 km/h to about 40 km/h. Every additional watt buys you less and less speed, because every additional km/h costs you exponentially more power to hold.
This is why the difference between a Category 4 time triallist and a national champion over 25 miles might be 80 watts — but the speed difference is only 5-6 km/h. And it is why position changes that reduce drag are so disproportionately valuable. A 10 per cent reduction in CdA is worth roughly 3.3 per cent more speed at the same power — without turning a single extra pedal stroke.
Aerodynamic Drag: What CdA Is and Why It Controls Your Speed
CdA — coefficient of drag times frontal area — is the single number that defines how much air resistance you create. The coefficient of drag (Cd) captures how slippery your shape is. Frontal area (A) captures how large a hole you punch through the air. Multiply them together and you get CdA, measured in square metres.
Typical CdA values for a road cyclist:
- Hoods, relaxed position: 0.35-0.40 m² — this is where most amateurs spend most of their time. Head up, arms straight, torso at 30-40 degrees from horizontal. Comfortable, visible, and aerodynamically expensive.
- Drops, elbows bent: 0.28-0.32 m² — tucking onto the drops with bent elbows and a flatter back drops CdA by 15-20 per cent. This is free speed that costs nothing but flexibility and core strength.
- Aero bars / TT position: 0.20-0.25 m² — arms together, torso near horizontal, head low. The UCI Hour Record position sits around 0.19-0.21 m². Dan Bigham rode his Hour Record at a CdA below 0.20 m², which required years of position refinement.
- Full tuck descent: 0.22-0.28 m² — supertuck variants where the rider crouches low with knees in. Since the UCI banned the supertuck in racing, this is training and sportive territory only.
At 35 km/h, moving from hoods (0.35 m²) to a decent drops position (0.30 m²) saves roughly 20-25 watts. That is the equivalent of several weeks of structured interval training — delivered instantly, every ride, for free.
This is why the conversation about getting faster should start with position, not power. I have worked with riders in the Roadman Cycling community who gained 2 km/h average speed on their regular loop without any fitness change — just by committing to riding the drops with their elbows bent.
Why Position Matters More Than Fitness Above 30 km/h
At 30 km/h, roughly 75 per cent of your power is fighting the air. At 40 km/h, it is 85 per cent. At 50 km/h, it exceeds 90 per cent. These numbers mean that above 30 km/h, your speed is determined primarily by your aerodynamic efficiency, not your engine size.
Consider two riders. Rider A produces 280 watts with a CdA of 0.35 m² (hoods, relaxed). Rider B produces 240 watts with a CdA of 0.28 m² (drops, aero position, tight jersey). On flat ground in still air, Rider B is faster despite producing 40 watts less. The maths is unambiguous.
Dan Bigham's work — both as a rider and now as an engineer — has demonstrated this repeatedly. His approach to aerodynamics is methodical: measure CdA in the field using the Chung method (power meter data, GPS speed, known gradient), then make small positional changes and re-test. No wind tunnel required for the initial gains, though professional testing tightens the margins further.
Sebastian Breuer, who won the Badlands 800 km ultra-distance race, applied this thinking to gravel. He ran 3D-printed aero extensions on his gravel bike — a setup that looks absurd until you calculate the cumulative time saved over 30 hours of riding at 25-30 km/h. At those speeds, aero bars save 15-20 per cent of the drag that a gravel rider on the hoods creates. Over 800 km, that is hours, not minutes.
The lesson for amateur road cyclists is the same principle at a smaller scale. Your 60 km Sunday ride at 30 km/h average is an aerodynamic event. Everything from your jersey fit to your head position to the width of your handlebars is either helping you or taxing you, and the tax compounds with every kilometre.
Rolling Resistance: Tyres, Pressure, and the Death of the Skinny-Tyre Myth
Aerodynamic drag dominates above 25 km/h, but rolling resistance is the second-largest force on flat ground, and the one most cyclists get wrong in the easiest way.
Rolling resistance is expressed as Crr — the coefficient of rolling resistance. A lower Crr means less energy lost as the tyre deforms against the road surface. Typical values:
- Fast road clincher (GP 5000 TL, 28mm, tubeless): Crr ≈ 0.0032
- Standard road clincher (training tyre, butyl tube): Crr ≈ 0.0050
- Budget road tyre with butyl tube: Crr ≈ 0.0065
- Gravel tyre, 40mm: Crr ≈ 0.0055-0.0070
The power cost of rolling resistance is:
P_rolling = Crr × m × g × v
Where m is total system mass, g is gravitational acceleration, and v is velocity. Unlike drag, this scales linearly with speed — which is why it becomes a smaller fraction of total resistance as you go faster, not a larger one.
At 30 km/h for a 75 kg rider on a 9 kg bike, switching from a training tyre with butyl tube (Crr 0.0050) to a fast tubeless setup (Crr 0.0032) saves roughly 5 watts. That is not transformative, but it is free once you have made the switch, and it stacks with every other marginal gain.
The skinny-tyre myth — the idea that narrower tyres are faster because they have less contact area — has been comprehensively demolished by data from SILCA, Bicycle Rolling Resistance, and independent testing. On real road surfaces, a 28mm tyre at 70 PSI is faster than a 23mm tyre at 100 PSI because the wider tyre deforms over surface imperfections rather than deflecting the entire bike and rider vertically. Those vertical deflections are wasted energy. The tyre pressure guide covers the detail, and the tyre pressure calculator will give you a specific starting point for your weight and setup.
Latex inner tubes reduce hysteretic losses compared to butyl — roughly 2-3 watts at 30 km/h — but tubeless eliminates the tube entirely, removing the casing friction between tube and tyre that accounts for a meaningful chunk of total rolling loss. If your wheels and tyres support tubeless, run tubeless. The speed advantage exists at every pressure point and every speed.
Weight vs Aero: The Gradient Crossover
This is the question I get asked more than any other, and the answer is surprisingly precise.
On flat ground, weight barely matters. A 1 kg weight saving at 35 km/h on flat tarmac saves roughly 0.3 watts through reduced rolling resistance. At the same speed, a 0.01 m² CdA reduction saves approximately 5 watts. The aero gain is more than fifteen times larger.
As gradient increases, weight becomes more important because you are doing more work against gravity and less against the air. The power required to climb is:
P_gravity = m × g × gradient × v
This scales linearly with mass, gradient, and speed. Critically, climbing speed drops — and with it, aerodynamic drag drops too, because drag depends on v³.
The crossover happens at roughly 4-5 per cent gradient for most riders at typical amateur power outputs. Below that, aero dominates. Above that, weight dominates. At 7-8 per cent, a 1 kg weight saving is worth 15-20 seconds per 10 km of climbing. At the same gradient, a 0.01 m² CdA reduction is worth 3-5 seconds. Weight wins by a wide margin.
For the practical road cyclist riding a mix of flat and rolling terrain with occasional climbs, the implication is clear: optimise for aero first, weight second. The aero vs weight breakdown covers the tradeoff in detail, and the race predictor will model both variables across a full course profile so you can see exactly where your time is won and lost.
W/kg — watts per kilogram — becomes the controlling variable on sustained climbs because it captures both sides of the gravity equation. Two riders producing 250 watts on a 7 per cent gradient will be separated entirely by their mass. The lighter rider climbs faster, full stop, because more of those 250 watts survive the gravity tax. Calculate yours with the W/kg calculator and compare against the benchmarks for your target events.
Drafting: The Physics of Riding in a Group
Drafting is the single largest performance variable available to any cyclist, and it costs zero watts to implement. Sit in the wheel of the rider ahead and your aerodynamic drag drops by 30-40 per cent, depending on the gap distance and the size of the leading rider.
The physics is simple. The rider in front creates a low-pressure wake behind them. By sitting in that wake, you ride through air that is already partially displaced. The closer you sit, the larger the drag reduction — at one bike length (roughly 1.5 metres), the saving is typically 35-40 per cent. At two bike lengths it drops to 25-30 per cent. At three bike lengths it is 15-20 per cent and diminishing rapidly.
In a peloton of 40 riders, the sheltered riders in positions 10-30 are producing roughly 40-50 per cent less power than the rider on the front. This is why a professional peloton can average 45 km/h for hours while individual riders are holding 180-220 watts in the bunch — a solo effort at that power would produce 32-34 km/h.
The cascade effect compounds through the group. Each successive rider creates additional wake disturbance, so the rider in fifth position experiences a slightly different — and often slightly better — drag profile than the rider in second position. Wind tunnel studies have shown that the third and fourth positions in a pace line are the most aerodynamically efficient, which is why experienced riders in echelons sit third or fourth wheel rather than fighting for second.
For the amateur cyclist, the practical application is simple: if you ride with a group, your average speed is determined more by where you sit in the group and how smoothly the rotation works than by your individual power output. A well-drilled group of four riders averaging 260 watts each will be faster than a solo rider at 320 watts. The group's collective aerodynamic advantage is that large.
The wind chill calculator is relevant here too — on cold days, the rider on the front faces a significantly higher effective wind speed (and thus greater cooling) than the sheltered riders behind. Dress for your position, not just the ambient temperature.
Flat Time Trials vs Mountain Stages: Two Different Sports
A flat 40 km time trial and a mountain stage finishing atop an HC climb are, from a physics perspective, almost entirely different events. Understanding why sharpens your thinking about what to train and what to spend money on.
In a flat TT, the rider's speed sits at 40-50 km/h for the entire effort. At those speeds, 85-90 per cent of power is fighting the air. CdA is the dominant variable. An FTP difference of 20 watts between two riders might produce a 30-second gap over 40 km. A CdA difference of 0.03 m² — achievable through position alone — produces a 60-90 second gap. The position is worth more than the fitness.
This is why TT specialists obsess over aerodynamics to a degree that seems pathological from the outside. Millimetre changes in bar width, arm pad height, and head angle are tested and re-tested because at 48 km/h, a 0.005 m² CdA change is worth 8-10 seconds over 40 km. Professional TT riders like Alex Dowsett have spent hundreds of hours in wind tunnels refining positions for exactly this reason.
On a mountain stage, the equation flips. A 10 km climb at 8 per cent gradient is ridden at 12-18 km/h depending on the rider. At those speeds, aerodynamic drag accounts for under 15 per cent of total resistance. Gravity takes 80 per cent or more. The controlling variable becomes W/kg — sustainable power output relative to body mass.
Two riders with identical CdA will be separated on a 10 km, 8 per cent climb entirely by their W/kg. A rider at 4.5 W/kg finishes roughly 3 minutes ahead of a rider at 4.0 W/kg. No amount of aero optimisation bridges that gap at climbing speeds.
For the sportive rider preparing for a mixed-terrain event — some flat, some climbing — both variables matter, but the relative weighting depends on the course. A flat-to-rolling 100 km sportive with 600 m of elevation rewards aero. A mountainous 100 km with 2,500 m of elevation rewards W/kg. Model the specific course through the race predictor to see where your minutes are won and lost, and the climb time calculator to set realistic pacing targets for each ascent. The VAM calculator gives you a real-time metric during the climb itself — vertical metres per hour — which correlates directly with power-to-weight and strips away the noise of varying gradients.
Practical Ways to Get Faster Without Getting Fitter
The cubic relationship means that reducing drag is more efficient than producing more power. Here are the specific changes, ranked by approximate watt savings at 35 km/h, that cost no fitness adaptation at all.
Position: the biggest free gain
Moving from a relaxed hoods position (CdA ~0.37 m²) to bent-elbow drops (CdA ~0.30 m²) saves 25-35 watts at 35 km/h. This is the single largest improvement available to most riders and it costs nothing. The limitation is core strength and flexibility — holding a low, narrow position for three hours requires trunk stability that many riders lack. Core work off the bike (planks, pallof presses, bird-dogs) builds the foundation. No heavy barbell work needed — bodyweight and light resistance is sufficient and safer for masters-age riders.
Narrowing the elbows is the highest-value positional change. Most riders ride with their elbows splayed outward, creating a wider frontal profile than necessary. Bringing the elbows in by 5-10 cm each side while on the drops reduces frontal area meaningfully. It feels awkward for the first few rides and then becomes normal.
Clothing: tighter is faster
A loose jersey flapping in the wind at 35 km/h costs 5-15 watts compared to a well-fitted race jersey. A skinsuit — even a non-aero road skinsuit — saves 10-20 watts over a standard jersey-and-shorts combination at the same speed. The fabric surface matters too: textured fabrics (dimpled or ribbed) can trip the boundary layer from laminar to turbulent at the right Reynolds numbers, reducing drag further. This is why World Tour skinsuits look rough to the touch — that texture is engineered.
For amateur riders not racing in skinsuits, the practical advice is simple: wear clothing that fits closely. No flapping zippers, no billowing back pockets full of arm warmers, no loose rain cape draped over the saddle bag. Every piece of fabric catching the wind is a watt you are paying for.
Helmet: shape over ventilation
An aero road helmet saves 5-10 watts over a well-vented climbing helmet at 35 km/h. A dedicated TT helmet saves 10-15 watts over a standard road helmet. The catch is that aero helmets vent less air, so there is a thermal penalty on hot days. For riders in the UK, Ireland, and northern Europe — where overheating is rarely the limiting factor — an aero road helmet is a sensible default for everything except the hottest summer days.
Head position interacts with helmet choice. An aero helmet only works if the tail stays aligned with the rider's back. If you ride with your head up, scanning the road ahead, the tail catches the air and the helmet becomes a drag device. Practice riding with your eyes up and your head tucked — looking through the top of your eye sockets rather than lifting your chin.
Wheels: depth over weight
Deep-section wheels (40-65mm rim depth) are faster than shallow wheels on flat and rolling terrain, not because they cut through the air (a common misconception) but because they guide airflow around the tyre-rim junction more cleanly. The drag savings are 3-8 watts at 35 km/h depending on depth and yaw angle.
The weight penalty of deep-section wheels (typically 100-300g heavier than shallow equivalents) is irrelevant on flat ground and costs single-digit seconds on most amateur climbs. Unless you are racing up Alpe d'Huez, the aero gain outweighs the weight cost. For a mixed-terrain sportive, 50mm all-rounders are the default answer.
Tyres and pressure
As covered above, running fast tyres at the right pressure — lower than most cyclists assume — saves 3-8 watts over a poorly-chosen setup. Tubeless, 28mm, at 65-75 PSI for a 75 kg rider. Use the tyre pressure calculator for your specific numbers.
Reducing frontal area: the small things
Bar width matched to shoulder width (or narrower) reduces frontal area. Most bikes ship with 42cm bars; many riders would be faster on 40cm or even 38cm. Shorter cranks — 165mm rather than 172.5mm — allow a lower hip angle and a lower saddle-to-bar drop without the same flexibility demands. Removing unnecessary accessories from the handlebar area (oversized computers, lights in commuting position, bar-end mirrors) cleans up the airflow around the widest point of the bike.
None of these changes is dramatic in isolation. But stacked together — position, clothing, helmet, wheels, tyres, bar width — a rider can realistically save 40-60 watts at 35 km/h. That is the equivalent of months of structured training, delivered without a single interval session. It is the closest thing cycling offers to a free lunch, and it is sitting right there in the physics for anyone willing to take it seriously.
Putting It Together: What the Maths Is Telling You
The cubic relationship between power and speed is not an inconvenience. It is an instruction manual. It tells you, with precision, where to invest your time and money if speed is the goal.
Below 25 km/h — on climbs, in headwinds, in stop-start urban riding — power and weight matter most. Above 25 km/h — on flat roads, in groups, in time trials, in any sustained effort where speed stays in the thirties or forties — aerodynamics dominates everything else.
Most amateur cyclists ride between 28 and 38 km/h on the flat, which puts them squarely in the zone where CdA determines speed more than FTP does. And yet most amateur cyclists spend 90 per cent of their improvement budget — both time and money — on the engine rather than the vehicle.
Train, absolutely. Build your FTP. Extend your endurance. Improve your W/kg for the climbs. But do not ignore the other side of the equation. A rider who trains hard and rides in an aerodynamically awful position is leaving the cheapest watts on the road.
Run your numbers through the power-to-speed calculator. Change one variable at a time — CdA, weight, power, gradient — and watch what happens. The maths will tell you exactly where your next five minutes of speed is hiding. For most riders, it is not in the legs. It is in the air.