Four hours in the Silverstone wind tunnel. Thirteen watts saved. Not a single one of them came from a new frame or a deeper wheel.
That is the headline from Sam Calder's session with Dylan Johnson, and once you sit with it for a minute the implication is uncomfortable. Most amateur gravel racers are spending in exactly the wrong order. The kit is doing more work than the bike. The fit is doing more work than the wheels. And the framework for deciding what to spend on next is not cost per gram. It is cost per CdA.
Listen to the full conversation on the Roadman Cycling Podcast →
Sam runs Rule 28, the apparel brand that has spent the last few years pulling watts out of fabrics most people would not think twice about. He took Dylan into Silverstone — the tunnel of choice for serious British aero work — and ran a four-hour session built around the 2025 gravel calendar. The data is the kind of thing the rest of us rarely see.
Here is what it actually said.
The 13-Watt Number Has To Be Read Properly
Thirteen watts on a four-hour session is a hell of a return. The way to translate that into language that lands at amateur level is a rule of thumb Sam mentioned in passing.
At the sharp end of a long race, one watt is roughly one minute of finish-time saving.
That is the relevant unit for an event like Unbound, where the front group is racing for ten hours. Thirteen watts becomes thirteen minutes off Dylan's projected finish. The slower you are on the course, the more time a single watt is worth, because you are out there longer for the drag to chip away at you. The amateur racer hanging on at the back of the lead chase group is not getting smaller savings than the pro. They are getting bigger ones.
The other thing worth pulling out is what was not changed. No new frame. No new wheels. No new bike at all. The 13 watts came entirely out of position, helmet, suit, sock and hydration choices. The bike Dylan rode in was the bike Dylan rode out. That is the part that should rearrange how most riders think about the next thing they buy.
The Sock Revelation Most Amateurs Miss
Start at the floor and work up.
Rule 28 has built two generations of aero sock. The newer One sock — single layer, woven rather than knitted — was the real find of Sam's recent fabric work. On their road test rider at 45 km/h, the One saved 10.5 watts versus the previous generation. Ten and a half watts. From a sock.
At Dylan's gravel race pace of 35 km/h the same upgrade was worth around three watts. The reason is the woven construction. Knitted fabrics are limited in how fine you can run the fibres and how tightly you can pack them. Woven gives you precision. The textured rib on the One steps up out of the fabric instead of stepping down into it, which changes how the airflow trips off the surface.
Read that back. A pair of socks saved a pro racer three watts at his event speed. A 3,000 dollar wheel upgrade rarely saves more than five. The ratio is not even close.
One nuance worth flagging. The dual-layer Neo sock — a track-speed product — only earns its keep above 50 km/h. If you are doing a 25 km/h ultra event, a less aero but better-breathing sock is the right call. If you are racing at 35 km/h or above, the woven One range is the call. Match the kit to the speed.
Helmet Size Matters More Than Helmet Shape
This is the bit of the testing that should change a lot of helmet purchases.
In a road TT position your head is in line with your torso. Helmet shape interacts with the airflow that then has to travel over your shoulders. That is why the wild new generation of TT helmets are growing into shoulder-shrouding shapes. The aerodynamic gain is not from being slippery. It is from filling in the gap behind your head.
In a gravel position your head sits above your shoulders. There is no helmet-to-shoulder interaction to optimise. Which means the helmet shape itself produces tiny differences — a watt or two between the best and worst aero road helmets Sam tested.
The big lever in gravel is helmet size. Sam tested a Met Manta in large and a Met Manta in medium. Same helmet. Different size. Four to five watts of difference. The medium won, simply because it presented a smaller frontal area to the wind. That is a free upgrade for anyone who has been buying the wrong size for comfort or for fashion.
The takeaway is direct. If you are racing gravel and your head is above your shoulders, do not chase the latest exotic shape. Chase the smallest helmet you can wear safely and comfortably. The Met Manta tested as the fastest of the helmets in the session, but the principle is broader than the brand. Smaller frontal area beats marginal shape gains in this position every time.
The Hydration Paradox
The wind tunnel keeps producing results that contradict café-stop common sense. The hydration pack data is one of the cleanest examples.
The 3-litre reservoir tested faster than the 1.5-litre. Not slower. Faster.
The reason is geometry. In a gravel position your helmet sits high. That creates a low-pressure void behind your head that the airflow wants to fill. A bigger pack fills the void and lets you draw a clean curve from the back of the helmet, across the pack, down to the hips. A smaller pack sits too low. The void stays open. The airflow has to deal with a draggy hole between helmet and pack, then re-attach to a separate body further down.
The lesson is not to ride with three litres of water you do not need. If the race is short enough that you can carry less, carry less. The point is that when you do need the volume, do not assume you are paying an aerodynamic price for it. You are not.
The pack-versus-suit-pocket question lands the same way. The Rule 28 gravel suit was not built primarily for aerodynamic gain over a backpack. It was built for comfort over ten hours. Comfort is also a CdA conversation in disguise — being uncomfortable wrecks your power output, and lower power means lower speed.
This is the bit amateurs underestimate the most. The fastest setup is the one that lets you produce your true power for the full duration of the event. Anything that compromises that is a hidden tax on the day, even if the wind tunnel says it is fast in isolation.
Bare Hands. No, Really.
A small point that earns its mention.
Bare hands are still faster than aero mitts. Other companies have produced numbers showing aero mitts beating non-aero mitts by three or four watts. Rule 28's testing has them slower than no mitts at all. The exception is a TT-specific glove, which is a slight gain over bare hands, but the saving is small.
So why do they make aero mitts? Because if you are racing Unbound, you will crash at some point. An aero mitt is faster than a torn-up hand. That is the only reason. If you are confident you will stay upright, bare hands are quicker. Looks better, too.
The principle behind that decision matters more than the choice itself. Every kit call is being made against a hierarchy that puts performance first, then risk, then comfort.
Head Position: The Upright Win
A nice happy coincidence in the data.
Sam tested Dylan in two positions. A tucked, lower head and a more sustainable, slightly higher head. The upright position was faster.
It should not have been. Lower is usually quicker. Except in this case the higher head was the position that fed clean airflow onto the hydration pack. The drop position created turbulence that wrecked what the pack was doing for him.
That is a win-win that almost never happens in aero testing. A more sustainable position that you can hold for ten hours is also the position that tests three or four watts faster. The lesson here is the broader one — gravel is not road. Optimise for the system, not the part. The same head drop that saves you watts on a road TT can cost you watts in a gravel race because the pack and helmet are doing different jobs.
Cost Per CdA Beats Cost Per Gram
This is the framework worth taking out of the whole conversation.
Most amateur cyclists shop on cost per gram. They will pay 1,500 dollars to drop 100 grams from a wheelset. It sounds rational. It is rational, but only on sustained climbs above six or seven percent — the part of the course where gravity wins and aerodynamics quietly stops mattering.
For everything else, the right metric is cost per CdA. The reduction in your drag area for the price you are spending. On flat and rolling gravel above 25 km/h, that metric puts your spending in a very different order.
A 100 dollar pair of woven socks: roughly three watts at gravel race pace. Cost per watt: about 30 dollars.
A 400 dollar properly fitted skinsuit-style gravel suit: 16 watts on Sam's testing of GCN's Alex Patton. Cost per watt: 25 dollars.
A 4,000 dollar new aero frame versus a 2020-era road bike: somewhere between five and ten watts. Cost per watt: 400 to 800 dollars.
The hierarchy is not subtle. Kit and position dominate everything until you have already optimised them. The frame is the last thing you should be replacing, not the first.
What This Looks Like In Practice
If you are a serious amateur racing gravel this year, the order is straightforward.
Start with what you have on your body. A properly fitting race suit. The right size of a good helmet — not the most expensive, the smallest you can wear without choking yourself. A pair of woven aero socks matched to your average race speed. Eyewear that does not leave a chasm between your pecs. That is the first 800 dollars and it will pay back more watts than any wheel set you can buy.
Then the wheels and tyres. Tubeless is the cheap rolling-resistance win. Wider tyres — Sam was clear that 28s and 30s now beat narrower options on most road frames designed in the last few years, and a 2.2 inch mountain bike tyre can be the right call in a gravel race where puncture protection and grip start to outweigh marginal aero loss. Match the rubber to the course you are actually racing, not the course you wish you were racing.
Then, and only then, the frame. By the time you are spending money there, the cheap watts are already in the bank.
If you want this thinking woven into a year of training and racing — not just a bike-park decision tree — that is what the Roadman coaching system does for athletes inside the Not Done Yet community. The aero work lands harder when it sits on top of structured intensity discipline — the kind of weekly load TrainingPeaks was built to track — and a body composition number that is actually moving. Kit savings compound when the engine underneath is being built properly too.
The Roadman Take
The 13 watts is the number that gets the headlines. The lesson is the framework.
Aerodynamics in gravel is not about the bike. It is about the system the rider sits inside — kit, position, helmet, hydration, eyewear, sock fabric, the order they are sized in. Every layer is a conversation about cost per CdA, and the cheap watts are hiding in the layers most amateurs ignore. The pros are winning because every layer has been audited. The amateurs at the back of the lead group are losing because half of theirs have not.
The whole point of the Rule 28 work — and the reason Sam took Dylan into Silverstone — is that this is fixable. Not slowly, not over years. Fixable in a single shopping decision and a single position session. You are not done yet. The watts are sitting on the table. The framework for collecting them is not complicated. It just runs against a few habits the cycling industry has spent decades training us into.
If you have a specific question about your own kit choice, your own course, your own next purchase — what to upgrade first, what is actually worth your money — ask Roadman for an answer drawn straight from the coach and engineer conversations on the podcast. That is what the AI coach is built for.
Cycling is hard. Our podcast will help.
