James Thinks

writing is a kind of thinking

Now that I have a power meter in the rear hub of my road bike I can compare not only my own efforts, but how much various bits of kit slow me down. In this case, I have compared a tough fast-touring Schwalbe Marathon Greenguard 25mm tyre with a Michelin Pro 4 Endurance 28mm. The latter should have lower rolling resistance by some 7W if the laboratory results from bicyclerollingresistance.com are to be believed.

However, those tests are done on a checker plate drum in controlled conditions. In some ways that's good because it removes a lot of noise from other factors that you'd get in the real world. However, I'm interested in getting some real-world data, that applies more directly to me and the bumpy roads I normally ride.

So I ran my own rather inexact experiment as follows:

Route: 3.2km climb up "Road hill climb" through Ditteridge. I didn't measure the descent as it depends more on wind resistance.
Power: 200W average. A power I can consistently and repeatedly produce for the time it takes to climb this hill.
Wheels: Front wheel only changed. I could've changed the rear as well, but that would've taken all day.
Measurement: The total lap time for climbing the hill. With consistent power this should show any differences in the drag of the set up.

Possible confounding factors

  • Wind - at a couple of points on the climb there was a headwind which seemed to increase around the middle of the session. I can't be sure how much difference this made. Ideally I'd do this on a still day.
  • Road surface inconsistencies - try as I might, I can't be sure I took exactly the same route up the hill each time. When cars passed me I may have been forced over rougher patches of road for a moment.
  • Different hubs - I didn't bother changing the tyres over to use the same rim and hub, so the Schwalbe was on my recently-serviced SON Delux hub. I guess there's a slight extra weight and possibly drag penalty for the Schwalbe tyre.
  • Variable power output - While the average power over the climb was 200W +/- 1W, I couldn't stay at exactly 200W all the way up with varying gradient and gear changes. 200W is obviously a lot faster on the flatter sections than the climbs, so there is more than one way of doing it. Arguably it may be faster to put out most of that 200W allowance on the steepest parts when less is being lost in wind resistance.
  • Psychology/Placebo effect - Although I'm not conscious of riding differently, I may have tried harder to steer a straight line and stay on the smoothest bit of road with some climbs than others.

With those caveats out of the way, here are the results.


Tyre Pressure (PSI) Lap power Lap time Relative drag (Watts)
Schwalbe Marathon Greenguard 25mm 95 200 00:13:21 13
Schwalbe Marathon Greenguard 25mm 80 199 00:13:31 16
Schwalbe Marathon Greenguard 25mm 70 199 00:13:27 15
Schwalbe Marathon Greenguard 25mm 60 199 00:13:14 12
Michelin Pro4 Endurance 28mm 85 199 00:13:08 10
Michelin Pro4 Endurance 28mm 70 199 00:13:12 11
Michelin Pro4 Endurance 28mm 65 200 00:13:14 12
Michelin Pro4 Endurance 28mm* 58 199 00:13:11 11
Michelin Pro4 Endurance 28mm 58 199 00:13:05 9

* Had to brake due to a turning vehicle.

Results Time
Av. Marathon Greenguard this 00:13:23
Av. Pro4 Endurance 00:13:09
Difference 00:00:13
Time diff per 100km (uphill) per tyre 00:07:01

The raw data copied from my Strava ride. Recorded on a Garmin Edge 500 using a PowerTap hub. "Relative drag" is based on my dodgy reverse-use of the Bike Calculator and is only useful for comparison within this table, if at all.


The results aren't as clear as I'd hoped, but I think the technique roughly works. Climbing at a consistent power and measuring the time did show some consistent results. The Michelin Pro 4 endurance tyre with a plain Shimano 105 hub was faster every time than the Schwalbe Marathon Greenguard on a SON Delux dynamo hub (disconnected). However, the difference is only a few seconds and multiplying this up to a 100km for two wheels would result in a longer ride by less than 15 minutes.

That number only really makes sense if the whole of that 100km was uphill, like my test. Downhill, tyre rolling resistance is small compared to air resistance. Perhaps a better estimate for a real 100km would be an extra ten minutes per 100km using the tougher Marathon tyres. On second thoughts, multiplying up like that would also multiply up my errors and confounding factors, so I ought to run a better experiment before declaring such a difference.

Future improvements

  • There's no noticeable drag from either hub, but I could obviously improve the experiment by actually changing the tyre between runs. This would also mean the rim/hub weight wouldn't affect the climb.
  • I think most of the runs were made with higher than ideal pressure for my weight. At 63kg I'm fairly light and so prefer lower pressures. If I did this again I'd try to work out a good pressure for each tyre and stick with it.
  • Repeated runs. Gusts of wind and wobbling around slightly rougher bits of ground could be averaged out if I repeat the test several times for each set up and take the total time (excluding descents).
  • Fit the tyres to both wheels. This should show twice the effect, maybe more as there's more weight on the rear wheel. I'd only need to change the tyres once during the session, so it shouldn't take too long.
  • Ride on a day with virtually no wind. At times this seemed to make a big difference, albeit for a couple of seconds, so removing it should be easy enough.

See follow-up post.

Mugshot of James cycling on a road in the sunshine.

James Bradbury

I write about whatever is on my mind. I do so mostly to help me think more clearly. If other people find it interesting that's good too. :-)