• en
  • fr
  • ja

Myth 19: 700C Wheels Are Faster

When we started this series to celebrate Bicycle Quarterly’s 15th anniversary, we thought we’d eventually run out of myths. But it seems that new ones are created as fast as we can debunk old ones. The latest is “700C wheels roll faster than 650B.”

This is stated with the same certitude as the old “narrow tires are faster” – and it’s just as wrong. Simply put, there is no evidence that 700C wheels roll faster than 650B (or 26″), and much data to show that they all roll at essentially the same speed.

In theory, it’s compelling to believe that a larger wheel will roll over road irregularities faster: The angle of attack is shallower, and the larger wheel will roll up onto the obstacle, rather than bump against it.

The reality is that the three most popular wheel sizes (above) are so close in size that the small differences are simply not important, even in an age of ‘marginal gains.’

Our testing confirms this: On smooth pavement, all three wheel sizes rolled at the same speed.

We ran otherwise identical 42 mm-wide tires (Schwalbe Marathon HS) at identical pressure. The only difference was the wheel size.

The dark gray area of each column shows the ‘confidence interval,’ which overlap for all tires. The small differences you see are noise in the data. (Ignore the fourth column: We ran one tire at different pressure to make sure small differences in tire pressure didn’t affect our results.)

What about rough roads and gravel? Above is the same test on highway rumble strips – pretty much the roughest ground you can find. (We chose rumble strips because they are uniform, not random like gravel or cobblestones.) Again, you see the confidence intervals overlap – the three wheel sizes roll at the same speed.

We’ve repeated these tests multiple times, with different methodologies (power meter, roll-down), always the same results: 700C wheels don’t roll faster than (slightly) smaller ones.

The real world confirms these results. Ted King has won Dirty Kanza on 650B and 700C tires, while most other racers were on 700C, because their sponsor’s bikes and wheels were designed for this tire size.

So where does this myth come from? Modern all-road bikes with disc brakes make it easy to swap wheels between 650B and 700C. Most riders use their 700C wheels on the road with fast-rolling racing tires. Their 650B wheels are shod with heavy gravel tires. And yes, the 650B wheels will be slower.

Of course, comparing supple and light 700C racing tires like the Vittoria Open Corsa CX or the Rene Herse Extralight with stiff and heavy 650B gravel tires like the Schwalbe G-One or the WTB Horizon makes little sense. The reason why the racing tires are faster isn’t the size of the wheels they are mounted on, it’s their supple casing. And if you made a 650B tires with the same supple casing, it would be just as fast. (We’ve done it and tested that, too.)

A 700C x 28 wheel has the same diameter as a 650B x 48 mm, so the angle of attack will be the same for both. This shows the absurdity of the argument. Of course, wheel makers want to tell you that you need two wheelsets, and tire makers would rather add a ‘fast-rolling 700C version’ to their lineup than develop a truly fast gravel tire.

Our OPEN U.P.P.E.R. test bike is gone now, but while I rode it, I knew that the 650B wheels (shod with our 48 mm Switchback Hill Extralights) weren’t holding me back. Nor was anything else – this was one of the fastest bikes I’ve ridden.

The U.P.P.E.R. also illustrates why 650B wheels make sense on all-road and gravel bikes: A 48 mm-wide 700C tire would require longer chainstays (adding weight and flex where you don’t want it), it would cause toe overlap, and it would take away from the nimble handling that made the OPEN so much fun.

Wouldn’t the U.P.P.E.R. be even faster with 700C x 28 mm tires? That question brings us full circle to the first part of this series: Myth 1: Wider Tires Are Slower.

Of course, Rene Herse offers all three popular tire sizes: 700C, 650B and 26″. That way, you can choose what works best for you.

Further reading:

59 Responses to Myth 19: 700C Wheels Are Faster

  1. Ford Bailey January 6, 2020 at 10:17 am #

    Around here people are set in their ways and don’t understand the advantages of not just 650b wheels, but wider tires in general.
    I’m kind of glad they don’t. They are my secret- they let me descend faster, ride with more comfort and fewer flats. And they’re light !

  2. Andrew January 6, 2020 at 10:20 am #

    I see that all of the test tyres (except one) were all inflated to 3 bar. What was the width of the 700c and 650b tyres used and were they all the same tyre model and casing?

    • Jan Heine January 6, 2020 at 10:33 am #

      They were all the same width (42 mm), same casing, same tire model. Otherwise, the test would not make sense.

      • Andrew January 6, 2020 at 2:22 pm #

        Both 42mm and no adjustment to pressure for the difference in tyre volume? If you inflate both to 4 bar the tyre tension will be different between to two wheel sizes and that affects rolling resistance. Surely you should equalize for tension not pressure?

        • Jan Heine January 6, 2020 at 2:36 pm #

          We tested one tire with 2.7 bars just to make sure that small differences in tire pressure don’t matter. We’ve also tested that separately on various occasions: small differences in tire pressure don’t affect performance.

  3. Allen Potter January 6, 2020 at 10:31 am #

    Okay, this is interesting since you’ve covered all kinds of reasons why 650B makes sense on a versatile bike. Here’s a follow-up question: do you think 650B is just as similar to 700C for a tall rider? My basic frame size is 62cm. I doubt you have research about this question, but do you have any anecdotes from tall riders using 650B wheels?

    • Jan Heine January 6, 2020 at 10:38 am #

      How fast a tire rolls won’t be affected by how tall the rider is, so speed-wise, I’m confident that you’ll be as fast on 650B as you’ll be on 700C.

      As to the suitability of 650B for taller riders, there is no reason to suggest it’s not. Ted King is 6’4″ and he rides our 650B x 48 Juniper Ridges. BQ team rider Ryan Hamilton is just as tall, and he loves his 650B x 42 Babyshoe Pass tires.

      As a tall rider, you have the luxury that almost any wheel size will fit your bike without creating toe overlap. So you can choose based on the handling you prefer. Lennard Zinn believes that tall riders often prefer a more stable, less nimble bike. If that is you, then bigger wheels may make sense. You get the ‘cornering as on rails’ feel, but you won’t be able to change line as easily if the corner decreases in radius or a pothole appears.

      • Bern January 7, 2020 at 1:55 am #

        Albert Eisentraut said that larger (that is, heavier, irrespective of height) riders need quicker handling front ends due to the inertial effect of the extra mass. So slightly steeper head tubes for steering, and heavier frame tubing as needed. It occurs to me that road racers who grew up with bikes set up that way might never have felt what it is like to ride a longer wheelbase bike that rails thru corners.

        And then there’s Grant Peterson’s mondo extremuloso wheelbase bikes, inspiring confidence in riders of all sizes…

        • Jan Heine January 7, 2020 at 7:31 am #

          Peter Weigle and I once did an experiment: He built a bike with the geometry of a 1954 Alex Singer (same tire size, too), only his rear end was 20 mm longer to accommodate a pump behind the seat tube. Peter and I took both bikes for a ride, switching back and forth several times. We could not detect any difference in the stability, cornering, etc. And when you think about it, 20 mm out of a total wheelbase of 1000+ mm is only 2%. Unless you go extreme – tandems are a good example – wheelbase won’t affect the handling.

    • Noel Hoffmann January 6, 2020 at 11:41 am #

      I am 6’1″ and ride a 61 cm frame. After many years of riding and racing (road and track) on 700c I have come to prefer 650b and 26″ wheels. My rando bike – used primarily for fast road riding and smooth dirt roads – is equipped with 650bx38mm. I can tell no difference in speed between it and a 700×23 equipped bike. I will note that I no longer care to ride tires narrower than 30mm or so, as they now feel unstable to me, and I cannot corner on them with the same confidence with which I can on a wider tire.

      My all-road bike has supplanted my rando bike (and my mountain bike) for nearly all things, and it is equipped with 26×2.3″ tires. It is among the best-handling and most confidence-inspiring bicycles I have ever ridden, and is only slightly slower than a good racing bike.

      I believe Mr. Heine is on to something with regard to total diameter. It seems to me that if your rim/tire combination approximates that provided by a 700×23 or thereabouts, you will have a workable combination, assuming correct frame geometry and regardless of frame size. At least that is how it has worked for me.

  4. Georgena Terry January 6, 2020 at 10:32 am #

    Excellent analysis. In 2011, I wrote a blog titled “Why are 650c wheels slow? They aren’t.” Hopefully bike manufacturers will eventually get the idea and realize that proper wheel size is essential to proper bike fit. Choosing the correct wheel size (diameter or width) doesn’t compromise performance.

  5. Mike Banks January 6, 2020 at 10:39 am #

    Larger diameter wheels do have lower rolling resistance. When all other things are equal, it boils down to simple physics. When you compare a 28 mm 700c and a 48mm 650B you are just making the 650B roll more like a 700c wheel. The comparison should have used the same width. Even with minuscule differences shown within your trials, you still cannot ignore cumulative differences over many miles of travel.

    • Jan Heine January 6, 2020 at 10:53 am #

      You make a good point: Simple physics indicate that larger wheels roll faster. The question is how much faster. After all, cutting off the zipper pull from your seatpack will also make you faster (less weight).

      Our tests were done with 42 mm Schwalbe Marathon HS tires, so the 700C wheel was a little bigger than the 650B and 26″ wheels. But those differences were too small to make a difference, when compared to all the other factors – which include variability between individual tires.

      In an ideal world, you’d take 50 tires of the same type and test them all, then use the 2 fastest ones. When you look at our test numbers, even though the differences between the tires weren’t statistically relevant, it’s possible that we had a particularly fast set of 650B tires, and a particularly slow set of 26″ tires, even though they used the same casing, rubber and tread pattern according to Schwalbe.

      The renowned cycling expert Jim Papadopoulos once said that we cyclists have a tendency to identify potentially useful metrics, and then argue over meaningless differences. If we had compared a 12″ wheel with a 29″ wheel on the rumble strips, we might have found a meaningful differences. But choosing 700C wheels over 650B won’t make you faster.

      • jjap January 7, 2020 at 3:15 pm #

        Would be fun to see a brompton and 36″ bike added to the test!

  6. Brendan January 6, 2020 at 10:45 am #

    Do you have the overall diameter for 26″x54mm to compare with the above two wheel sizes?

    • Jan Heine January 6, 2020 at 11:04 am #

      Easy to calculate: 559 + 54 + 54 = 667 mm, so about 10 mm smaller, which is how you can get nimble handling with such a wide (and slightly heavier) tire.

  7. mats_norden January 6, 2020 at 10:48 am #

    Interesting all this. Question is now who makes a standard, not bespoke, race bike that can take 650Bx48 or at least 700×40 and fenders (thinking of your 2020 predictions). What comes to mind is Trek Checkpoint, both aluminum and carbon versions. A few small niche ones in UK like, Mason, Fairlight, Kinesis and a some more. Who else will help to fulfill the prediction?

  8. Andrew January 6, 2020 at 11:12 am #

    Jan, didn’t Rene Herse mostly use 700c wheels for his performance bikes from the 1970s onwards? I remember reading that somewhere but of course could be completely wrong.

    • Jan Heine January 6, 2020 at 11:20 am #

      During the 1950s, tires became narrower and 700C wheels became popular, even for randonneurs and cyclotourists. The Concours de Machines had popularized ultra-light all-road bikes with wide tires in the 1930s and 40s, but the last of these events was held in 1949. At the same time, TV brought racing into every living room, and most cyclists emulated racers. Rene Herse resisted this trend longer than most, but when there were no more high-performance 650B tires, it no longer made sense to build 650B bikes. So even Herse switched to 700C tires, and most of his 1960s and 1970s bikes rolled on the 28 and 25 mm tires that his customers demanded.

      Having experienced both the 1950s and 1970s machines from Herse, I found them all wonderful to ride, but I still prefer the older ones with the wider 650B tires. They are faster, more agile and carry a load better. (They are also lighter, because the Concours had every rider dream of a superlight bike.)

  9. Corey January 6, 2020 at 11:25 am #

    Can you say more about the testing methodology here? These bar charts just show wattage by tire.

    • Jan Heine January 6, 2020 at 11:38 am #

      We rode the same bike with three wheelsets and an SRM power meter crank. 3 runs per setup. The smooth surface and the rumble strips were next to each other. No wind and constant temperature. Robust statistical analysis to make sure we were detecting differences in performance and not noise in the data. We did find many differences – tire pressure, tire width, flexible fork (both suspension and flexible steel blades) all affected performance. The full report is in Bicycle Quarterly 29.

  10. Brendan January 6, 2020 at 11:36 am #

    Your article mentions, “was one of the fastest bikes I’ve ridden.” Are you able to list for us the top 5 (or 10) fastest bikes you have ridden thus far?

    • Jan Heine January 6, 2020 at 11:55 am #

      The list of fastest bikes includes a variety of bikes. Basically, once a frame ‘planes’ optimally for us, we can put out the maximum power. Then it’s just a matter of weight and aerodynamics. So my Herse without bottles climbs as fast as the U.P.P.E.R. with two full bottles, etc.

      Many modern carbon bikes work extremely well – it seems that many makers have really optimized the stiffness in the right places. Some ti bikes – like my Firefly – are just as amazing. And the best steel bikes are handicapped only by the extra equipment most of them carry (fenders, lights, etc), and that handicap is much smaller than most people would think. When we pitted the U.P.P.E.R. against Mark’s custom ‘Six Hands,’ the difference in uphill speed was too small to measure.

  11. Tim Cupery January 6, 2020 at 11:41 am #

    It seems pertinent to cover why the MTB market (including racers) moved from 559 to 622. I think you’ve covered this elsewhere and argued that the reasons may be valid, but don’t apply to roads, even rough roads.

    • Jan Heine January 6, 2020 at 11:50 am #

      Yes, mountain bikes are a different thing. I have not much experience with mountain bikes, so I can’t really comment. It’s hard to test off-road conditions and get replicable results. Our rumble strips are probably the best if you want to test rolling resistance/aerodynamics alone. I suspect that many other factors play into what makes an mtb faster…

  12. TIrvine January 6, 2020 at 12:48 pm #

    How do you keep aerodynamic efficiency consistent on an open road test? Also how long did you ride on the rumble strips?

    • Jan Heine January 6, 2020 at 12:56 pm #

      The rider used the same position for each run. We’ve confirmed in the wind tunnel that we’re able to replicate the position reliably. We tested each setup at least three times, to allow comparing the runs. The statistical analysis showed that there was very little variability between runs. Not all open-road testing is reliable – you really need to make sure that conditions don’t change (temp constant, wind must be zero) and that the rider is able to be totally consistent on the bike.

      You’ll have to look up the article about how long the rumble strip section was – somewhere around 400 m (1/4 mile). I was the test rider on that day, and I was pretty beat up after riding more than 15 miles on the rumble strips.

  13. Dr J January 6, 2020 at 1:15 pm #

    So are 700c wheels with 38mm wide Barlow Pass Extralight tires faster than 650b wheels with 38mm Loup Loup Extralight tires?

    • Jan Heine January 6, 2020 at 2:38 pm #

      That is what many ‘experts’ will tell you. When I did some research for this post, I even found a web site that said that the bearing resistance with 650B wheels was greater, because the wheels turn faster for the same speed. Fortunately, our real-road tests take all those things into account…

      • Bern January 7, 2020 at 2:02 am #

        Bearing resistance…at the relative rpm’s of 2 almost identical bicycle wheels?
        Sorry, no…

        • Jan Heine January 7, 2020 at 7:31 am #

          Agreed. In theory, it could make a difference. In practice, it’s too small to be meaningful.

      • Jon B. January 7, 2020 at 3:04 am #

        Interesting that the test was done with heavy duty touring tires, which perform much differently than supple, lightweight tires. How do we know the results would be the same?

        • Jan Heine January 7, 2020 at 7:33 am #

          We tested heavy-duty tires with stiff sidewalls on the bumps because with them, small differences in tire pressure don’t affect the performance. Also, a supple tires provides better suspension, which could mask the wheel’s ability to roll over bumps. We wanted to isolate the wheel’s influence, hence we used a tire that did as little as possible.

          • Derek January 7, 2020 at 4:38 pm #

            It’s good science to try to isolate one variable, but I think by using heavy, stiff tires you may have done the opposite. The overall resistance is so high that any contribution by the variable you want to measure, wheel diameter in this case, becomes a much smaller fraction of the total Watts and therefore harder to detect. I don’t worry too much about my wheel size, but I hypothesize that tire stiffness could cause other very small yet still measurable differences to fall out of significance, rather than emphasize them.

          • Jan Heine January 8, 2020 at 7:19 am #

            Supple tires have more suspension, so the impact of bumps will be reduced. So stiff tires actually should give you a better measure of what the wheel size alone does. Just look at stagecoaches with steel rims on wooden wheels: Their wheels were huge to roll over bumps. Once you introduce some give into the tire, you don’t need such large wheels any longer.

            In any case, we’ve done the test with our own tires, and the result was the same: 650B rolls as fast as 700C.

          • Derek January 9, 2020 at 7:52 am #

            If you’ve already done the the test with your own tires and got the same results, then why on earth didn’t you say that first! Why continue to push a point that you know is moot?

          • Jan Heine January 9, 2020 at 8:55 am #

            We didn’t test our own tires on the really rough surfaces, so the test with the Schwalbes is a better indication that even there, bigger wheels don’t roll faster.

  14. Danno January 6, 2020 at 4:06 pm #

    It seems to me that there is some terminology confusion that goes with this myth. I tend to think of “wheels” as separate from “tires”. One can have a variety of tire sizes for a particular wheel size (700 x ?) but when referencing an individual tire it usually is accompanied by a particular wheel size and width (650b x 48).

    Your graphic that showed a 700 x 28 tire with an equal diameter 650b x 48 was more of an apples to apples comparison to me. What that means is the smaller wheel size needs a wider tire. Then there is the suppleness factor…

    Thanks for the stimulating discussion.

    • Jan Heine January 7, 2020 at 7:25 am #

      The drawing includes the rim, so I labeled it a ‘wheel.’ I’ll change it to ‘tire’ to avoid confusion.

  15. kai January 6, 2020 at 9:26 pm #

    for commuting i tend to advice people to get a a used 559, mostly because they are out of fasion, and thus really nice bikes can be bought for little money.

    moreover they can be parked in the street and not be stolen, because no one wants them.

    on pavement with good tyres they roll as fast as it ever gets.

    myself i commute with brompton having 16 inch wheels. they actually come with surpricingly fast tyres and i have gone in quite fast club peletons no problem. of course 16 inch is on the extreme, but goes to show that on pavement tyre type and suppleness is more important than size.

    as a bonus that bike is always taken inside, so theft is almost not an issue.

  16. singlespeedscott January 7, 2020 at 12:06 am #

    Offroad, the switch from 26″ wheels to 29ers, in the early 2000’s, was like the difference between night and day. 29er’s just made riding technical trails easier, even when only using a hardtail with 80mm of fork travel.

    Look at all the top XC race bikes. The vast majority run 29er wheels

    • Jan Heine January 7, 2020 at 7:27 am #

      Totally, mtbs have gone to bigger wheels, no doubt about it. But road bikes – even all-road and gravel bikes – are a different thing, otherwise, we’d also all run 75 cm-wide flat handlebars!

  17. mobk January 7, 2020 at 4:43 am #

    24″ wheels, 20″ wheels? Do you think they are any slower? If not shouldn’t disciplines like crits and cyclocross with lots of acceleration be ridden on bikes with (relatively) tiny tires? Lighter, stronger, less rotational inertia…

    • Jan Heine January 7, 2020 at 7:35 am #

      They should, because everybody knows that lighter wheels ‘spin up’ faster. I think the main reason smaller wheels (UCI rules allow them 10 cm smaller than 700C x 25) aren’t used on racing bikes with narrow tires is that they make the bike handle less-than-optimally. Also, there may be an optimal rotational inertia to ‘push against’ as you sprint out of the saddle.

  18. DaveS January 7, 2020 at 6:36 am #

    If 700c wheels are faster than 650B (or 26″ wheels), then 36″ wheels would be even faster. Why hasn’t the 36r bicycle taken off? It has been a small niche market for years (google it to see examples) and does not seem to be making any inroads to mainstream bicycle markets. Maybe this will change in the future and Bicycle Quarterly will be doing tire test on these tires?

    • Jan Heine January 7, 2020 at 7:38 am #

      I’ve wondered about the same thing: If 29ers are so much faster than 27.5″ mtbs, why isn’t anybody testing whether a 30.5″ wheel is yet another quantum leap faster? It’s a typical example of the bike industry just following trends, rather than actually thinking outside the box.

      If Bicycle Quarterly was a mountain bike magazine, we’d be testing this as quickly as we could get a tire mold for the bigger tires made. It wouldn’t be cheap, but neither was our four-bike double-blind test of frame stiffness and ‘planing.’ A custom frame and a custom-welded set of rims wouldn’t be a problem, and with tubeless, we won’t even need larger inner tubes. 😉

      • kai January 7, 2020 at 11:28 am #

        i actually since 2 years back own an ride a 30.5er, i e 3 inch tyres on 40i 622 rims. it rolls tarmac about as fast as anything else, outrolls and outperforms smaller and thinner tyres in seriously rocky terrain (obviously), but are heavy uphill.

        i prefer 2.6 tyres on the same 622/40i rims, but do not mind att all riding my old 26er, which has improved tremenduously now being tweaked with plush 2.5 tyres on 559/40i rims. with that setup all in all much closer to the performance of 29ers.

    • Grant January 7, 2020 at 9:05 am #

      mountain biking is a totally different beast because you hit bigger bumps and you hit the more frequently. any time you hit a bump, a portion of the force is directed up (pushing the bike up and over) and a portion of the force is directed backwards, slowing you down. the larger wheels strike bumps earlier, which gives the rider more time to flow up and over and shed less speed. if you’ve ever ridden a skateboard, you know what i’m talking about. you have tiny wheels and you can feel the board drag backwards on “larger” bumps. all that said, larger wheels are worse in every other way. they’re heavier, weaker, more difficult to design the frame around.

      there’s no easy answer for what’s ideal. It’s all about where you ride and how you design the bike. smoother trails, a 26 is faster. bumpier, a 29. if you’re going downhill, you combine longer-travel suspension with the strength and frame-clearance of a smaller wheel, since parasitic losses from the suspension bobbing don’t matter as much, and absolute strength is paramount. i am sure that somewhere on this earth there exists a trail where a 36″ bike would be faster than 29.

      but you’ve picked up on something important. the bike industry tends to operate in the image of professional athletes, and attempts to sell whatever the pros ride. wisdom at the shop begins with realizing that you’re not a professional athlete and have your own needs.

      If you wanted to test the effects of bumps on speed, you need to use bumps that are larger than the tire carcass. So, with a 2″ tire, you’d need to hit 3″ tall bumps. In terms of what is practical, I would go in the other direction and get some 24″ tire/wheel combos, and 20s as well. When you hit 3″ bumps on a 20″ wheel, even with a 2″ tire you can literally feel the bike drag backwards underneath you and shed speed.

      • Grant January 7, 2020 at 9:24 am #

        Jan, another point is that as you hit significant bumps repeatedly, a bump that is over 3″ tall, the larger wheel will not descend as far between each one of them since it will catch the next one sooner.

        I thought of a test:

        Find an uphill staircase and set a camera up on the side. Attack the staircase at a consistent entry speed that is high enough to make it all the way up, but low enough so as not to be dangerous, say, ~12mph.

        I predict that you will find that, no matter how much you try to flow over the stairs, the 29er will have a higher exit speed at the top of the staircase. It will be a small difference, obviously, but I think it will be consistent.

        • Jan Heine January 8, 2020 at 7:16 am #

          That is an interesting point. But would that make a tire at high pressure more efficient on chipseal, since it doesn’t go into the recesses in the pavement like a more supple, softer tire would? In the real world, it doesn’t seem to work that way…

  19. Pk January 7, 2020 at 11:44 am #

    I really think you’re ignoring the Trail Nerds out there who still run 700a and 700b rims on frames so as not to change the trail when swapping tire sizes! /sarc

  20. Tommy January 8, 2020 at 9:23 am #

    This is another great article. So, in theory, if you already had a gravel bike with 700c wheels and can run up to 40-42 mm tires a choice for a second wheelset would be 650b’s so you could run some larger tires? Thanks for your input!

    • Jan Heine January 8, 2020 at 11:48 am #

      Totally. That was how 650B first became popular in the U.S.: People took 700C bikes and converted them to 650B, so they could fit wider tires. With disc brakes, this has become a lot easier.

  21. John C. Wilson January 8, 2020 at 1:02 pm #

    I will continue to agree with Archibald Sharp and David Gordon Wilson and with simple physics that large wheels have less rolling resistance. And can agree with JH that 1% or 2% differences are immaterial. A 1% diameter difference does not make anything like a 1% difference in rolling resistance.

    At the 10% spread between 559 and 622 would note that few wish to ride 559×32 and that even 584×32 is not popular. This was so when tires were stiff and continues today. After adding the tire it is less than 10% but plainly perceptible.

    Nearby is a neighborhood of brick streets. Until fatbikes came along I’d never seen any bike on those streets but my own. This over a period of forty years. And it still remains rare to see bikes on those streets. I never found a good tire and rim combination for that neighborhood until it was possible to get 622×50. Currently Snoqualmie Pass in 622×41 actual is good enough. Those tires of course being in a different league than tires of past. 559×54 (older and stiffer) bogs down on those streets and the ride is so harsh that riding is a stunt, not something done for pleasure.

    • Jacob Musha January 8, 2020 at 5:08 pm #

      Of course, supple tires are now available in sizes like 559×52 and 584×48. So today, bikes with those wheels roll just as fast as a bike with a 622×50 wheel and tire.

      Unless you are saying that the smaller wheels are still slower despite using the same tire construction. If so, please cite your source. “Simple physics” predicting 1-2% is not good enough as small differences may be lost in the noise.

  22. Mitch Harris January 8, 2020 at 2:01 pm #

    “So where does this myth come from?” One origin of the myth might be the huge wheels (2 meters tall rear wheels IIRC) some fast carriages were adopting by the late 18th Century. There could be several reasons, including fashion, for adopting large wheels for fast carriages for travel on roads before tar-macadam surfaces, but they seemed to believe they allowed faster speeds or smoother ride at a given speed which might be a related effect. If those larger wheels were faster/smoother on carriages then maybe iron-shod wheels are more sensitive to diameter effects than pneumatic-tire wheels. Perhaps the angle of attack matters more for an iron wheel and becomes less prominent an effect with pneumatic tires. Early safety bicyclists found this effect too. Some late 19th Century bicyclists who preferred the safety bicycle design still preferred the hard rubber-shod wheels their ordinaries had but this did not become a trend because these smaller hard non-pneumatic wheels were harsher and slower than the big hoops on ordinaries (approaching 2m tall) and also slower than the pneumatic tire show 28″ safety bicycle wheel. The pneumatic tire took over all cycling so fast and so pervasively despite expense added reliability problems of the time, that I can well believe that pneumatic tires made a dramatically noticeable difference people could not ignore. Making wheel size almost irrelevant might have been one of those effects.

    • Jan Heine January 8, 2020 at 2:43 pm #

      I’ve been thinking about this, too. As you mention, the suspension effect of a pneumatic tire will change how the tire rolls over bumps. I suspect this works like this: You first compress the tire as it hits the bump, and on rebound, it pushes the bike upward.

  23. Dg January 9, 2020 at 6:17 am #

    Question: is there a possibility that the difference that is not statistically significant over the relatively short distance in the test would be statistically significant over a longer distance such as a century?

    • Jan Heine January 9, 2020 at 8:49 am #

      Absolutely. The more data you collect, the more you can separate the noise from the real differences. If we ran 100 more tests, the dark gray ‘error bars’ would get smaller, and the actual measurement would move a little inside that dark gray aera: We’d get more precise results.

      Then we could answer some questions: Are the 650B tires really faster, as our raw data indicates? And the 26″ tires really slower? If that is the case, it’s probably because each tire has some tolerances in its thickness, rubber, etc. We’d have to test multiple sets of tires of the same model if we want to isolate the effect of wheel size alone.

      We’ll probably find that a ‘slightly thinner’ 650B tire will roll slightly faster than a ‘slightly thicker’ 700C tire… and vice versa. In other words, the differences due to wheel size are small compared to other factors.

      If you are trying to set a record or win the worlds, it would in theory make sense to take 100 tires and test them all, then select the fastest. In practice, you’ll be better off using that time to train (or rest) and increase your power output by a percent or two. That is the crux of all the marginal gains. In theory, they all add up, but in practice, there are bigger gains available elsewhere, and time and resources are limited.