I'm currently reading The Bicycle Wheel by Jobst Brandt, and I'm reading all of the issues with offsetting the rear rim toward the drive side of the hub...

Asymmetry in a rear wheel built for a multispeed gear cluster makes it weaker against side loads coming from the right side. This means that the wheel bends to the left more easily than the right. The greater the offset, the weaker the wheel. In order to reduce offset for six-, seven- and eight-speed clusters, hubs have been built with narrower flange spacing. Although they reduce the disparity in tension between left-side and right-side spokes, they weaken the wheel against lateral forces.

So, I'm curious, why is it the standard to do so? Why not center the rim to the hub just like the front, and then offset the frame's drive side rear triangle to accomodate the difference? It may shift the chain line a bit, but this doesn't seem like anything that challenging to adjust for. The strength and durability benefits are huge, and the only downside I can see is that you couldn't arbitrarily change rear hub widths (not that it seems they change that often anyway), or you may have to introduce the offset again. Why don't bicycle manufacturers start doing this instead? Do they make frames that you can use a rear wheel built with the rim centered on the rear hub?

Rear Wheel Offset --> Rear Wheel Offset

  • I know they make offset rear rims to accomodate this, but this is another compromise as the rims are weaker on one side. It seems we try everywhere else, Why not the frame?
    – Ehryk
    Apr 14, 2012 at 8:16
  • 1
    1) Tradition, 2) It would make mating wheel to bike (and getting brakes centered, etc) harder, 3) It places uneven stress on the axle and bearings. Apr 14, 2012 at 9:47
  • 1) if it's just tradition, then it's time someone shakes it up, 2) How do you figure? It would go in centered just like any other wheel, and 3) what causes the uneven stress? That one bearing is farther from center than the other? Both should be under their limit, so is there something wrong with uneven stress?
    – Ehryk
    Apr 14, 2012 at 16:10
  • For (2) the rim would not be centered between the axle ends, so the bike would have to be set up with the same offset as the wheel. Change number of cogs, eg, and you change centering. Apr 14, 2012 at 18:07
  • Not exactly true. Change rear axle width, and you're right, but if you take a mountain bike hub, for example, you can put 6, 7, 8, 9 speed (6 and 7 with spacers) or single speed conversion cassettes on the same hub and everything would still be centered. 10/11 may require a different frame.
    – Ehryk
    Apr 14, 2012 at 18:15

4 Answers 4


The biggest functional reason actually would be chain line.

You could make the hub shell wider, and run an offset dropout to allow the space for the gears, but then you would have to run a similar offset on the bottom bracket to maintain a usable chain line.

Running the offset on the bottom bracket would affect Q factor positioning on many riders, and while some riders like or need additional stance width, most do not.

Because a change like this requires such a radical redesign of pretty much every bicycle component, it would require a consensus from the entire range of bike manufacturers to be put into practical or wide spread use.

So you could call it tradition, but I think momentum would be more accurate. The existing body of design work would need to radically change to accommodate something like this, and that is a lot of "mass" to shift.

Since existing wheel structures and designs are working for riders as is, most companies choose to stay within the range of difference that allows the bike frame and other component designs to be affected as little as possible.

After all, there is no guarantee that an offset design would work in practice, and the benefit to it isn't major, given that existing designs work.

  • Agreed. Q factor is the distance between the pedals, and any multi-geared bike has the crank arms already passing very very close to the chainstays. A modification of this could be problematic. And, as pointed, after all the current designs are working fine (I personally think Jobst Brandt's arguments are greatly speculative and lack any experimental proof, not to say some of them seem not to make sense) Apr 14, 2012 at 14:50
  • BUT, since I started using wide-spaced, big-diameter, symmetrical flange hubs (front and rear), the robustness and durability of the wheels (measured as "resistence to abbuse and neglect") are very objectively perceived. Apr 14, 2012 at 14:55
  • 1
    The benefits to me seem non-trivial: stronger wheels instead of lopsided torsional strengths (stronger in one direction than the other), even spoke tension between sides, equal spoke lengths which make ordering and building much easier, no need for offset rear rims (which are weaker anyway), even flange tension between DS/NDS....
    – Ehryk
    Apr 15, 2012 at 1:57
  • 2
    The benefits do not outweigh the disadvantages. When was the last time you heard about a wheel failure that wasn't the direct result of some stupidity on the part of the rider? Current wheels don't fail when you're just riding.
    – zenbike
    Apr 15, 2012 at 3:13
  • 1
    Hello! Cannondale Ai since 2017?
    – Noise
    May 11, 2021 at 9:07

At first I read you to mean an off-center wheel plane, but not so. Interesting question, i.e. does a bike remain stable/rideable if rear axle is not symetrical? I guess yes, as long as front/rear wheels are in line, but practical clearance issues immediately become apparent.

Tolerances in modern frames and wheel-drivetrain designs already nearly max out every clearance there is. 2mm is a huge gap, exceeding the clearance between chain and right rear seatstay in the smallest rear sprocket; since the wider rear hub of 130mm (135mm in tandems) years back, clearance between rider's ankles and the right chain stay can cause contact. Solutions to spoke issues have included narrower rear hub flanges, and asymetrical rear rims, as you mention, but I have seen no studies documenting significant weaknesses in either design solution; weaknesses instead occur in poor manufacturing criteria, quality control, loose spokes, etc. No studies (beyond pure theory) indicate a well-made rear wheel has a dramatically weaker resistance laterally from one side versus the other? Has there been a rash of left-turning rider crashes due to rear wheel failures I haven't heard about?

As mfg. insist on solving more non-existent problems (ie 47-sp rear cogsets) eventually every current frame will be obsolete, and your solution of a wider rear axle standard* will be reality, but not to solve your problem. It will instead create new problems, all of which will only be "solvable" by purchasing all-new bikes, totally incompatible with all that came before. * It's here - Shimano 11-sp rear hubs actually are about 1.5mm wider than 130, which is considered safe to cram into rigid carbon rear drop-outs. It may or may not provide for chain clearance on current frames, and requires all-new incompatible chain as well. Why are the frogs in the warming pot not screaming yet?


There are two alternatives to asymmetry in the spoke tensions of rear wheel:

  • Asymmetric frame
  • Asymmetric rim (with non-centered spoke holes)

Traditionally, the way asymmetry is handled is in spoke tensions. It should be noted that the author of The Bicycle Wheel is a prophet of doom. There are the following examples in the book as the degree of asymmetry:

  • Six-speed: 1.6 : 1
  • Seven-speed: 2.0 : 1
  • Eight-speed: 2.5 : 1

It is implied (but not said) that adding every speed increases the ratio by approximately 0.5 so that eleven-speed would be 4.0 : 1. This is not the case. Firstly, all Shimano freehubs between eight-speed and ten-speed use the same freehub width because they are cramming more sprockets in the same space. Secondly, 2.5 : 1 is not certainly true for Shimano 8-speed freehubs. For example, Shimano FH-T610 135mm OLD freehub has 1.8 : 1 as the asymmetry. This is partly caused by it being 135mm as opposed to 130mm which road freehubs are. However, even Shimano FH-5600 is 2.1 : 1 as an example of a 130mm OLD hub and certainly not 2.5 : 1. Even Shimano FH-5800 (130mm OLD, 11 speed) is only 2.2 : 1 and not 2.5 : 1. So it appears that even Shimano 11-speed 130mm OLD freehubs, the most extreme asymmetry there is, are only 2.2 : 1 so the 2.5 : 1 stated in the book is very suspicious.

Secondly, road freehubs are moving due to adoption of disc brakes towards 142mm M12 thru axle (which roughly corresponds to 135mm quick release and is certainly much wider than 130mm quick release). FH-R7070, an example of 11-speed 142mm M12 hub, is 1.8 : 1, the same as 8-speed 135mm OLD quick release rim brake hubs.

So the problem is not as bad as it is claimed in the book. Certainly however an over 100 kg rider should avoid using a 130mm OLD 11-speed road bike in offroad situations if using 23mm tires pumped to over 120 psi, especially if the spoke count is lower than the optimal 36, but that should be obvious anyway.

So let's consider the alternatives to spoke tension asymmetry:

Rim asymmetry

The problem of rim asymmetry is that it typically modifies the rear rim to be different from the front rim. Thus, every bicyclist having a stock of rims in home needs after this modification two stocks of rims: a stock of rear rims and a stock of front rims. The doubled rim stock costs makes this solution rather poor. There are two alternatives to this: firstly, it is possible to make the rim asymmetry half of the rear wheel natural spoke tension asymmetry. It might allow using the same asymmetric rims in both front and rear wheels. Secondly, with the advent of disc brakes, it is possible to take the rear rim, reverse it and install it to a front wheel. Disc brake front wheels too are asymmetric but in the opposite way.

Traditionally bicycles used rim brakes and the desire to have only one rim type made rim asymmetry quite rare. However, disc brakes and reversing the rear rim in the front could make this solution viable today. I wonder why asymmetric rims are not becoming very common today. Perhaps bike manufacturers have not realized the usefulness of asymmetric rims as the standard rim in disc brake bicycles.

Frame asymmetry

The problem of frame asymmetry is chainline. Today the best bottom bracket standard is probably Hollowtech II and Hollowtech II cranksets have fixed chainline. You can't adjust it like you could with square taper bottom brackets. So an asymmetric frame would have a chainline optimized only for the big ring, and the little ring would be useful only with the very largest sprockets. Not very desirable. Of course an option could be to use non-standard crankset parts for modified chainline.

Today, with the advent of mid-drive electric bikes this frame asymmetry may actually be useful. Mid-drive units are rather wide so they typically are used with only one chainring. If you want a typical road bike 2x setup in a mid-drive bike, the logical way to install the second ring is to increase the Q factor of the crankset and put the second chainring in the freed space. With standard non-asymmetric frame, the second chainring would be too outboard. However, with an asymmetric frame the second chainring could be just in a useful position.

My electric bike, Cannondale Synapse Neo EQ, uses this concept of using asymmetric frame with mid-drive. It has a frame asymmetry of 6mm, called Cannondale AI. I think it's useful in this very bike although for non-electric bikes it would be less useful.


As a builder of home bicycling contraptions myself (recumbents not bicycles) but also as a nationally competitive TT'ist (seniors) I have a comment on centering.

Of course making the spoke offset symmetrical would (as mentioned prior) push the cogset far out, perhaps in 35-40mm range (just a rough guess).

To keep the chain line correct would force the right pedal to shift out to match (as mentioned prior also). It has been noted in cycling literature that it is no coincidence that the world's best TT'ist also ride with their knees close together. It has also been documented in MIT's wind tunnel that the gap between the legs is a major contributor to air drag (noting that a bottle on the seat tube sometimes lowers air drag because it fills that gap).

Now, if you are talking about riding MB's, aero doesn't matter much but if you a road guy who'd sell his soul for .1 mph, then it does.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.