falling off bikes since 1975

I’ve had a longer look through the experiments proposed by the 29er faithful for the ultimate normalised 26 versus 29er test. This was in response to their criticism of the Cycling News comparison.

For all their whinging all bar one of them has failed to address all of the concerns that they themselves raised against the Cycling News/Angry Asian tests. Subjectivity remains - though many of them recognised and acknowledged the fact that there are too many variables and that whatever you test you have ignored other variables.

Some of them even realised that the first hurdle “What is the CRITERIA!!!???” was a difficult question to answer in itself.

That’s why I didn’t give them one.

If they now look back over their own experimental proposals with the same critical eye that they applied to the Cycling News comparison, then I hope they are man enough to realise that they owe James Huang an apology.

There is just one exception, I’ll mention him later.

Here’s the main faults they listed against Cycling News, and whether they addressed them, or not.

Head Angle/Trail

Plenty of them picked up on the head-tube angle/trail and offset issue, but no-one solved it. I think they should now realise that it is impossible to normalise this.

But is it really such a big issue for performance? In the real world there can be more variation between two regular off-the peg 26″ wheel bikes, or between two 29ers than there can be between a 26er and a 29er.

Well, handling is always subjective, but to test pure rolling performance of 26 or 29 inch wheels it is irrelevant.

That’s not to say it can be ignored. Once you’ve shown, separately, that big wheels roll better then it follows that 29ers have a whole new world of line choices open to them that can take advantage of this and thus increase the overall advantage.

What no-one addressed is whether in this case rolling beats damping. By limiting the test bikes to rigid or short travel designs they created unrealistic 26 inch wheeled bikes - you try getting hold of a fork with less than 100mm of travel these days. When 26 v 29 tests restrict the 29er in such an unrealistic way all hell breaks loose and the test is immediately dismissed as biased.

Chainstay Length & Wheelbase

While plenty of them said that they’d normalise these no-one had an ideal solution. Plenty of them have problems with current tests clearly favouring the 26er. It’s interesting that instead of justifying different chainstay lengths, as they are in the real world, 29er advocates design experiments that equalise chainstay lengths, but immediately throw objectivity out the window by making the 26er chainstays too long.

Sorry, but as soon as they’ve done that they’ve invalidated the experiment. And not just on performance grounds either. Longer chainstays subjectively feel different, and objectively perform differently. Otherwise they’d not riase such a hoo-haa about them in the first place.

Gearing

Cycling News have not given their 26 and 29 inch comparison bikes the same gear ranges.

For all their whining, apart from those who made it singlespeed, only one person addressed this issue.

Wheel stiffness

Not one person addressed this issue.

Note that this can be tested completely separately, using strain gauges, and some big weights.

The Rest…

Some of the experiments raise even more variables, and didn’t say how they’d be addressed. Can you imagine the flaming they’d have to give themselves if they printed these experiments in a mainstream magazine.

It’s also interesting to see that current tests out in the public domain are criticised for optimising for the 26 inch wheeled bike, and trying to replicate that with a 29er. I agree with those of them who cry that this is a faulty basis/bias to begin with. So what do you make of tests which then optimise for the 29er and try to replicate that with a 26er? Will they admit that this too is flawed?

And the prize goes to…

Only one person followed the scientific method and realised that there are too many variables, and proposed a hyopthesis that attempted to minimise these variables and could be tested. He designed a test that nearly achieves it.

Like I said he has eliminated or explained away most of the variables. He hasn’t addressed handling or wheelbase which will have an effect even on his experiment. If you think that under motor-power they won’t you might like to consider why some racing motorbikes come with adjustable headstocks.

He has also limited the bikes to rigid forks. This is not very representative of the real world, but is necessary to eliminate one of the test variables. Subsequent tests could be done to see if suspension went some way to compensating for smaller wheel diameter, but the initial test serves as a good baseline. I know this test is not meant to consider whether rigid or suspension is better, but as the line choices open to 29ers has been debated as a factor on experimental design and one of the reaons why 29ers feel better I would have liked to see rigid v suspension discussed and explained away. Especially as the test involves riding over baby-heads.

To accurately test PeTs hypothesis and eliminate the remaining variables his experiment could probably be performed using an electric motor driving a single wheel over a series of rolling roads with variable sized obstacles/surfaces eliminating the bike altogether. I have no doubt that this would prove that the 29er wheel is more efficient.

But the real prize goes to…

Those few of them actually got the real point, there is no point in these sorts of tests. You should just ride what works for you.

The rest of them?

Looking for concrete proof smacks to me of insecurity and a need to have their choices validated. In which case only a test that proves that 29ers are better will ever satisfy them.