[sdkjfwiluf]

what was the conclusion ? it's not gyroscopic effects because they still balance with counter-rotating wheels, it's not trail because they can still balance with negative trail. Is it some of each of these effects, if so how much ? What else is involved ?

[jessaustin]

According to TFA (and to this avid cyclist it makes sense) it is the combination of the two effects you mention and weight distribution. Novel bikes can be constructed that tweak one of these inputs into an unsuitable range and yet still balance, but I really doubt you could mess up all three and still have a bicycle.

[sliverstorm]

I think negative trail works pretty much the same way as positive trail; I think the question is, does a bike with zero trail work. I'm guessing it can be done, but experience with various motorcycles (which have a wide range of rake) suggest trail is desirable for other reasons too.

[IshKebab]

Well I haven't read the article (sorry), but on a bike with a rider you balance by active feedback. If you steer the way you are falling to right yourself.