They have much more interesting videos on the homepage of stringbike itself: http://stringbike.com/

The lowermost video on this page shows the real bike from an angle where you can see what happens to the wires:

http://www.stringbike.com/Internal%20symmetry%20of%20both%20...

Think of the system on the back wheel like a tape measure. The string is inside wound around the axle. As the pedal moves the string is pulled out winding the axle and turning the wheel.

I think the advantages are because of the location of the torque. When the left pedal exerts force, it torques the left side of the bike which coincides with the the natural movement of the bike. Imagine when a cyclist pedals very hard and the bike sways from side to side from the force exerted on the pedals. Now I'm not sure if they are trying to reduce the movement of the bike, or make it feel more natural.

The thing I cannot make out is the switch from gears to a single system. Is the output of work the same as the force going in, which would mean that you have to pedal the same amount going up a hill as going down, or is there a variability of work like a rowing machine? The rowing machine system would seem to apply here but they do not address that so I'm not 100% on that.

Also, there's no explanation of the symmetrical aspect of the design. I'm intrigued by it, but other than vague allusions to its superiority over asymmetrical designs no analysis is given.

So, I ask HN: why aren't chain/belt bike designs symmetrical?

I think what happens is that the wires are fixed length, but wrapped around a ring on the back axle. As the pedal is pushed down, it moves the pulley, unwrapping the wire from the ring, which rotates the rear wheel through a ratcheting mechanism.

I'm not sure what causes the wire to rewrap itself around the ring on the upstroke... probably some sort of embedded spring.

Yeah I'm confused too. Looking at it, it seems like the string is elastic and pedaling builds tension or some thing like that.