CAN wiring question

So after reading up on the CAN bus outside of the preview of FIRST I realized the wiring of a normal CAN bus is not series it is parallel like I2C for instance. Why do we do it in series? It would seem like parallel would provide more reliability.
I hope this isn’t an obvious question and I’m just missing something but it seems off.

Added a photo displaying a non FRC CAN network.





It actually is setup like that.
The CAN bus is continuous straight through the devices, with each of the devices tapping into the lines as they pass through.

For instance, if we remove power from one Jaguar in the middle of the chain, the rest of the chain still operates.

The stubs are very short as a result.

All of our CAN devices have a straight pass-through from one port to another - a direct electrical connection with nothing in between. The pass through helps to encourage proper design while keeping the spurs (the lines from the bus to the device) as short as possible. The picture you posted is how things work with our CAN bus from an electrical perspective… the “series” configuration just makes wiring a little easier.