It's interesting to consider this given canyoners are constantly getting ropes to operate in ways where the forces are not longer evenly spread between the fibres. A knot is an obvious one, but so too is the rope bending over an edge (particularly sharp edges where the rope bends more quickly) or running through a descender.A bent rope means some of the fibres on the outside of the curve will have to carry more load while fibres on the inside may take none of the load. In addition, parts of the rope in a knot may be compressed and the fibres unable to move to share load.
According to one manufacturer website I saw, the assumption for knots like a figure of eight or overhand in nylon or polyester ropes should be that they will result in about 50 per cent reduction in rated strength. Other fibres, types or ropes, and knots can reduce strength even further.
I recently saw a discussion where someone was talking up using a 6mm pullcord to abseil on due to its rated strength of 1000kgs (most canyoning ropes are rated to about three times this). It might sound like a good safety margin, but the moment you run it through a descender (particularly one that turns the rope sharply like an ATC / tube style device) you're going to take it down a lot. Likewise, if abseiling over a sharp edge you'll have the combination of reduced rope strength with higher cutting risk. Extend that to situations where you may use the rope in a rescue, whether hauling someone or abseiling tandem with an injured person, and it starts getting very dicey.
Something to think about when assessing the real-word situations our gear finds itself in. Also a good reason to research the benefits of different ropes and fibres. Depending on what a rope is made from, and how it is made, it can have vastly different properties.