I'm looking into coercion-resistant voting schemes and one of the key elements of coercion-resistance is resistance to randomization attacks.
Juels et al. define a randomization attack as one where adversary coerces the voter to vote for a random candidate (as drawn from some distribution). Many other articles have since repeated this as a requirement for coercion-resistant voting schemes, so it seems to be important.
Since coercion-resistance is an extension of receipt-freeness, and receipt-freeness already guarantees that a voter can't be coerced to vote for a particular candidate, I'm having a lot of difficulty imagining a scenario where both of the following properties hold:
- Adversary is unable to force a voter to vote for a particular candidate
- Adversary is able to force a voter to vote for a random candidate
Can you provide an example of such a scenario?