Voting with 1 against your most preferred candidate, 2 against the next and so on, instead of just an X, or punching a single chad, or whatever.
Why would you vote against your preferred candidate?
Say your preferred candidate is X, but you know he's an underdog and likely to take a bath. Rather than totally wasting your vote you have a second preference, Y, which has a much better chance against your hated non-preference Z. There is a slim chance that X will get in, so you vote for him first. If he fails, your vote has a fallback plan.
This system is more likely than a single vote system to result in a consensus being formed.
Take, for example, a situation where:
10,000 voters love candidate A, but failing that would like C. They hate candidate B.
9,000 voters love candidate B, but failing that would like C. They hate candidate A.
8,000 voters love candidate C, but failing that would like B. They hate candidate A.
In preferential voting, candidate C would be elected. 8,000 very happy voters and 19,000 mildly satisfied voters.
In a FirstPastThePost
system, candidate A would be elected, resulting in 10,000 very happy voters and 18,000 very disappointed voters, though if the voters are aware of the likely voting pattern, the candidate C voters would manually switch their votes to candidate B - a best case situation resulting in 9,000 happy, 8,000 satisfied and 10,000 unhappy. Vote switching tends to only be partial though - possibly resulting in a victory for A despite knowledge of the system. This will have a greater effect where B and C have roughly equivalent popularity.
A more realistic situation would be:
49% candidate A (ie. Republican)
3% candidate B (ie. Green)
48% candidate C (ie. Democrat)