I clearly misinterpreted what you wrote, but what you wrote doesn't seem to constitute a complete strategy. I still don't know who to vote for unless you can define "candidate you like". If it means better than the average, than this strategy is hardly optimal.However, the optimal strategy with information obtained from polls, is to vote for yourNo again. If your favorite candidate is in last place in the polls, this strategy suggests you should vote for every candidate, which is obviously a waste.
favourite and any other candidate you like that the poll says it would beat your
favourite.
At any rate, the rest of what I wrote, about the well-worn optimal approval strategies, stands.
What you probably meant to say is, "vote for **the candidate you would vote for in a plurality election** and any other candidate you like that the poll says it would beat **that candidate**. This assumes that you would vote "lesser-of-two-evils" in a plurality race.
This strategy is pretty good, but it's not quite the best - there are situations where this approach could make you regret your vote, even if the polls are fairly accurate. A better strategy is, "vote for anyone you like more than the front runner, and the front runner as well if you like him/her more than the second place candidate." The only difference here is that if you like the second place candidate more than the first place candidate, you will vote for candidates you like more than one but less than the other.
-Adam
----
For more information about this list (subscribe, unsubscribe, FAQ, etc), please see http://www.eskimo.com/~robla/em
