>The DH3 story is implausible for voting methods that elect within the <br>top cycle.<br><br>wrong.<br><br>> Each voter considering raising the turkey candidate over <br>rival candidates will know that, unlike with Borda, the outcome won't
<br>improve if only a small number of voters strategically raise the turkey. <br>The outcome can change only if so many voters raise the turkey that the <br>turkey reaches the top cycle.<br><br>for a voter who prefers A>B>C>D, using this strategy may cause the election of A or B instead of C - an improvement.
<br><br>> Strategically minded voters would <br>understand that the strategic vote could only help if it also creates a <br>large risk of electing the turkey.<br><br>and the total "expected value" of the strategy is the probability of a benefit times the amount of benefit minus the chances of electing the pawn times the detriment of doing so. and it's pretty plausible that the expected value is positive.
<br><br>for example, say i preferred<br><br>Obama>Edwards>Clinton>Biden<br><br>if i think you're right that voters aren't going to use this strategy, then i'm going to vote<br><br>Obama>Biden>Edwards>Clinton
<br><br>and i'll be assured that this has a better chance of helping obama beat out edwards and clinton than of electing biden. and if the average of my preference of obama over edwards and clinton is greater than the average of my preference of edwards and clinton over biden, then the strategy is all the more justified.
<br>