People vote differently whenit is real from when it is not
People vote differently when the choices are more closely balanced from when more distant
People may not vote for their choice if they believe it will win
& may vote against it if they believe this.
Assumed losers are not voted for if they appear they may lose clearly
The greater the perceived differences between the choices, the more likely voting, and the more similar they seem as each other, the less likely, depending on the perceived likelihood of each winning of course.
Phrasing the question differently may change the result. In an ideal world, people should be able to see all ways a question can be phrased, and it ultimately relies on understanding the problem better, or more deeply, or from different biases, and assumptions of the future etc.
People will vote differently if they feel their vote may change the result, ie closer to 50:50 before their vote. More considered, but maybe more conservative and less risk averse unless risk is not risking. Decisions that make risks too great maybe should be broken apart.
Systems with feedback that prevents 3rd options developing can be biased badly when 3rd option is different andbecomes 2nd close option. Then voting may occur for 2nd choice believing it wont win, where voting for option 2 is used as a tool to send a message to option 1.
Often voters can believe they are right and if they lose, it is not because of intelligent, opposition with legitimate moral concerns, but by the lowest common denominator, uneducated, and wrong for bad reasons. – This can be dangerous as it may be seen as others as a view by a population that doesnt exist, that supports them perhaps.
When trailing slightly, this is the best state to be unless winning by a very long way.
When polls oscillate, assume the result will continue the pattern.
By Jim Cook of UK ThinkTank .com