In a first‐past‐the‐post electoral system a vote is tactical when it is cast for a candidate that the voter believes is more likely to win than their preferred candidate, to best influence who wins in the constituency. The classic example involves a supporter of a party placed third or lower in the constituency choosing to vote for one of the front‐runners because they are wary of ‘wasting’ their vote. The wasted‐vote logic is sometimes referred to as ‘Duverger's psychological effect’ and is one of the factors that is thought to drive Duverger's law. Formal rational choice theory shows us that it is not necessarily utility‐maximizing to desert third or lower placed parties. Someone indifferent between the two front‐runners will have little incentive to vote tactically for either of them. Also, tactical voting is not simply about voting against the incumbent, the most likely winner, or even the most disliked party. Although it often involves these things, people can vote tactically for the incumbent party even when the most disliked party is likely to come last. In recent general elections in Britain between 1987 and 2005, between 5 and 10 per cent of voters are estimated to have voted tactically. Perhaps as many as forty seats in 1997, 2001 and 2005 were lost by the Conservatives as a result, although not all tactical voting is anti‐Conservative.
There are also strategic incentives to misrepresent one's preferences in other electoral systems (see Gibbard–Satterthwaite theorem). Typically they are much weaker and the frequency of strategic behaviour is correspondingly lower. Whilst the terms tactical voting and strategic voting are synonymous, sophisticated voting generally refers to behaviour in (particularly US Congressional) committees where there are defined agenda rules and legislators vote in a sequence of decisions.