Could some one explain US polls in the context of UK polls?

[Harry Hayfield]

Here in the UK when a poll is published we get the following information:

Conservatives leading
Labour losing
Liberal Democrats within striking distance of Lab
Others with half the support of the Lib Dems
which is then followed by the number of people polled, the margin of error and the date of the poll being conducted.

Most UK polls do list those who are undecided but use a weighting mechanism so that all numbers relate to those who will actually vote. Do US polls use a similar method?

[Verily]

Generally, no. Most pollsters require voters to volunteer "Other" unless a significant third candidate is running (see the Maine and Texas gubernatorial elections in 2006). This makes sense in the US context as minor parties are far more irrelevant here than in Britain (where results such as Wyre Forest and Brighton Pavilion are not uncommon and where regional parties win seats).

Undecided voters are sometimes pushed to make a choice ("Who do you lean towards?"), but there are always some undecided voters included in a release.

Many pollsters do not weight their results at all; those who do weight only by partisan affiliation and demographics, not by past vote recall the way UK polls do (or with ranked likelihood to vote).

Polls vary from being of "registered voters" (those who say they are registered), "likely voters" (those who say they are registered and likely or certain to vote) or "all voters" (only question asked to determine likelihood to vote is age).

[Alcon]

By the way, an important UK/US difference:  UK polls report the sample as the total number of interviews conducted; US polls report only the total number of accepted, processed responses.

That is, in the UK, sample size includes voters who aren't registered voters (or likely voters), or refused to answer the rest of the survey for whatever reason.  Thus UK polls appear "bigger" when they are not.

[Hash]

Sadly, no.