A trio of things are worth noting about this poll, especially for longer-standing readers who remember the, er…, patchy predictive value of the constituency polls I reported before the 2015 general election.
First, it’s a properly conducted poll by a reputable company who has published full details of it.
Second, naming candidates such as Luciana Berger in such polling questions has a mixed record. It’s often proved to be a more accurate methodology, especially where a popular Liberal Democrat is in contention. But not always. See 2015, for example, where it sometimes worked better and sometimes didn’t. It’s a reasonable assumption, though, that this is the sort of constituency with the sort of Lib Dem candidate where naming all the party candidates in a poll is going to be a better rather than worse guide.
Third, as Anthony Wells, in particular, pointed out ahead of the 2015 general election, even if each published poll is properly conducted there still can be bias from which polls parties choose to publish. However, what you can tell from the polls that are published is what sort of contests are producing the sorts of results parties are happy to release. In this case, that for the Liberal Democrats a seat where the party was all-but-nowhere before can see a dramatic rise in support with a high-profile candidate. That’s knowledge worth having.