according to the brext book by mark goodwin which analysed survey and stastistic results that 34% figure is definately incorrect.
given a sample size of a million referendums, the bell curve suggests that remain would have a ‘win’ by a very slight margin, however it is that close that if it was re-run right now the result would more than likely stay the same and that the leave side would win. the book is based on solid verifiable data
Which book is that? I could only find a book by a Matthew Goodwin, which was published in 2017, so surely not relevant to gauge sentiment now.