Political forecasters are having to become much more sophisticated to be credible after the notable lapses of opinion pollsters in recent elections. Stephen Fisher, the Oxford academic behind the Elections Etc website now sensibly draws on several sources, including not just polls but other normally more reliable predictors, such as bookmakers, prediction markets and so-called volunteer forecasts, to calculate the odds on the likely outcome of the Brexit referendum vote. His latest verdict: averaging out all these different sources, Remain is given a 67% chance of winning the vote, but the margin of victory will, his data suggests, be no more than 54% to 46% – similar to the final Scottish referendum vote in fact. Equally surprising to me is that the variation between the results produced by the different methodologies is so narrow (no more than plus or minus 2%). Is it really that close and that certain an outcome? Wiith nine weeks still to go, and so many undecideds still to arrive at a conclusion, there is plenty of room for movement in opinion between now and 23 June. Will the prospect of a close vote, if the figures remain this way, stun the EU into making a much better last minute offer on terms to David Cameron, like the one the Scots were given after that late rogue poll last year? Many undecideds would I am sure vote tactically for just such an outcome if it could be achieved – an intriguing thought, but given how laboriously the EU works, not an option in the real world. Even allowing for the likely existence of a “shy” Remain vote, what does seem clear is that if the figures stay this close, the nervousness in financial markets will continue to deepen as the vote approaches. Watch out also for a “wobble Wednesday” effect if we get another rogue poll, whether reliable or not.