I got predictions of that result, from those in the know, to within about 1%. I ignored them.
What makes that even more stupid is that the same sources told me the outcome of the referendum pretty much spot on. I ignored them then as well and clearly failed to learn the lesson.
Its intelligence gathered from street level campaigning and canvassing. It tends to reflect the polls reasonably closely, usually with a particular slight one way but the prediction is recognisable. On these two occasions it wasnt, at all. I assumed it was wrong. It was wrong in 92 and again in 2015. I ignored it both times and was proved right.
So, its roughly 2-2