Behold my far-fetched but statistically sound election result forecast. You will by now have heard and seen any number of polls and analyses, confidently "predicting" Thursday's outcome. I expect you are all worldly-wise enough to realise they are a load of tosh. Weary call centre slaves heckle the unsuspecting few by phone, and smug pundits recklessly extrapolate the results to several decimal places of precision. Knowing that you will be unsatisfied with such piffle, I have assembled the most reliable prediction yet - read it if you dare...
Rather than relying on a precarious forecast drawn from a few thousand (possibly imaginary) people across the whole country, this forecast is based on the entire turnout in the North Thanet constituency from the last 6 general elections, some 285,00 people in all. With such a large sample, predictions can be made with a much higher level of confidence.
Starting with the universally recognised statistical technique of drawing a straight line that touches NONE of the data points, we can be assured that the turnout on Thursday will be 43,300.
We can now repeat the exercise with each of the main party's percentage share of the vote. (One of the lines does come perilously close to a data point - I do apologise.)
So, the results will be as follows:
Conservative: 19,139 (44.2%)
Labour: 16,367 (37.8%)
LibDem: 5,153 (11.9%)
Given the sample size, this forecast can be relied on to be about 100 times more statistically significant than any other you have seen. The margin of error is about as wide as the page.
Vote on Thursday, or hang your head in shame.