Whilst scientific investigations that aim to understand the natural world can undoubtedly be demanding and difficult, the last three hundred years have seen huge leaps in our knowledge. In comparison unravelling the complexities of human behaviour sometimes appears more challenging. Even so, the inability of pollsters to predict the recent intentions of voters both in the USA and UK is rather noteworthy. Post event explanations have varied from devious foul play to poor polling sampling methodology and analysis. However I think Theodosius I has a lot to answer for. As the last emperor of a united eastern and western Roman Empire he abolished haruspicy. Consequently YouGov, Gallup and others have missed out on nearly two thousand years of refining entrails analysis as a prediction tool.
Perhaps for the sake of a few chicken livers things might have been very different. Fowl play?