When it comes to polls, it’s been the year of inaccurate polls. When we finally realised they can be wrong, and actually, a bit of a liability. When they don’t get the simple binary result right – yes or no – then what use are they?

Let’s not forget that polls are statistics, which are rarely accurate if they’re not scrupulously scientific. A poll takes a sample size of the population and projects that sample size onto the population as a whole. All of a sudden, polls become ‘the truth’.

Of course, this is flawed in a number of ways. The sample size is probably not be a reflection of society at large. People may lie or change their mind between when they are polled and when they vote. Polls are run by people of a certain societal group prey to their own prejudices and predispositions. The way the poll is calculated may be wrong. And, perhaps most important when you think of the 2 major instances when the polls got it wrong in 2016, the polls don’t reflect the complexities and intricacies of how votes are counted.

Is this the death knell for polls? Probably not, but it’s a useful reality check, a reminder that statistics can be a guide, but often a faulty guide; misinformed, biased, made up or just plain wrong.

As our US friends might say: polls, schmolls…

Advertisements