I made more than £3000 betting on the US presidential election. Here’s how I got everything right.

Subject: I explain how I made a tidy sum of money betting on various political outcomes and why the pollsters got it all so wrong.

Length: 868 words, 6-7 minutes

Trump card

trump bet

How exactly did I figure out what the likes of Nate Silver and hundreds of polls could not? I’m not the only one of course, Cernovich, The Wall-Street Playboys and Scott Adams all made similar predictions. And yet the bookies all offered insane odds on what was clearly a sure win for Trump.

So what was everyone else missing? There are two factors that skewed the polls and as a Brit, the Brexit polling was a clear warning of what was about to go down. Recall that ‘leave’ were behind 10 points the night before the results were announced.

Why the polling was so inaccurate

  1. Social desirability bias

This cognitive bias describes how poll respondents will tend to answer questions in a way that will be viewed favorably even when the information is anonymous. The mainstream media were ruthless in their efforts to excoriate both Trump supporters and Brexit supporters, which lead to a much higher level of social desirability bias. Respondents would tend to claim to either support Hillary, support Gary Johnson or be undecided when in fact they were secretly supporting Trump. The same thing was observed both in the UK Brexit vote and the general election just prior to it. Media in the UK call it ‘the shy tory effect’.

If pollsters want a more accurate view of people’s opinions, it is on the media to stop the hate mongering against political opponents.

2. Poll manipulation

Recently, polls have become abused as political tools. There is a belief that if you can manipulate the polls, you can create a sort of self fulfilling prophecy, whereby your opponents will be too demoralized to bother voting. Of course, given the results for Trump and Brexit, it would appear that the opposite is true. Being the underdog is only galvanizing people.

In the US election polling, we saw massive oversampling of democrats and women and undersampling of independents and republicans. Even though the unskewed data showed Trump in the lead by a significant margin, bookies still set their odds based on skewed data, leading to Trump having odds of 7 to 1 just a few weeks before the results.

How to get a more accurate read on public opinion

The elections of the future will be predicted through social media interaction. Trump was the first presidential candidate to make full use of social media. Take a look at the below comparison of the presidential candidate’s social media interactions; the two snapshots were taken at the exact same period. If it isn’t clear from the numbers who is more popular, it should be clear from the content. Trump is engaging with everyday Americans in a way that they understand and appreciate.

Trump social media

Google search results are also a much better predictor of who will win the presidency than any poll aggregator. Take a look at the image below, which correctly predicts the last 3 presidential election winners.

1469135757932

And if you needed anymore proof, take a look at what happened on November 6th, when the number of people googling Trump actually overtook the number of people googling porn.

Google predictions

This was all just a fluke! Why are you making such a big deal about this?

Sure, you could make the claim that I simply got lucky betting on Trump. At least you could, if that had been the only bet I’d made. But I didn’t just predict that Trump would win the presidency. I also put money on a number of other obscure long shots – and won those bets too.

I correctly predicted that Gary Johnson would receive less than 5% of the popular vote, despite the fact that he was polling at much more than that. I knew that social desirability bias meant that most proclaimed Johnson voters were actually Trump supporters.

PA betAnd I also predicted that Trump would win Pennsylvania, a state that hasn’t voted republican since the 80s. This was an easy call, given the high levels of disenfranchised steel workers that support Trump’s isolationist economic policies and large numbers of Amish and evangelical residents that like Trump’s conservatism.

Why go to such lengths? For social proof. On this blog I give out a lot of life advice. You should only follow the advice of people that can demonstrate they know what they’re doing. I knew that my world view was accurate, but I needed to prove that to others. The easiest way was to make some predictions about something. And by putting money down, you can see for sure that I’m not faking. I not only believe in my own views, I’m willing to risk capital to show that they’re correct.

Hopefully reading this has increased your faith in the things I suggest. If you want to stay in the know for the next political event, or receive other valuable life advice for free, sign up to the mailing list at the top of the side bar, or follow me on Twitter.