Polls and the B.S. Detector
Polls and the B.S. Detector
August 14, 2016
In my book – that you might enjoy because it is full of words – I talk about developing your own personal bullsh*t detector. I won’t rewrite that chapter here, but I’ll tell you how my own B.S. detector sees the accuracy of the presidential polls.
One of the things I look for when I’m trying to detect B.S. is to see if two or more sources of information are in agreement. The polls showing Clinton with a solid lead seem to be in agreement, so that is strong evidence that it is a true snapshot of current sentiment. Some polls could be biased, and probably are. But all of them? That seems unlikely.
Common sense tells me there are enough anti-Trumpers in the world that they might indeed form a solid majority at this point in history. That’s not impossible on the surface of things. Common sense does not conflict with the idea that Clinton is ahead.
Now consider the size of the gatherings for Trump rallies versus Clinton events. That suggests a huge under-polling of Trump supporters. But another explanation is equally credible: Trump is more entertaining. That alone could explain the difference in event attendance.
Trump also dominates on Internet engagement stats, and he does better with online polls than with phone polls. But that could be nothing but a sign that he has more energy on the Internet. It doesn’t directly translate to votes.
So what are the strongest arguments that the polls are wrong and Trump will win in the end?
Anecdotally, many Trump supporters know other Trump supporters who won’t admit their Trump support, even to loved ones, much less to pollsters calling their homes. It just isn’t safe to support Trump in many parts of America. I live in one of those places, and that’s why I endorsed Hillary Clinton for my personal safety. It’s just safer. (And yes, I am totally serious.)
The so-called Shy Trump Supporter is real, but we have no accurate way to measure them. Likewise, we have no way to measure people who haven’t yet been motivated to register but might later.
We also have to ask ourselves whether it is possible for all of the separate polls to be “rigged” in favor of Clinton. My B.S. filter says that whenever you have a situation in which there is a lot to gain, opportunity for cheating, and a low risk of getting caught, shenanigans always happen. So I expect, based on that universal law alone, that SOME of the polls are rigged and SOME of the actual election will be rigged as well.
But since all polls show Trump behind, and it is deeply unlikely that all polls are rigged, my best guess is that only the outlier polls are rigged, or at least inaccurate. Trump is probably down, but not as much as poll averages suggest.
My best guess is that Trump is genuinely behind in the polls, and unless something big changes, he will lose the election.
But something big always happens. Probably several big things will happen between now and November. And it might include one or more of these things:
1. New Clinton health issue or revelation
2. New Wikileak that is more damaging than what we have seen.
3. New Clinton Foundation revelations worse than what we have seen.
4. Trump makes an uncharacteristically empathetic speech that shows he can take advice, is not irrational, and that he loves all Americans.
5. The Shy Trump Supporter is really a monster size.
6. Godzilla changes sides.
7. I take sides.
8. A major terror event.
9. Trump outperforms expectations in the debates (especially the first one).
Anything can happen. But I think there are more potential shocks on the Clinton side because any bad news about Trump’s character or business dealings are already baked into the cake. He is virtually shock-proof. Clinton is not.
I still predict a Trump landslide, based on the 3rd act movie formula. Trump is in his deepest hole right now. This is when the surprise happens (next two months) if it is going to happen. He’s had other deep holes, but none as deep as this. This is the big one because time is running out.
For new readers, I don’t vote, and I don’t support the policies of either candidate. My political preferences are quite different from both. And I think it is insane to elect 70-year-olds to a job that requires so much energy and mental agility. You wouldn’t hire a 70-year-old for any other type of job that they had never held. Why does it make sense here? (Answer: cognitive dissonance)