The most recent predictions by the pundits is that Republicans have a 15% chance of keeping the House. The same pundits say Republicans have an 85% chance of keeping the Senate. It should be pointed out that in 2016, the same people predicted that Donald Trump had a 16% chance of winning the presidency. That is why Hillary Clinton ordered a barge loaded with fireworks for her victory party.
But we now know that at least some people running polls for Hillary Clinton were warning her that Trump could win. These people were, as expected, shouted down by more brilliant, but wrong, advisors.
This is happening again. Only this time those people predicting the blue wave are already starting to back off their predictions. I thought it would be interesting for people to know where the 15% came from in the first place. It all has to do with the bell curve.
Following is a typical bell curve. I used to teach statistics at the college level, so this is very familiar to me.
This is basically the standard graph used to predict results based on a series of data. In this case the graph is showing that about 68.3% of predictable data points falls within one standard deviation and 95.4% falls within two standard deviations. Hence a prediction of results falling with the 85 percentile is actually less reliable than a normal distribution because one would expect 95% confidence that the results would fall within two standard deviations.
In addition, bell curves are not perfect. In the case of elections, the problem may be with regard to the data used to develop the curve. In 2016 I looked at the underlining data and discovered that every pollster was predicting turnout models that were significantly different that those from more recent elections. When I adjusted the data to compensate for that, I was shocked to see that it was Trump who was more likely to win. I did write about this in my blog, but this was so different from what the experts were reporting that I was pretty subdued in my analysis. Following is what I wrote on November 1, 2016. It turned out to be very accurate:
On November 2, 2016 I wrote about reports that Republican turnout in early voting was higher than expected. The MSM ignored this, to their ultimate peril:
On November 7, 2016 I noted that if the underlying data was accurate, Trump would win. This was such a departure from what was being reported on TV that I almost didn’t publish this blog.
The purpose of this is not to gloat about my analysis, it was pretty rudimentary. It was rather to point out the flaws in the analysis by the talking heads on TV. This year we don’t have all those polls and we don’t have the underlying data. Neither do they. This makes it much harder to predict. But, I can say this with confidence. Those bold predictions that Democrats have an 85% chance of winning the House are based on scarce or none existent data. In addition, this is based on a bell curve showing a normal distribution. In real life, particularly regarding politics, the least likely outcome is a normal distribution. That is why everyone is talking about turnout. Turnout changes everything. So, the real question is who is motivated to vote and who is not.
It is clear that Democrats have motivated people to outrage demonstrated by public protests and rude behavior. But you don’t win elections with protests and rude behavior. Just ask the people who ran the 1972 George McGovern campaign about that. He lost to Richard Nixon in the fourth-worst loss by popular vote in American History.
On Tuesday will Democrats turn out in droves to vote for more ObamaCare, for open borders, and for impeaching Kavanaugh and Trump? Perhaps, but I sure wouldn’t bet on that, particularly when early voting trends, again, show that it is Republicans who are turning out in force.
Or will Republicans turn out in droves because they see the threat to our national security represented by caravans of people we don’t know flooding through our borders. Will they turn out in droves because they are disgusted with the gutter tactics of Democrats desperate to stop the Brett Kavanaugh from becoming a Supreme Court Justice? Will at least some Hispanics and some African Americans realize that it was Donald Trump who cut their taxes and made it possible for them to get a job? If any or all of this happens, that blue wave will look pretty red on Tuesday night.
We will soon know. If I am wrong, then brace yourself for Speaker of the House Nancy Pelosi, Chairman of the House Committee on Finance Maxine Waters and Chairman of the House Intelligence Committee Adam Schiff. That will be very bad indeed, but also very temporary. These clowns are guaranteed to muck things up. But if I am right, we may see the worst public meltdown in the history of TV broadcasting. Either way, be sure to vote and then hold on because this is going to be quite a ride.