Who Won the Debate? (ABC Lost)

Meanwhile, Here's How to Predict Elections

The debate was infuriating to me but not because of the two candidates. It was a debate in which the mainstream media showed its true colors. They fact-checked Trump at least four times but never fact-check QueMala even when she was caught in outright lies, like saying she never was against fracking or how Trump called Neonazis “fine people.” The questions were skewed toward the Democrats and when the Democrat candidate avoided answering a question, there was no push back. So who lost the debate? ABC and the American people.

I suspect that Kamala will get a short-lived bounce in the polls because (a) she didn’t cackle and (b) she spoke fluently, albeit not on point. But that bounce will probably last about as long as a soap bubble in a hurricane. Polls are notoriously inaccurate.

Polls are a type of survey, a recognized but particularly sketchy form of research. Most scientists discount the value of surveys or polls altogether. The only reason we still use them is that they’re easy to do. And that’s why our lazy media class likes polls.

To conduct a proper poll (or survey), you need to “sample” a large group of people. The ideal sample is a representative of the whole group. For instance, to sample the American electorate you need to find a small group of people who represent the entire 165M plus American electorate. If you sample too many registered Democrats or too many urban dwellers or too many college students, you may wind up with a skewed sample that does not represent the whole population. For instance, I could poll 10 people I know and get results that favor Trump. It doesn’t mean it’s an accurate sample. A good poll requires a sound sample and no poll I ever saw discloses its methodology. That’s why political polls rank right up there with phrenology and voodoo in terms of sound science.

Political polls very often use ridiculously small sample sizes—like 100 people. That is hardly a good way to gauge over 165M potential voters. Plus no political poll I have ever seen states what scientists would call the “n” number or number of people sampled. So with a political poll, most of the time, you don’t know how many people were surveyed and where they came from. It would be 12 guys at a White Dudes for Tim Walz rally. In the 2016 election with poll after poll favoring Hillary, it’s clear the polls were oversampling Democrats and, in particular, Hillary supporters.

Plus science has long known that surveys and by extension polls are inherently biased. You can’t survey a person without that person agreeing to be surveyed and it has been determined that the kind of person who takes surveys is fundamentally different than the kind of person who just wants to be left alone. So any poll you do is going to under-sample (or not sample at all) the people who do not want to talk to survey takers.

What kind of person do you think that is? Trump voters have long been insulted and castigated—few MAGA people will talk openly to strangers about their politics for fear of being attacked, called liars, stabbed, or beaten to death. Plus MAGA people often work and raise families, so they don’t have time to be bothered taking some stupid poll. Another thing that makes survey-takers (the scientific lingo calls them “respondents”) different from the general public they are supposed to represent is that those who take surveys tend to have strong opinions. People who are more relaxed or moderate avoid polls. Karens all want to take polls. People who froth at the mouth like to take polls. People who read Ricochet Cafe are more inclined to tell the pollsters to buzz off.

So anyway, you can’t trust polls. So how do you navigate the rest of this election season? There is a solution.

Go to Polymarket or any number of other similar sites. (This is not an ad for them—they don’t know me although I visit their site often. Ricochet Cafe has no relationship to Polymarket.) I don’t know if any of these kinds of sites are markedly better than the others, but I’m just most familiar with Polymarket.

Polymarket is a betting site and, among other things, it allows you to bet on election results. These are not people taking a survey. These are people who are putting their money with their big fat mouth is. You can bet on all sorts of things from election results to sports to the date the Feds will cut the interest rate. And the odds shift constantly for some bets, many times a day. QueMala got a little bump with the debate but that should evaporate by the weekend. (In fact, some of it was lost in the time it took me to write this article.)

Polymarket bills itself as a “prediction market.” You can find out pretty reasonable estimates on how enterprising Americans think about any election. You can even bet of fringe strange things like whether or not Trump will drop out of the race (5% chance). There is a 4% chance Kamala will drop out. You could have bet on whether Trump and Kamala would shake hands before the debate. (They did—and think about this. I think it was the first time they had ever met face to face.)

You can also bet on stupid stuff like how many times Elon Musk will tweet this week or when Taylor Swift will announce her engagement.

I am not an advocate of betting, but I do consult the Polymarket trends often. Polymarket is an offshore company, meaning it operates outside of the United States because the issue of whether you can bet on elections is somewhat controversial. And there are heartening bets, as well. Right now on bets regarding the “balance of power,” the odds favor a Republican sweep (White House, Senate, Congress). And odds also favor that this debate will not be followed by QueMala’s proposed second debate.

Polymarket is big business. Over $125M has been bet on the 2024 election right now, making presidential politics one of America’s favorite sports. Billionaire gadabout Peter Thiel, co-founder of PayPal, has invested in Polymarket, and he was an early investor in Facebook before running away with bags of cash and distancing himself from Mark Zuckerberg.

I don’t recommend betting at all, but it is worth noting that Polymarket uses blockchain technology and so gamblers need to be versed in crypto to place bets. You even need to get a crypto wallet.

Let’s say you want to bet on the Presidential campaign. When I started this article, the odds are evenly split, 49 cents for Trump and 49 cents for Harris. So it will cost you 49 cents to buy one share of “Trump will win.” If Trump wins, you cash out that share for $1. (All shares cash out at $1 each, no matter what the bet is.) If you are insane and want to bet that Joe Biden will win, that will cost you a fraction of one penny per share, but if by some miracle Biden does win, that will pay off $1. Michelle Obama costs 1.8 cents a share, but you’ll get a buck back if she wins.

The old-fashioned handshake rules of “a bet is a bet” do not apply on Polymarket. You can make a bet and then wriggle out of it before the outcome is determined. You do this by selling your shares. Now if you bought shares that Biden will win the next election, good luck offloading those. But if you bet on Harris before the debate and now you want to switch to Trump, you can make the move. You might take a minor loss (or not), but bettors can jump ship freely. However, once the outcome is determined, the window closes and everyone settles up.

Polymarket has been around since 2020, and it’s better than the polls, because it reflects people who actually do things (like follow politics, go online, make money, place wagers) and those are people who vote. They are also people who think about things. And when people put money on their predicted outcomes of politics, that means more than just polling some affluent white liberal urban ladies (AWFULs) at the mall.

By the way, Polymarket did not run a particular bet that was most interesting to me. It was: Will Kamala cackle in the debate? They didn’t take bets on that, likely because one man’s cackle is another man’s giggle.

Polymarket changes constantly and when I started writing this, the odds were 49% for Trump, 49% for Harris. It’s now 50% for Trump, 49% for Harris. By the time you read this, it likely will have shifted again.