Tomorrow is election day. Let that anxiety wash over you however you want to take it.

It’s well established I’m a data nerd, so I wanted to take a final look at some of the polling, probabilities and such. I have a few go-to’s: 538 and now the Nate Silver offshoot, gambling sites and Real Clear Polling. So, what do they all say? Insert shrug emoji here.

Real Clear Polling

I’ll start with Real Clear Polling. Their no-toss up state projection has Trump 287, Harris 251. But of course, this exercise can not be that simple. According to their state polling averages, there are four states with an average of less than one point of difference. Two others are less than two and a seventh is less than three, the generally accepted margin of error for a polling aggregate.

This exercise does not factor in probability math as far as a winner forecast.

538/Silver

Over at 538, their forecast does include some probability math. They have it basically as a coin flip, Trump winning 52 percent of the time and Harris winning 48 percent of the time. So, nothing decisive either way. The electoral college prediction: Trump 272, Harris 266. Yeah, that close.

Looking at their state averages compared to RCP, they are mostly in line. They have the same four all within a point or less: Nevada, Pennsylvania, Wisconsin and Michigan. The next three are again North Carolina, Georgia and Arizona, all leaning Trump.

As for the national vote averages, 538 had Trump at 48.0 and Trump at 46.8 percent. With the way the electoral college is set up, it is tough for a democrat to overcome “only” winning by 1.2 percent. That scale is tilted.

Nate Silver is no longer at 538, but his methodology and presentation is still very similar. He has Harris at 48.5 percent nationally, 47.7 percent for Trump, so slightly closer than 538 at a .8 difference, as negligible a difference as there is. State wise, he also has two states less than one percent: Pennsylvania and Nevada. He has Harris up slightly more in Michigan and Wisconsin than the other two do, but again barely.

One of my favorite visuals 538 does is the electoral snake, or, the path to 270.

This is a great visual representation of the path to victory. This is sorted by state polling averages, those likeliest for Harris on the far left, for Trump on the right and then those contested states in the middle. As you can see, Harris will likely need both Nevada and Pennsylvania to win. She could possibly win North Carolina or Georgia, but if she is losing Nevada, she is less likely to win those two. This is where Harris might be in a bit of trouble. The four states that are polling as coin flips (Michigan, Wisconsin, Nevada, and Pennsylvania)? She might need that coin to land on her all four times.

Betting Sites

You can’t bet on the Presidential race here in the US, but you can oversees. RCP tracks a few and spits out an average. These are based off no-vig odds with implied win probability. If that doesn’t make any sense to you, then your probably skipped over the betting site section anyway.

I’ve spent most of the last eight years hearing from Trump supporters that you can’t believe polling data because he was not projected to win in 2016 and did. I’d guess those people only bet favorites in sports as well, since they don’t understand probability math. The betting average has Trump at 58 percent, with a high of 59 and a low of 57.

Now that the polling averages are showing this as a contested race this time around, I’ve heard from more Harris supporters that you can’t believe polling. So, apparently both sides know better than data scientists.

I’ve read the herding philosophies, and I can see some credibility in that theory–that pollsters are afraid to release polls that go against the norm so the outliers are not there. But I’ve also heard stuff like you can’t trust Nate Silver because he advises Polymarket, which a Trumpican invested in . That ignores that Silver’s numbers are actually slightly better for Harris than the others included in this here little exercise, and that Polymarket is right in line with the other betting markets.

So, What To Expect?

What does all this mean? Well, the pollsters are likely to not be right. They are also likely to not be wrong. That is how coin flips work. If the national average is off by more than 3-4 points, or multiple states are off by more than 5-8 points, then we should examine this process. But within those margins, this race could still go wildly in either direction.

Pennsylvania is the likeliest tipping point state, at just over 25 percent that it alone is the state that puts the winner across the finish line, according to 538. Cool, and east coast state, so we could know early, right? Not likely. Pennsylvania looks to be razor thin, and was one of the last states done counting last time. I would expect the same here.

I don’t expect to go to bed Tuesday knowing who the winner is. I’m not even sure about Wednesday. We are likely looking at another Red Mirage, where Trump looks to be ahead early in the night. The counties that tend to be done counting early tend to vote Team Red. They are less populated, vote mostly in person and don’t take as long to calculate. I have a feeling at some point once again, Trump will try to claim victory early.

I’ll say it again, this thing could go either way. Leaks have already come out that Trump plans to claim fraud in Pennsylvania if that vote does not go his way. Those seeds are already planted.

And I’m writing all this not even knowing if I will watch the election coverage or not. In 2020, I did not. In the middle of a pandemic, I sat outside in a buddy’s driveway and ran the rail on Bourbon County Stout, drinking one of each release from 2012 to 2020, plus a couple more variants. This year, I’m having a surgery the day after, so I’ll be out of it in a different way.