Polling should be pretty simple. Basically a mini, trial-run version of an election. Voters interested in participating in the election should be interested in participating in a poll. It should be pretty easy to get a straight fucking answer from enough people that you can get a reasonable enough estimation of how an election is going to turn out… Well that’s how it should be.

The reality is that it necessitates putting together shit like this – which to explain, is a table of recent numbers from the firms whose crosstabs contain the actual numeric quantity of the respondents who said they favored convicted felon former President Trump or Vice President Harris.

That number is NOT easy to come by. It really only comes from three pollsters: The New York Times, YouGov (but not CBS News/YouGov), and douchebag Mark Penn’s “HarrisX” outfit. The table columns are as follows: Poll/Date, number of respondents, the stated percentages for Trump, Harris, and “Else,” the “N” for Trump, Harris, and “Else,” and the N based percentages for each, the code for which reads = ROUND((N CANDIDATE / (N CANDIDATE + N OPPONENT + N ELSE)) * 100, 2).

The last group on the bottom is a complicated story, because the Times’s Nate Cohn called it a “poll within a poll within a poll.” They oversampled the city of Philadelphia – which they didn’t actually publish as its own crosstab page – within the state of Pennsylvania within the United States. Really only included it just to show how close to an actual tie their national poll becomes when you take out the state of Pennsylvania. It’s understandable why they had to recalibrate those numbers.

What’s less clear is why most of the other individual polls the Times/Siena ran look like they just did a little bit of rounding to pump out the presidential race number… Except North Carolina. Seems like they saw Harris was winning North Carolina and thought “Hmmm… Yeah that doesn’t look right. Let’s just put Trump in the lead there. Just feels like he should be up. Yeah, that’ll do.”

Would rather stay out of the weeds of statistical science here. It just looks like they fuck with the numbers and make up whatever they want. Obviously that’s not literally what they’re doing. The real question is do they know what they’re doing? Do they understand that they decisions they make here feed into a narrative when democracy hangs in the precipice? Like this isn’t some jerkoff coding project or a Bitcoin wallet, this is our nation’s life here and the people who take its pulse should maybe do a bit more to explain themselves, forcing the audience to take a lot of faith to the heart.