People fundamentally misunderstand statistics if they think that a 70% chance means something will happen. It means it is expected to happen seven of ten times, but every time it doesn't happen (approximately three of ten times) is also an expected outcome.
FiveThirtyEight understands statistics. They also do a good job of trying to understand the variables. While they didn't expect a Trump win in 2016, they did identify variables that would lead to it: late-breaking swing voters going for the outsider, and dissatisfaction generally hurting the "insider." Those same things are less likely to help Trump this time.
This whole thread--and the Trump supporters on it--make me shake my head, even those friends of mine whom I know personally. I'm stunned that you don't see him as I do, and I'm terrified by what that means--already--for our country. I hope CD is right that it may not make much difference, but I already see a huge difference.
The confidence I see here from Trump supporters fits my narrative of Trump supporters almost too well. You have confidence despite the massive 2018 beating the Trump-led Republicans took, and the polls, because they are similar to the numbers between Clinton and Trump in 2016. But they encapsulate much different dynamics, including the lack of apathy that we saw in 2016, and an opponent who isn't nearly as unpopular as Clinton was (/is). Voters are motivated this year, and yes, that includes Trump voters. We won't see 200,000 fewer votes in Wisconsin than we saw in 2012--even with COVID-19.
Yet you are convinced. To borrow from our dear leader, it is what it is. Nothing this partisan can say will change it--nor your views about Trump's role in Kenosha and Portland, nor anything else. To Bwar's point:
https://theoatmeal.com/comics/believe