> I'm also deeply skeptical of the notion that polling is even remotely correlated to actual results.

But it has been strongly correlated to the results in basically all elections so far in all democracies on the planet. Taking 2016 as an example there has been a very strong correlation between polling and the results. The national polling averages were only 3 points off from the actual result. If that's not correlated I don't know what is.

The shy Tory factor[0] is likely to be even stronger this year than in 2016 when it comes to polls. After 4+ years of being harangued and called every name in the book by the vast majority of national culture (movies, music, TV, news, social media, news-entertainment), along with increasingly hostile projects such as https://donaldtrump.watch/ I imagine less of his more subdued supporters are going to be honest with pollsters.

[0]: https://en.wikipedia.org/wiki/Shy_Tory_factor

Not necessarily.

There is a 0.750.750.75*0.75 or a 31% chance of no rain at all

No it doesn’t, and this is a fundamental misunderstanding of how probabilistic forecasting works. If it rains 9 out of 10 times a weatherman says there is a 30% chance of rain, they are a bad weatherman, but they aren’t much worse than if it rained 0 out of 10 times they predicted a 30% chance of rain. A weatherman accurately assessing the probability of the weather forecast would see it rain around 3 out of 10 days they say there is a 30% chance of rain.

Only if they say that every day, and it rains every day. If the most likely outcome happened every time, then the model is likely wrong/under-confident. The prediction is never going to be 100% accurate until the event is happening/has happened. Up until then, there's always a chance you're wrong or something can change. Being wrong once isn't necessarily a sign that the whole system is messed up.

"He won" is not sufficient evidence upon which to do that.

If they were wrong, but “less wrong” than all others, you should pick their model (unless you have an oracle, because the alternative - flipping a coin or “relying on your intuition” is rarely better).

Also, it doesnt make sense to look at a single prediction to evaluate a model.

Out of all the predictions they have made (did you look at individual state predictions?), how many were correct (and how confident were they?) - how many were wrong (and how close to 50% were they?).

That is how you evaluate a model (aka cross entropy)

It's not "wrong" to predict a low chance for something that eventually happens. Unlikely events can happen.

It seems like you are conflating probabilities with absolute certainties. If I had a 1:10 probability of winning the lottery, I would probably take it. If I had a 1:20 probability of getting injured if I leave my house today, I’d stay home. If I did get out but didn’t get injured doesn’t mean the model was wrong.

If you have a die that rolls a one 1/6 of the time, do you consider the die wrong?

If I tell you you're not likely to get two heads in a row, and you do, does that make me un-reliable?

It's unfortunate we can't just run the election again a few times, and actually find the rate at which Trump is elected given the polls.

And it's not empty signalling if 538 assigned Trump a higher chance of winning; they were pretty much the only ones saying he has a chance. That is why people think the models are useful.

If everyone was wrong, it is reasonable to believe a low probability event occurred. However, given the extent to which people predicted a Trump loss (say 1 in 20), which is significantly rarer, given that the event occurred it suggests the model that predicted a Trump win with the greatest probability to likely be a more accurate model.