Look. This will be my last reply in this debate, but they are separate events. You can't look at the probability of having at least one out of 17 wrong, which is what you're doing, and is indeed the 80-90%, and have that same exact probability apply for 16 out of 17 wrong, when each county has independently had that 8 in 9 chance (or more) of being correct. You could look at the results as a whole per year, and see that 4 out of 9 elections, at least 1 was incorrect. And that drops dramatically to 1 out of 9 elections where 2 were incorrect. That chance will continue to drop further and further until you get all the way down to "at least 16 out of 17 voted incorrect". You can't argue statistics, and then completely ignore the actual stats.

And yes. There are behavioral components, but when I argued that point earlier, saying that these counties likely share a similar demographic spread to the USA as a whole – which is why the likelihood is weighted so heavily toward success — you dismissed it. Your argument was initially based on mere odds and chance, with no reasoning as to why there might be any behavioral change. Just chance.

Look. This will be my last reply in this debate, but they are separate events. You can't look at the probability of having one or two out of 17 wrong, which is what you're doing, and is indeed the 80-90%, and have that same exact probability apply for 16 out of 17 wrong, when each county has independently had that 8 in 9 chance (or more) of being correct. You could look at the results as a whole per year, and see that 4 out of 9 elections, at least 1 was incorrect. And that drops dramatically to 1 out of 9 elections where 2 were incorrect. That chance will continue to drop further and further until you get all the way down to "at least 16 out of 17 voted incorrect". You can't argue statistics, and then completely ignore the actual stats.

And yes. There are behavioral components, but when I argued that point earlier, saying that these counties likely share a similar demographic spread to the USA as a whole – which is why the likelihood is weighted so heavily toward success — you dismissed it. Your argument was initially based on mere odds and chance, with no reasoning as to why there might be any behavioral change. Just chance.