Yes, that's definitely right. I think the path for Trump is more like: - Start at 50 vs 40, gap of 10 - Flip 2.5% to 47.5 vs 42.5 gets the gap down to 5 - 3% advantage in Dunno/Other and it'… (View Post)

One thing that could be worth looking at that I was about to look at but then my browser crashed and I can no longer be arsed is, how does the lead vary depending on the other/don't-know proport… (View Post)

It's here: https://www.realclearpolitics.com/epolls/2016/president/us/general_election_trump_vs_clinton-5491.html So July 1st is Clinton 44.6 Trump 39.8. Graph for this time for comparison: http… (View Post)

My reply to Stodge also applies to this - in any poll you can always find a specific subsample that seems wrong, it doesn't mean polls don't work. That said, you're comparing to the 20… (View Post)

Is this poll (or the Marist one) *supposed* to be demographically balanced by region? Generally with any poll you can look at the various subsamples and find something in it that will allow you to sa… (View Post)