and to widen back in favor of Clinton.
Yesterday, Five ThirtyEight had been at Clinton 53.3 to Trump 46.7. Since then it has bumped twice, and as I write this, is at 53.7 to 46.3, or a widening of the margin by 8/10 of a percent.
Similarly, Huffington Post Pollster had been a 1 point Clinton lead nationally, but as of a bit more than an hour ago is now back up to 2, 45-43.
Also similarly, The Upshot blog had had its percentages as 68% CLinton to 32% Trump. Yesterday it moved to 68-31 and is still there. There are some changes within their state predictions, with NC now leaning towards Trump, but the margins for Clinton in FL and IA (and I think OH) widening again.
Looking at the 7 prediction models, on a national level (which you can see if you scroll down on Upshot, and can click through for details), only RothenburgGonzalez shows it as a tossup, with all of the other 6 showing an advantage to Clinton ranging up to 80% likely at Princeton Election Consortium (which only uses state polls).
We still need the closing night of the convention, and then a chance for post convention polling starting at least one week after the end of our convention. to see where we really are, but if overall you are seeing a solid lead nationally for Clinton, and 60% likelihood or better in sufficient states to win the electoral college, absent something totally unexpected, the results are exceedingly unlikely, based on past precedent.
UPDATE as of 6:41 Upshot now up to 70-30, and Princeton Election Consortium is up to 83%. And that is before the impact of tonight.