All times in this article are UK time.

The first two hours, from 11pm to 1am, will be very difficult to interpret. Because of early voting, and Democrat precincts often reporting before Republican ones, you’re likely to see Clinton take the lead in a few Trump must-win states. That doesn’t necessarily mean she’s going to win them, far from it. Trump is likely to come back in those states, so don’t switch off your television and automatically assume that you know the result!

There are more Trump must-win States, particularly early on, than Clinton ones. That’s just the nature of the Electoral College. Obama carried Florida and Ohio in 2012; he won. It’s very difficult to think of a plausible path (there are some implausible ones with very similar states voting in wildly different directions to each other) for Trump to the White House if he doesn’t win both of them.

Turnout will be pretty critical as well. It seems to be up as far as early voting is concerned, but what does that mean? Does it mean that the Democrats have just done a good job of getting their voters to the polls early, in which case fewer people will vote on the day and turnout will remain fairly constant? Or is it a genuine increase in turnout over 2012? Trump certainly needs to get previous non-voters who support him to turn out on the day if he’s going to win the White House. Clinton seems to be getting a huge increase in Hispanic voters going her way; on the flip side, can she replicate Obama’s success with African American voters? That’s one to watch in South Carolina especially.

Watch out also for how Gary Johnson and Jill Stein are doing early on: Stein will almost certainly take protest-votes from Democrats, whilst Johnson will take a mixture. One of the real uncertainties of this election: in 2012, the Democrats and Republicans were polling 96.9% of the vote on average between them, and they got 98.3%. This time, Trump and Clinton combined are averaging just 90.4%. What’s happening with the other 10%? Are they gravitating back to the major candidates or not? If third-party candidates are holding up in some States, are they doing less well in the swing States where it can affect the outcome? You might be able to get a handle on this quite early.

Some States close their polls at different times. This is often due to time zone differences, where a State straddles two time zones. You’ll be getting results from one half of a State whilst the other half is still voting in some cases.


Unless you’re a complete political anorak, don’t pay too much attention to what happens between 11pm and midnight  The only polls to close are those in the Eastern half of Indiana [Indiana straddles time zones, so the Western half polls close later] and some of Kentucky, and we’re pretty sure that Indiana and Kentucky will vote for Trump. Expect a double-digit win for Trump in Indiana. If he’s winning in single-digits, it’s some cause for concern. If he’s up by 15% or more, then it’s good news for him.

There’s a knife-edge Senate race in Indiana, and we probably won’t even learn too much about that. Your best bet for 11pm to midnight is to take a quick nap. You won’t have missed anything, but you’ll have the energy to stay up later for the more interesting stuff.

Over the midnight to 1am hour, you’ll get real indications of what’s going to happen. Information will come in from Trump must-win States – Florida, North Carolina, Ohio. It’ll come in from the Clinton must-win State of Virginia, and from the truly swing State of New Hampshire.


Polls close in more states. The early results in these won’t be indicative – two will lean more pro-Clinton from early results, and two will lean more pro-Trump. It’s important to keep an eye on this or you’ll see things that aren’t there – on either side:

Florida (a Trump must-win State), New Hampshire (swing state) (first polls): Expect the early results to be more pro-Clinton than the states as a whole, because the Republican areas are usually slow to report. Treat with the appropriate pinch of salt. In Florida not all polls close at midnight, so you’re waiting for 1am for the rest.

Virginia (a Clinton must-win State), Georgia (a Trump must-win State): Expect the early results to be more pro-Trump than the states as a whole, because the Democrat areas are usually slow to report. Treat with the appropriate pinch of salt.

South Carolina (safe-ish Trump State): Unless Trump is losing by a landslide, he’ll win South Carolina. I’d say ‘par’ for Trump is something around a 7-9% lead in South Carolina if the election as a whole is on a knife-edge. More than that, he’s doing better than expected. Less, and it could be a difficult night for him.

Vermont: Will go Clinton by 20+ points most likely. You can usefully ignore Vermont going Clinton just as easily as you can ignore when Kentucky goes Trump by 15(ish) points. Those numbers could be way out, but it’s not worth reading much into it.


Ohio (a Trump must-win State): Again, don’t read too much into the early results in Ohio. This is something of a Trump sandwich: early results will favour Clinton, and if 2012 is anything to go by, so will the late results. The ones in the middle? That’s where Trump would have to cash in if he’s going to win Ohio. Polls suggest that he will, but it’s fairly close.

North Carolina (a Trump must-win State): As so often happens, the early votes are counted first. Clinton has a better ground game than Trump for picking up early votes, so she’s likely to start off in the lead. That doesn’t mean she’s winning the State as a whole.

West Virginia: Romney won this by 26 points in 2012; expect a Trump landslide and move on. Nothing to see here.


All hell breaks loose. You’ll see television networks falling over themselves trying to report data as we get results in from (in alphabetical order so I don’t miss any): Alabama, Connecticut, Delaware, [Washington] DC, Illinois, Kansas, Maine, Maryland, Massachusetts, Michigan (first polls) Mississippi, Missouri, New Hampshire (final polls), New Jersey, Oklahoma, Pennsylvania, Rhode Island, South Dakota, Tennessee, Texas (first polls). The final polls will close in the Western parts of Florida at 1am as well.

First, the minor things to keep an eye on. States like Illinois and Missouri matter only in terms of the Senate, and Texas is only highly competitive if Clinton’s winning by a landslide. Maine isn’t a winner-takes-all State. It’s just worth keeping an eye on Maine’s second Congressional District as I have a feeling that Trump might just sneak it, and pick up an extra one Electoral College vote. Not too important, but there are some highly plausible tied maps so if it’s really close that one vote could make a difference.

Look out for Michigan: it’s normally Democrat, but quite demographically favourable to  Trump and it’s notable that both Clinton and Trump have been campaigning there in recent days. Pennsylvania is possibly similar, but I think Michigan is a more realistic chance. That would cover for Trump losing Nevada later in the evening if he were to sneak that. Why Michigan? Well, Detroit is overwhelmingly Democrat. But the rest of the State is more mixed; it’s Detroit that keeps it in the Democrat column. That means Clinton needs to get out the African American vote in Detroit, if Trump surges amongst white working-class voters – and there are many of them in Michigan. Perhaps I’m talking myself into it, but I could easily see Trump getting a breakthrough in Michigan. Detroit tends to report later though, so Trump could lead all the way and then lose Michigan.

Other than this, it’s probably still Florida, New Hampshire and the ongoing counts in States where polls closed earlier that are worth looking out for.


Arkansas: Republican now. Don’t let the Clinton name fool you.


Polls close in Arizona, Colorado (first polls), Louisiana, Minnesota, New Mexico, Nebraska, Wisconsin, Wyoming, Michigan (final), Kansas, South Dakota, Texas (final polls). Another huge influx in States coming in, so I’ll just cherry-pick the important ones.

Arizona: This was Republican in 2012, but it’s getting tighter. If Clinton gains any new States that Obama didn’t, Arizona is top of the list as most likely. It’s also going against the norm in that the early results will probably be biased towards Trump.

Minnesota: A long-shot, but not totally impossible for Trump to win. If he did, he’d probably be on course for a big win nationally. It’s unlikely but just about within the realms of possibility.

Colorado, New Mexico: A probable Clinton win in both of these, but it’s going to be close. Win one of these, and Trump may not need (say) New Hampshire or Nevada. New Mexico should be Gary Johnson’s biggest result. This could impact on the final result, especially as he appears to be taking more Clinton votes than Trump ones in New Mexico. The same probably isn’t true across America.

Wisconsin: Clinton should win this, but this is one where the earlier reporting favours Republicans. If Trump is leading here early on, don’t assume he’ll go on to win the State. It’s another just-barely-competitive State, so there’s an outside chance of a Trump win but I wouldn’t bet on it.


Idaho (first polls), Iowa, Montana, Nevada, North Dakota (first polls), Oregon (first polls), Utah

Nevada: The key question is whether Trump is already dead in the water in Nevada. Polls have him only just behind on average, but early voting in Nevada is huge – and many, many more Democrats have voted than Republicans. I’m expecting a Clinton win here, despite the polls being so close.

Utah: The only realistic chance of someone other than Clinton or Trump winning any Electoral College votes. I think Evan McMullin will get 20-30% of the vote, but come up short. It’s not impossible that he could win, though, which would lead to a whole range of possibilities for a tie in the Electoral College. If no candidate gets 270, then things get very interesting indeed. (It’s not impossible to conceive under the arcane tie-breaking rules that he could win the Presidency, in fact – though it’s a one-in-a-thousand chance at this stage)


Polls close in California, Hawaii, Idaho (final polls), North Dakota (final polls), Oregon, Washington.

It’s an absolute long shot that Trump could gain in Oregon, but highly unlikely that anything else will change hands. Idaho and North Dakota will stay Republican; California, Washington and Hawaii will remain Democrat. There’s a lot of Electoral College votes in California (55) as it’s the most populous State in the Union so expect Clinton’s total to suddenly shoot up once we pass 4am. Clinton could well be behind in the Electoral College at 3.55am but very much certain to win, because of California et al.


Polls close in Alaska. Almost certainly Republican, though not quite a 100% nailed-on certainty this year. The Republicans don’t seem too keen on Trump in Alaska. Don’t wait up for that result; it often takes them days to finish counting. It’s only 4 Electoral College votes anyway. If Alaska’s Electoral College votes are critical, there’s probably a long recount happening somewhere else anyway…

Photo by Diego Cambiaso