Here are five critical things to watch for on Election Night

Nervous as hell? Don't think you can take election night? Girding yourself for weeks of counting torture?


There is unfortunately nothing that can completely alleviate that stress. But as the night of unfolds, there will be a few things to keep an eye on that could provide some real and significant early clues as to whether events are trending in the right direction, or if you should be stocking up on a two-week supply of bourbon (or a four-year supply of Prozac).

Before diving into them, it is helpful to note a few important assumptions up front about the dynamics of the vote.

One is that record-breaking early voting has favored Democrats, and the Election Day vote will favor Republicans. This is fairly well documented: as of a week before Election Day, Democrats were ahead by about a 2-1 margin in the states that track returned mail-ballots by party registration. On the other side of the coin, polling has consistently shown Republicans preferring to vote in-person on Election Day. The upshot is that in general, mailed and early ballots will skew Democratic, and votes cast on Election Day will skew Republican.

A second, related presumption is that there will be at least some "Blue Shift" after Election Day in states that are still counting. Ballots counted after Election Day tend to skew Democratic, both because Democrats' ballots are more likely to be challenged and end up as provisional ballots, and also in cases where Democrats are mailing in more votes that get counted late (this tends to happen especially with younger voters). So Election Night counts – i.e., before all the votes are counted – can often look a lot redder if there are still mailed votes outstanding, and effect that has been termed "Red Mirage."

And finally, it is likely that any close result in a swing state will enter the Twilight Zone. The Supreme Court has allowed late-arriving mail ballots to be counted in a number of critical states, but has opened the door to re-considering. Signature mismatches, "naked ballots" that aren't enclosed in a security envelope, and other ballot issues will come under close scrutiny and almost certain litigation – in fact, more than one million people could ultimately lose their vote due to these issues. The bottom line is that unless a state's reported winning vote margin is greater than the number of ballots that could end up being contested, the state will carry a big asterisk.

With that general view of the landscape, here are five things to pay attention to that could help cut through the maze of information and speculation and give some meaningful early indications about the outcome.

1) Trump's Mission-Critical States

As reported in Axios – and according to multiple sources –the number of likely state wins for Trump has dwindled in recent weeks, but Trump campaign manager Bill Stepien has plotted three remaining paths to victory. The first is for Trump to win Arizona, North Carolina, and Pennsylvania. The second one swaps out Pennsylvania for Michigan. The third requires Trump winning the combo of North Carolina, Michigan and Nevada.

Notice the common points in those scenarios: all three require North Carolina, two depend on Arizona, and two need Michigan. So those are the mission-critical states, and Biden winning any of them would slam a door in Trump's face. Forecasting models agree: for example, FiveThirtyEight's interactive forecasting tool shows that a Biden win in North Carolina raises his win probability from 90m percent to 99 percent, in Arizona to 98 percent and in Michigan to 92 percent.

Since Arizona and North Carolina start counting mail ballots either upon receipt or otherwise before Election Day, there is a real possibility of getting results relatively quickly—either this evening or tomorrow. So if they are looking strong for Biden, it would be a meaningful sign.

One important caveat though. Arizona officials plan to release initial results at around 8:00 pm local time based on the ballots cast before Election Day. These votes are therefore likely to look better for Democrats. They then plan to count the ballots cast in-person on Election Day (skewing Republican), followed by votes received by mail that day (back to Democrat). So before popping a champagne cork or driving your fingernails into your palm with each fresh tranche of numbers, pay close attention to the time and estimated percent of votes tallied: Arizona's figures could gyrate throughout the evening. North Carolina's probably won't have the same flip-flop pattern, but could start out by leaning Democratic and then migrate steadily toward Republicans because state officials intend to start with ballots cast pre-election and then count Election Day votes last.

2) The "Red Zone" States

In football, the last 20 yards before the goal is called "the red zone." There is a red zone for each presidential campaign too. The three scenarios above first require Trump to hold closely contested states Ohio, Florida, Georgia, and Iowa that Stepien explicitly cited as foundational to the paths laid out above (Texas apparently went without saying).

Losing any of them would be a powerful sign that Trump's drive will fail – so powerful that the FiveThirtyEight forecasting tool moves the Biden win probability above 99 percent if he takes any of those states. Earlier absentee vote counting in Florida and Ohio means that there is a decent chance of having significant information by Wednesday morning. And again, based on the order of counting, be wary of initial tallies in Florida, Texas, Ohio, and Iowa which may first skew Democratic and then trend steadily in a Republican direction.

And of course, Biden's got a red zone too. His pathways to victory assume solid win probabilities in swing states like Minnesota, Nevada, and Michigan. A reversal in any of them would be significant. Take Minnesota, for example (which counts mailed votes upon receipt), currently forecast as 94 percent likely to go to Biden by FiveThirtyEight. But if he does lose it – which we expect to know no later than noon on Wednesday (Minnesota combines early and Election Day counts so there should be fewer wild swings along the way) – his overall chance of winning the election drops to just 29 percent. Nevada similarly counts absentee early (a loss drops Biden's win probability to 37 percent), and while Michigan election officials expect a slower count, a loss there could be devasting (11 percent).

A final, critical note about those forecast probabilities: they bake in the idea that there is some correlation among state results. So a win in Ohio is given credit not just for the Electoral Votes it delivers, but also for what it augurs elsewhere. But of course, other states may not break the same way.

3) Winning Before the Blue Shift Sets In

The two indicators above are focused on swing states where we could get reasonably firm results earliest. But another way to think about it is in the other direction: what if Biden is already doing well in the slow mail-counting swing states? If those states are already looking good by the wee hours of November 4 with most in-person votes tallied, that would be a strong indicator that Biden is in a winning position.

The two states to pay attention to here are Pennsylvania and Wisconsin. In Pennsylvania, not only do election officials wait until 7:00 am on Election Day to even start opening mailed ballots, but some key counties wait until every other kind of vote is tallied and then start counting the mail-in votes last (Republican-run Cumberland County plans to wait until the next morning). Wisconsin similarly waits to open mailed ballots and then conducts precinct counts which could result in slow results.

And again, there is a flip side: if states like Arizona, Minnesota, Nevada, and Georgia that plan to tally absentee votes quicker are looking good for Trump as the sun comes up on November 4, it is a lot less likely that there will be a well of outstanding Democratic votes to turn things around.

4) Exit Poll Turnout and Demographics

Data so far indicate that more than half of this year's total vote will come in before Election Day (Texas, for example, had already surpassed its total 2016 vote). The more this Democratic-leaning vote is in before Election Day, the more Donald Trump needs the swing state electorate to be very Republican on Election Day itself in order for him to close the gap.

Figuring out if the Election Day vote is coming through for Trump is something that can potentially be estimated via early exit poll data. There is a fraught history around the accuracy of exit polling, which has often been a mess. But there have been steady improvements, and the two competing approaches to exit polling this year should give a better range of plausible numbers. If media release Election Day-specific turnout demographics, which they could start to do as early as 5:45 pm, those data could signal whether Trump got the highly Republican electorate he needed to close the gap.

Take Georgia for example. Several days ago, 30.4 percent of mail-in votes had come from Black voters, 14.3 percent from voters under age 35, and 56.5 percent from female voters. We know that these groups favor Democrats (in the latest Monmouth poll of the state, Black voters favored Biden over Trump 90-7, young voters 54 -40, and women 53-44). Given Republican voting preferences, one would expect the proportion of voters from those Democratic-leaning demographics to dip significantly in the Election Day electorate. If it doesn't, that is good news for Biden.

5) Bellwether Counties

There is a long history of looking at "bellwether" counties to try to determine how the election will go, but it's a mixed one, and evolving populations are continuing to scramble the picture. Still, there are a handful that are worth paying attention to.

One is Arizona's Maricopa County. It is significant because it is close and it is large. Home to about 60 percent of the voting public, the result there has driven the winning side in the last 9 Arizona elections. Ever since Donald Trump won it by almost exactly the 3.5 point margin that he won the state in 2016, it has been trending toward Democrats. A Biden win there would be big.

In another Trump state linchpin, North Carolina's Jackson County could be an indicator. It is a true swing county that is one of only four of the state's 100 counties to vote with the winner in at least 10 of the last 12 elections. It is on the small side as North Carolina counties go, so it will not drive the outcome, but it could provide a useful litmus test.

In Florida, it is actually worth keeping an eye on two strong Republican counties – Sumter and Pinellas – where Trump needs giant margins in order to hold the state, but where there have been hopeful signs for Democrats so far in early voting. Anything less than a 2-1 margin for Trump in those places would be a flashing red bulb over Trump's chances in Florida and, by extension, the entire election.

And in the rust belt, there are a few places to watch. Kenosha County in Wisconsin is a true bellwether having voted with the winner of the last 7 partisan statewide elections. And in Pennsylvania, Erie County is considered a useful barometer, having split the statewide races (while voting for Trump) in 2016 and preferring Democrats in 2018 en route to being the only county to vote with the winner in all seven of the last partisan elections. However, county officials plan to count in-person votes starting at 8:00 pm, followed by absentee starting around 11:00 pm, then stop counting at 2:00 am before resuming in the morning of November 4. So bear that pattern in mind when watching returns there, since there is likely to be Blue Shift into the early morning.