On November 8, 2016, Donald Trump won the presidency despite garnering millions of votes fewer than Hillary Clinton. Since that time, few American institutions have been subject to as much controversy as the Electoral College, the body through which US presidents are elected every four years. Many Democrats have spent the last four years complaining that the Electoral College disenfranchises urban voters and bestows an unfair advantage upon Republicans in presidential races. In response, Republicans have insisted that such outcomes are features, not bugs, of our republican system of government. (As an irresistible editorial aside, these arguments would undoubtedly be reversed if the current Electoral College landscape favored Democrats over Republicans.) In any event, four years later the Electoral College maintains its position at the forefront of presidential politics, with no indication that the process of electing presidents will change in the foreseeable future. Accordingly, as the 2020 election looms a mere matter of days ahead, any analysis of the presidential candidates’ prospects must start and end with one magic figure in mind – a figure alluded to in the title of this piece.

The Electoral College consists of 538 electors, divided among the US states and DC on the rough (but not precise) basis of population. In order to secure the presidency, a candidate must win a majority of all electoral votes – which means that the winning candidate is the one who manages to cobble together, as Donald Trump’s and Joe Biden’s advisors have been patiently explaining to them, the votes of 270 electors. While it is technically possible that no candidate will obtain 270 electoral votes, this would require either that a third-party candidate wins electoral votes or that Trump and Biden deadlock at 269 – an exceedingly unlikely, though not impossible, scenario. In such a case, the presidency would be determined by the House of Representatives and the vice presidency by the Senate.

The purpose of this article is to analyze the current state of the Electoral College playing field heading into Election Day, assessing the candidates’ respective positions in the key states that are likely to determine the outcome of the election, and considering each candidate’s most straightforward path to 270 electoral votes and his likelihood of achieving it. In conducting this analysis, I am all too cognizant of the potential for my personal biases to creep into my predictions. As anyone who regularly speaks to Republicans and Democrats about politics can no doubt attest, for all their differences, one thing that Biden and Trump supporters share in common is that neither can seem to envision any scenario in which their favored candidate is likely to lose. If I similarly allowed my own hopes to color my prognostications herein, I would undoubtedly be very embarrassed when Libertarian Party candidate Jo Jorgensen fails to win all 538 electoral votes next week. So with that in mind, let us all endeavor to set our personal biases aside, acknowledge that there is much about this election we do not know, and nonetheless seek to draw the best conclusions possible on the basis of what we do know.

The first step in Electoral College math is identifying each candidate’s “safe” states – the states in which the opposing candidate has little to no chance of victory. Safe states constitute the base on which each candidate will need to build by winning enough swing states to cross the threshold of 270. For Donald Trump, these safe states would be Indiana (11 electors), Tennessee (11), Missouri (10), Alabama (9), South Carolina (9), Kentucky (8), Louisiana (8), Oklahoma (7), Arkansas (6), Kansas (6), Mississippi (6), Utah (6), West Virginia (5), Idaho (4),Nebraska (4), Alaska (3), Montana (3), North Dakota (3), South Dakota (3), and Wyoming (3) – for a total safe-state base of 125.

Joe Biden’s safe states are fewer in number but richer in electoral votes. They are California (55), New York (29), Illinois (20), New Jersey (14), Washington (12), Massachusetts (11), Maryland (10), Connecticut (7), Oregon (7), New Mexico (5), Hawaii (4), Rhode Island (4), DC (3), Delaware (3), Maine (3), and Vermont (3) – for a total of 190.

For the sake of accuracy, it is worth noting that Nebraska actually has 5 electoral votes and Maine has 4; however, these two states apportion some of their electors based on who wins each congressional district. Of Maine’s 4 electoral votes, 3 are safely Biden’s – while the 2nd Congressional District could go either way. Of Nebraska’s 5 electoral votes, 4 are safely Trump’s – while, as in Maine, the 2nd District could go either way. Accordingly, I have assigned 4 Nebraska electors to Trump and 3 Maine electors to Biden.

Also worth noting is that there are several states which, although not entirely safe, are generally expected to be won by a certain candidate – call them the “likely” states if you will. For example, Texas polling has been competitive and Republicans are concerned about a surge in early voting there; however, the smart money is still on Trump winning the Lone Star State and its 38 electoral votes. It’s a similar story in Georgia (16 electors), where polling has been tight but conventional wisdom holds that Trump will likely win – or, that if he loses Georgia, he will have been beaten so badly nationally that the state won’t matter anyway. On the Biden side of the equation, Colorado (9), Minnesota (10) and Virginia (13) are states that Trump has attempted to compete in but are unlikely to tip his way even if he wins the election. For the sake of confining our limited remaining space to true swing states only, we will add these states’ electors to their respective candidates’ tallies. Thus our updated score is Biden 222, Trump 179.

With all of that said, let us proceed to the key states in this race. Florida and North Carolina, a pair of southeastern states with large populations and correspondingly rich deposits of electoral votes, have been hotly contested all campaign season long, with both candidates pouring time, attention and resources into them. The RealClearPolitics (RCP) polling average has Biden ahead by just 1.5% in Florida and 1.8% in North Carolina as of the time of this writing. It is easy to envision either candidate winning each of these states – and it is probable that both will be won by the same candidate. The polling suggests that Biden maintains a slight edge in Florida and NC, but by margins so slim that they could be dismissed as ordinary polling errors. The stakes here are very high, because Florida sports 29 electoral votes and North Carolina 15.

Arizona, a traditionally Republican state featuring a respectable haul of 11 electors, is in a similar boat. Biden leads by 2.4% in the RCP polling average, but a Trump win here would be far from an upset. Indeed, Arizona, like Florida and North Carolina, is virtually a must-win for Trump – so the close margins in these three key states make the race a whole lot more exciting.

Now we turn our attention to the Upper Midwest. Trump won both Wisconsin (10) and Michigan (16) by narrow margins in 2016, and both are now must-wins for Biden. The good news for the former vice president is that polls show him with strong leads in both states. If either of these states goes red next week, however, Donald Trump will more likely than not have earned himself a second term.

Ohio and Iowa are a different breed of swing state. Both states voted for Barack Obama in 2008 and 2012, and both were won by Donald Trump in 2016 by convincing margins. Iowa and Ohio polls show competitive races now, with Trump leading by 0.6% in Ohio and Biden leading by 0.8% in Iowa, according to RCP. Ohio is a big prize with its 18 electoral votes, while Iowa weighs in at just 6. I would wager that both states will be won by the same candidate, probably Trump – but if Biden ekes out a victory in either or both, the president is in deep trouble.

New Hampshire and Nevada both lean Biden, but for different reasons. Nevada is home to a large Hispanic population, which helped deliver the state’s 6 electors to Hillary Clinton in 2016. However, Trump has made inroads with Hispanic voters in the last four years, and the state is not out of reach for him this year. New Hampshire’s population, on the other hand, is almost entirely caucasian – and although Trump ran close to Clinton here in 2016, his deteriorating standing among white voters deems it highly unlikely that he will win New Hampshire and its four electoral votes this time around.

Finally, we arrive at what I see as the ultimate swing state (technically, swing commonwealth) in this election, the one most likely to determine the outcome of the election: Pennsylvania. It is virtually impossible to paint a realistic path to 270 for Trump without Pennsylvania and its 20 electors, and the math likewise gets rather dicey for Biden if Trump manages to eke out another victory in the commonwealth. Unfortunately for Donald Trump, he trails his Democratic opponent here by 5.1% according to the RCP average as of the time of this writing. Even Republican polling has Biden ahead here, which does not bode well for the president’s reelection prospects.

For Trump, the most likely path to 270 electoral votes looks something like the following: He wins all his safe states as well as Texas and Georgia, and he also wins Arizona, Ohio, Iowa, Florida, North Carolina, and Pennsylvania. Doing so would place him in the high-270s and earn him the upset over the Democrats and pollsters once again.

For Biden, the math looks to be considerably easier. His most likely path to 270 looks something like the following: He wins all his safe states plus Virginia, Minnesota and Colorado, and he also wins Michigan, Wisconsin, New Hampshire, Nevada, and Pennsylvania. This, which the polling suggests is the most likely outcome, already places him in the high-270s – with the ability to pick up any slack in toss-up states like Arizona, North Carolina and Florida.

So where does all this leave us? Clearly, Biden is the frontrunner. He could blaze a path to the presidency through either the South or the Midwest, while Trump needs a good deal of both in order to win. Moreover, Biden is ahead in a number of key states which Trump must win in order to win the election. That said, there certainly remains enough wiggle room for Trump to find a path to victory and earn himself another four years living rent-free in the White House and in Democrats’ heads – but a significant, nationally-persistent polling error would likely be needed for this to happen.

Speaking of polling errors, I imagine that some readers may be thinking about the “fake polls” (does Donald Trump read TLS?) and dismissing my analysis based on the premise that pollsters got it wrong in 2016 and therefore the polls are meaningless. Addressing this erroneous viewpoint really merits an entire article of its own, but I will briefly mention just a few of the counterpoints to this argument.

First of all, we must acknowledge that a polling error of such magnitude is in fact possible, but not probable. As a general rule, presidential election polling has been remarkably accurate throughout the years, normally predicting the result within 3 percentage points at the national level and 4 points at the state level. So while the polls did get it wrong in 2016 by a larger-than-usual margin, recency bias may be coloring our perception of polling and the accuracy thereof.

Secondly, pollsters have identified certain errors in their polling methodologies from 2016, such as under-polling the “poorly educated” (whom Donald Trump loves – and who, it turns out, love Donald Trump). Pollsters say they have learned from these errors and sought to fix them in subsequent polling – and the accuracy of polls in the 2018 midterms would suggest that methodologies did in fact improve following the 2016 debacle.

Third, Biden’s edge in state (and national) polls is more significant than was Hillary’s at this point in 2016.

Fourth, and perhaps most importantly, the fact that polls got it wrong in 2016 does not necessarily mean that they are wrong now. While voters are justified in suspecting that the polls may be wrong, and a substantial dose of skepticism is always warranted, that does not mean that the 2020 polls are meaningless. As noted at the start of this article, there is much that we do not know about this race. But polling is one of the best indicators – if not the best indicator – at our disposal to predict how the election will go and which states are most critical to each candidate’s chances. Much like stocks might perform poorly over a short period of time but have always acquitted themselves well over the long term, so too polling has shown itself over a matter of generations to be generally reliable, although of course not infallible.

Even though the polls are probably accurate to an approximate degree and Biden will probably win, it never hurts to be prepared – so I’m going to leave off here and go plan my Jo Jorgensen victory party.