The daily prediction published by EL PAÍS keeps the Democratic candidate Joe Biden ahead: he would win the elections on November 3 83% of the time, by 16% of Donald Trump. That is, the Democrat would be president 5 out of 6 times. It’s a clear advantage, one that has been widening since September, when Trump’s options were twice as high as now. Since then, three things have happened: his contagion by covid-19 weakened him among bettors, the polls have not been favorable, and time is running out. Every day that passes without something positive happening to your candidacy is one more day that Biden wins.
But that Biden advantage is not final. 16% probability events, like the one Trump has, can occur, as we can all perceive. I’ve used this simile before: Right now, Trump’s chances of winning are the same as losing playing Russian roulette with a bullet in the drum. There are five options that nothing happens, but we do not think of playing.
To explore those Trump options it is useful to ask a question: what states are they passing through? The country’s electoral system gives enormous importance to the map. Each territory distributes a certain number of electoral votes, which all go to the most voted candidate: the winner takes it all. In total, 538 are distributed and the presidency is won with 270. The following graph shows the polls and the probabilities of victory for each candidate and for each of the territories.
There are a dozen key states because they can change hands. Some cast many votes and others are representative. For example, imagine that they could give you only one result: I think they would ask for Michigan. If Biden wins there, his chances of winning go up to 19 out of 20, based on FiveThirtyEight’s statistical model . But if Trump surprises and takes Michigan, he would become the presidential favorite with 5 out of 6 to win. Something similar happens with Pennsylvania, Wisconsin or Minnesota, and to a lesser extent with Nevada.
The Republican’s Options
Let’s explore the ways of Trump. The first fundamental state is Florida, where you basically need to win. Deals 29 delegates and is open (Biden has a 3-point lead in polls). From there, the Republican must ensure his victory where he is leading: Iowa (+1, in the polls), Texas (+2), Ohio (+2), Alaska (+6). To later also win all or most of the disputed States: Georgia (+0), North Carolina (-3), Arizona (-3). This would leave him 10-15 electoral votes out of the 270 needed to be president.
For Trump to complete that sum I see three alternatives. The first route consists of repeating the 2016 road and winning some Rust Belt States: Pennsylvania, Michigan, Wisconsin or Minnesota. But Biden seems more popular than Hillary Clinton in those territories, so Trump needs to turn the polls by six points or more.
The second arithmetic option for Trump is to surprise in the west and win Nevada, New Mexico or Colorado. It is complicated, of course, because they are States that Clinton won in 2016 and where now Biden also has 6, 11 and even 12 points of advantage.
The third route is a hybrid: it happens because Trump improves – a lot – his polls in the northeast and wins Maine and New Hampshire, but also takes some State of the previous options (rust or west).
None of these three paths is easy, which is why Biden is the favorite – nothing special needs to happen. But we should not take these three routes to Trump’s success as impossible, or others even more exotic. Now they seem crazy to us – and it is possible that they are – but one sign that they are not unthinkable is that, if one does occur, we will surely find a way to make it logical. Because, as the psychologist Amos Tversky wrote: “People predict very little and explain everything.”