Clinton wins the debate but not the war

American presidential debates have very little to do with policy and debate – they are about who wins the narrative – be it by body language, demeanour, or a single hard-hitting ‘zinger’ aimed at the opposition’s fatal flaw.

Hillary Clinton – by means of a calm demeanour and a string of solid attacks on Trump – clearly won the narrative of Monday evening’s debate. BrandsEye’s social media analytics shows– in real time during the debate – a decisive victory, nationally, for Clinton.

But the social media data also shows that this will not be a decisive victory for Clinton – the numbers and structure of the race point to a long, hard slog toward the November finish line.

Clinton bested Trump before, during, and after the debate, both in terms of greater number of positive mentions, and lesser number of negative mentions.

Sentiment towards Clinton and Trump Nationally

This tallied into a solid boost for her overall social media advantage for the day, in which she extended her total positive sentiment lead for the week by close to 15% - opening up a lead of 17% as compared to just over 3% for the week leading up to the debate:

Sentiment towards Clinton and Trump in North Carolina

Normally, this would probably represent something of a knock-out punch in an election, especially against such a widely disliked candidate as Trump is. In fact, most of the international media have been reporting the debate in these very terms.

But this is not a normal election – because Clinton is almost as widely disliked nationally as Trump is. In the list of least popular presidential candidates of all time, she is second only to her competitor.

This is clear from the half-hourly data collected on American social media during the debate: Clinton’s positive mentions only eclipsed her negative mentions during one brief time period the entire night. The graph also shows the negative mentions generally happening earlier than the positive – social media users almost reflexively criticise both candidates.

This negativity has the effect of almost cushioning the crushing rhetorical defeat suffered by Trump – who himself recorded vastly more negative mentions than positive mentions the entire night, but still benefits from Clinton’s own negative sentiment.

Before the debate, Hillary bemoaned the fact that she could not understand why she was not up by 50% in the polls – here’s her answer. She has a likeability problem which means she is battling to absorb US citizens repelled by Trump.

In short, a rampant Clinton victory in the first debate has not altered the six fundamentals of this election; and these fundamentals all point to a very tight race in November – edging towards Clinton – but seemingly unable to be decisively tilted towards her even after a crushing debate victory.

The six, unaltered fundamentals remain:

1. The national polls and social media sentiment will likely continue to tip toward both candidates alternately over time as the US weighs up the two most intensely disliked candidates in history.

2. This election, as per the US system, will be decided state by state. This is vital for outsiders to note: each state holds a number of ‘electors’, according to their population, which are almost universally awarded in a winner-takes-all manner – meaning a candidate can win the election whilst losing the popular vote – as Bush did in 2000. Analysts need to look at the election state by state, with all their varying demographics, electoral importance, and political allegiances.

3. The way the states fall in terms of general party allegiance, matched with the latest data at hand – including polling, other historical trends, and social media sentiment – allow us to narrow, in all likelihood, the result of the election down to three crucial swing states: Ohio, Florida, and Pennsylvania - with the caveat that Trump holds a swing state Romney won in 2012 - North Carolina.

This is where Clinton has a built-in advantage. Because of overwhelming Democratic advantage in big states like California and New York, Trump must win all three of these states to win the election.

4. And Trump is showing consistent strength in North Carolina – putting the three-state game into play. In fact, in North Carolina, his debate performance strengthened his social media sentiment:

Sentiment towards Clinton and Trump in Florida

5. Ohio and Florida, whilst showing varying data points, are both within reach of Trump. Trump seems likely to win Ohio with consistently good polling. But he was eviscerated on social media in Florida on debate night. There does seem to be a data issue in Florida, where Trump is doing surprisingly well in traditional polling while struggling on social media. Reportedly, Clinton’s campaign is in panic mode concerning a lack of mobilisation amongst black voters, as compared to their enthusiasm within Obama’s two victories there. Meanwhile, it is likely Trump is doing well amongst the older, populous demographic of Florida who are likely to be fairly inactive on social media. The volatility of the race is exemplified by Florida’s current data. The debate should give Clinton some bounce in the Sunshine State – but will it hold?

Sentiment towards Clinton and Trump in Pennsylvania

6. If Trump can somehow hold onto his ascendancy in Ohio, and win a close one in Florida as a result of a low black voter turnout there, the race will probably come down to Pennsylvania. Intriguingly, his weak debate performance hardly damaged his sentiment there. He is still underwater, however, both in sentiment and polling.

Sentiment experienced towards Clinton and Trump on back of 1st Presidential Debate

But there is a twist here - Pennsylvania is a highly diverse state politically. Legendary Bill Clinton politico, James Carville, famously described the state as Pittsburgh on one side, Philadelphia on the other, and Alabama in between. The non-urban counties are coal and old industrial country. These are the people who Obama said cling to guns and religion in the midst of economic hardship.

Trump's anti-trade and immigration rhetoric is made for these people. Clinton should be wary of one vital statistic that could push Trump to an upset in a state like Pennsylvania: in 2012, 24 million white men without college degrees (the demographic most favourable to Trump) did not bother to vote – and Obama only won the popular vote by 5 million. The fact that Trump is holding on in the social media data in this state should be very concerning for Clinton.

What is ultimately clear is that this is not a normal election that can be analysed merely by traditional polling and thinking. Trump’s gaffes just don’t hurt him as it would a normal candidate. In a strange way, it almost helps to craft his anti-establishment persona. Meanwhile, Clinton just can’t put Trump away – her support is seemingly weighed down by her low favourability numbers.

Analysts are going to have to closely watch all the data emerging from the key states of Florida and Pennsylvania in order to understand and stay ahead of the curve of what is shaping to be one of the most important and fascinating political spectacles of our generation.