THE LAST MILE TO VICTORY: NEVADA

We have updated our analysis and present our final view.

This is a nail biter. Down to the wire. 

The only certainty we have now is the Polls got it so wrong. Again, and again, and again.

Polls Kill Democracy

I will do a separate write up in the coming week on Polls and WHY we need to stop using them.

Where are we right now?

As at 1.30pm AEST, there is one clear path towards victory for Trump or Biden: through Nevada.

Nevada was off the Maven radar. We did not analyse Nevada in our prediction on election eve, 3 November 2020. So the global prediction team has fired up our engines to analyse this crucial state and present our final update on this election. 

The rest is up to the stars.

Quick check-in: Maven US Election 2020 Scorecard

So far, the State predictions by Maven have been on the money, with 4 out of 5 correct in the declared states.

Our prediction was built up by analysing each Battleground State across 6 key attributes that determine voter favour. Our prediction was that Trump would win. Meanwhile all major polls suggested that Biden would win by a landslide - however this is beyond statistical error. (Polls are clearly not a valid indicator and should not be used anymore, please). Refer to my previous articles to learn more about how this election has unfolded over the last 4 weeks.

Our data and methodology clearly works. Team Maven has done us proud.

Biden flipped Michigan. Trump flipped Florida (the Spanish vote counted, again).

Biden is now at 264. He needs 6 to win through Nevada. 

Based on the current tally for the remaining Battleground States - Georgia, North Carolina and Pennsylvania - Trump is leading, which is consistent with Maven predictions.

Biden can win Presidency if he wins Nevada. And currently, he is leading by a margin with 16% of votes left to count.

I have no more nails. This is knife edge.

Source: The Associated Press

FOREWORD: NEVADA

Nevada was not on our radar 48 hours ago. So our analysis of Nevada is not ‘like for like’ as our analysis of the 8 Battleground States. It’s not apples to apples, but we can still make this fruit salad taste good.

The content we are analysing on Nevada is no longer locally focussed. The data and our measurements on engagement and emotion will contain content around the context that Nevada is a decider in this election, as part of the national discussion. 

This is important to note. 

When we work with data, we work with what we have, which so far, has worked very well. We used the same data and methodology (refer to previous article for explanation).

We applied the same logic to arrive at our updated prediction.

Context: Observations from Sydney, Australia:

Democrats have held Nevada since 2004 and Trump fell shy of winning Nevada in 2016. This year Trump campaigned hard. 

In 2020, Nevada has seen high voter turnout, surpassing that in 2016 - indicating that this market is committed and wants their voices heard.

Nevada is largely a desert state with low inhabitants. It is home to Las Vegas which thrives on retail, tourism and hospitality, which has been severely impacted by Covid. Nevada is also home to a significant Hispanic and Latino working class population, who work ‘back of house’ supporting local and internal talent ‘front of house’, and possibly come from all over the country and the world. So in short, workers in Nevada may not constitute as registered voters in Nevada.

This makes the postal vote element a curveball. 

Why does this impact our read of Nevada?

Our method of predicting this market is based on petabytes of online content surrounding this election and the state of Nevada. Our read on this market measured engagement, affect and emotion of people interested in this content. We don’t know who they are or where they live (that’s Cambridge Analytica territory). We look at the market as a collective. We may be analysing the emotions and engagement of people who are not registered voters of Nevada. So in this state, in particular, our read on emotion may not result in predicting favour in this election.

However, we are going to do this anyway and see what we see. So far, it has worked in 4, if not 7 of the 8 Battleground States.

I back our data and our methodology.

FIRST LENS: TRUMP VS BIDEN (English and Spanish market)

You should all be experts in reading this now. In case you need a refresher on how to interpret the Emotion Eyeballs, refer to my first article in this Trump 2020 Predictions series.

So what does this say?

Trump is favoured in the English and Spanish speaking market. The Spanish speaking view is more intensely negative towards Biden. I’m calling out deep ‘anxiety’ and ‘distress’. 

In this view, Trump wins Nevada.

SECOND LENS: TRUMP VS BIDEN on Jobs

We know Jobs has played a critical role in 2020. So we went straight to analysing both the English and Spanish speaking views on Jobs in Nevada.

It is very clear that the emotions towards Biden are negative. Whereas the Spanish market is deeply positive towards Trump. 

Interestingly, the Spanish speaking market once again favour Trump.

On this basis, Trump wins on Jobs.

THIRD LENS: TRUMP VS BIDEN on Covid

We assessed Covid as a top 3 critical issue for this election in most of the Battleground States. Considering the importance of tourism, retail and hospitality to Nevada, we analysed Trump and Biden on Covid in English and Spanish.

Let me introduce a new chart: Engagement Quadrant. Put simply, the top two quadrants represent narratives that are relevant and engaging.

• Timeless means relevant to culture, society, deeply engaging. 5% of narratives in general are Timeless
• Transformational means Timeless but has momentum and will change the world. 2% of narratives in general are Transformational. These are the gold nuggets
• Transient and Tribal are two definitions of ‘noise’, meaning they are not relevant, will fall away over time. 93% of narratives fall in these categories and we should ignore them.

Okay, ready to look at Covid?

We see that in Nevada, Covid is an election issue for English speaking but NOT Spanish speaking. Interesting. The blue collar back of house workers are not concerned about Covid, and hence we could assume they would vote in person. Note this point for later.

For those that cared about Covid, the emotions around Trump are very negative.

Based on this view, it’s a tough one to call. On one hand, Trump is not favoured by some Nevadans based on his response to Covid. 

Neither Trump or Biden wins on this lens.

So where are we at now?

Based on the summary above, our prediction is Trump to win Nevada. Hence, our Maven Prediction is Trump will win the US Presidential Election 2020.

But you know that the story doesn’t end here…

Analysing Nevada, after the votes have started to be counted, has now created more anxiety. We also have more facts to work with.

How do we see this end? Do we revise our Prediction? 

Let’s consider one last angle.

THE CURVE BALL: POSTAL VOTE

This is becoming clear as mud. So let’s stick to the facts and see where we land:

• Our analysis on Nevada favours Trump to win this State
• The current vote count says that Biden is ahead, with the lead narrowing
• 16% of votes are yet to be counted, most of which are postal votes
• We don’t know the composition of postal votes (are these Nevadans in other states, or overseas?)
• Postal votes so far have favoured Biden.

My head is spinning. I am getting demented.

To take this prediction to the end with some level of certainty, we would have required:

• AI read down to counties - which we didn’t do
• A cleaner AI read on Nevada the day before Election Day - which we didn’t do
• Data on voter registrations and where they live - which we don’t have.

Still, pushing through with some logic again...I’ll lay it out for you to decide. 

This prediction game is now out of our hands. 

FINAL WORD AND PREDICTION:

Consider this:

• We saw that Spanish speaking Nevadans are not concerned about Covid, so are likely to have cast their vote in person (i.e. will not add to the postal count...so no more postal votes for Trump expected)
• We saw the English speaking market broadly favour Trump and are more concerned about Covid, so are likely to not have cast their votes in person...so more postal votes for Trump expected
• Trump has been encouraging people to vote in person...so no more postal votes for Trump expected
• Postal votes - generally - swing Democrat...so more votes for Biden expected.

And lastly:

• The current vote count is showing Biden with a comfortable lead.

What do you think?

So based on ALL the above, and the weeks of analysis to reach this point..

 *cross my heart, hope to die, stick a needle in my eye*

I revise my prediction and call that Biden looks clear to win Nevada and hence, the 2020 US Presidential Election.

There we have it.That’s a wrap.

But wait, there’s more:

If only democracy played out democratically. 

Trump will challenge this result. This election will be drawn out in the courts for weeks. Trump may still take the White House in December, ala 2000 Bush vs Gore.

By that time, my nails would have sufficiently grown back for more nail biting drama.

America, brace yourself for some ugly displays of human behaviour. The world at large supports you through the change and rebuild that needs to happen in 2021.

MAVEN PREDICTION: BIDEN TO BE THE 46TH PRESIDENT OF THE UNITED STATES.

….and once again, POLLS ARE DEAD

I hope you enjoyed this 3 part US Election 2020 series. 

If you like this content, connect with me on Linkedin, Instagram and YouTube. Interested in your take on this approach. Did you get demented?

*****************************************

Elisa Choy is an economist and strategist with a magic eye for translating data. She is also a professional speaker and media commentator who elegantly translates big data and AI into engaging and thought-provoking stories on human behaviour, society, culture and business. She predicted the winner of The Voice 2019 and 2020, and MasterChef Australia 2020.

She is Founder of Maven Data, Australia’s first AI-powered market research agency, and Strategic Data Central, a boutique data analytics consultancy transforming organisational internal data into actionable insights.

She has worked with Australia’s leading brands, large and medium enterprises from SMEs to ASX10 including Woolworths, Caltex and Seven West Media.

Subscribe to her content on mavendata.com.au, Instagram and YouTube.

Elisa acknowledges her Partners in this research:

Maven Data uses proven advanced analytics technologies to analyse market emotion and predict industry trends. Elisa partnered with Significance Systems which developed Earth.ai, a platform that identifies deep, underlying narratives and their key drivers, for any topic, in any language, over the open internet. The Australian company's founders, Darrell Berry and John Ricketts, were heralded by Forbes magazine as developing the idea of ‘social media’ back in 1994.