How the news cycle (and links) will predict the election

“The stats say she is winning right now and while [Donald] Trump has a chance, it has to be lower than Clinton’s,” he added.

Whereas Clinton’s link data is consistent, much like in life perhaps, Trump and his link data have wild fluctuations. He is also gaining in terms of the number of links, as well as in Trust Flow, which is Majestic’s measurement of the overall influence of his website. And because his links and his Trust Flow go up when he dominates a story in the news, the outcome of the election will depend which way the pendulum happens to be swinging at that moment and whether his Trust Flow is therefore experiencing a high or a low.

“If [Clinton] makes another blunder as big as the email scandal or the pneumonia or we suddenly find she has two illegitimate children, then Trump will walk it, but if she can keep her nose clean and be statesmanlike, Trump will never catch her,” Jones said.

And so we just have to wait and see.

Swing states

In the meantime, Jones said Majestic does not break down findings by state, but the data nevertheless shows Clinton’s voters are more likely to swing key states better than Trump’s will. In other words, Clinton’s more consistent links mean it will likely be harder for Trump to eat into left-leaning states than for Clinton to eat into right-leaning states.

“A stable link profile would suggest that Clinton would be better off pushing a positive message to non-supporters, whilst Trump may just as well try and bring Clinton down with headline stoppers,” Jones added.

Which seems to be precisely what the candidates are doing.

Majestic’s roadmap

Per Jones, Majestic is one of the biggest crawlers on the Internet, but instead of on-page content like Google, Majestic analyzes the way pages link to each other.

“A good way of interpreting that is if you imagine a roadmap of all the roads in America linking cities and towns, you don’t need to know the population of each city to get a sense of the importance of [particular] cities,” Jones said. “Links can have the same kind of importance and we can build a map of how every page on the Internet impacts others.”

Spun another way, Jones said, “If you believe the mark of a person is not what they say, but what others say about them, then looking at what webpages link to each candidate reveals quite a story.”

Predicting elections

Jones said he uses Majestic’s Link Graph, which maps all the links and sites that link to a particular site, to make predictions like this. Previously, Majestic has made similar calls for the 2012 London mayoral race, the 2012 US election and the 2016 Republican primary.

In fact, Jones said Boris Johnson’s 2012 candidacy for mayor of London was not unlike Trump’s campaign today.

“[Johnson] came in and started to create waves and news, which was not a dissimilar campaign, although he was not quite as audacious as Trump in trying to create news and a character for himself that eventually got him elected as mayor of London,” Jones said.

Further, when Majestic was tracking the absolute number of backlinks to mayoral candidate websites in 2012, Jones said Majestic saw Johnson increasing link equity over time, which was around the time he started to go ahead in the polls. And this could very well happen with Trump, too.

But, Jones noted, using links to make predictions does not always work. Case in point: Majestic got the Scottish referendum wrong, but Jones said this might have been because Majestic analyzed links from around the world, including those linking from well outside the voting area.

In the case of Clinton v. Trump, Jones said the “world effect” is indeed likely to skew Majestic’s data – probably toward Clinton – but he noted the difference is that the US is large compared to Scotland and, as a result, the skew will not be so abnormal.

Gaining links

As of mid-September, Majestic said Clinton is leading on every metric it tracks — including who has the most links, how trustworthy those links are and how many websites are involved in linking to each candidate:

Majestic’s Trust Flow is an overall score between 0 and 100 that attempts to quantify the relative influence of every page on the Internet. Majestic can then use this to compare the relative Trust Flow for each candidate’s site. As of September 14, Clinton leads by a score of 58 to 55, but, Jones noted, this could easily change.

And while Clinton has more links — 688,000 to Trump’s 565,000, to be precise — it’s much closer when you look at the domains linking, Jones said.

“What’s really interesting to me is Trump over time has been increasing his Trust Flow score…which has been going up over the last few months, whereas [Clinton] has had a static Trust Flow score over the last few months,” Jones said.

In fact, Trump has what Jones called a “weird chart” in which Trust Flow goes up and down, which he said is unusual.

“It appears every time he makes a big announcement or claim – ‘Let’s build another wall,’ ‘Let’s throw another million people out’ — whatever his PR piece is, it has an affect on his Trust Flow,” Jones said. “I can’t say he’ll beat [Clinton], but he’s getting more serious websites linking to him in the past 90 days and it would appear more trustworthy websites are linking through to his website right now.”

Trump is also gaining links faster and will have more links by the election, Jones said.

Further, Jones noted if Trump’s Trust Flow surpasses Clinton’s, he will have started to take the lead, which could happen by the election based on the way he is managing the news cycle.

Clinton’s score, however, is only likely to go down if respected groups remove their support – and their links.

“But [Trump] is on the up,” Jones said. “Every time he says something to a different group, he’s gaining new stuff, [but] that up and down seems to make it less confident he is keeping people that are voting.”

‘Like a yoyo’

Here’s a closer look at Clinton and Trump’s Trust and Citation Flow scores since July. Citation Flow is more about the absolute volume of links and starts from the basis that every page has X webservers linking to it, while Trust Flow initially discounts the sites Majestic does not know were curated by humans, Jones said.

Here, it’s easy to see Trump’s Trust Flow jumping up and down “like a yoyo,” Jones said.

“This is unusual for a website, but reflects how the Trump campaign is making more waves in the news cycle than Clinton,” he added. “However, when the story drops off, so does Trump’s Trust Flow score.”

The up and down likely means Trump is getting links from powerful publications, but because it’s news, the links disappear when the news cycle subsides, Jones said. That’s because the link goes off the home page and onto a less important page.

“In order to maintain ongoing Trust Flow, [Trump] needs to continually get into the papers to get Trust Flow to go up,” Jones said. “It would seem to me that if he starts to calm down between now and November and put less alarmist stories out there and [act] more [like] ‘I can be President,’ he will do scarily well. It seems that when he stops doing a bad story, people forget the outrageous things and start to believe some of the other things he may or may not be standing for.”

Still anybody’s game

According to Jones, the takeaway is that if Election Day falls when the US is “in the right mood,” Trump could very well win the day.

It also confirms what pollsters are saying – it is hard to call before the debates.

“I think that the live debates will make a significant difference either way. If one of them messes up…the other one has it in the bag and I would say that’s probably [Trump’s] best opportunity to get through a more reasonable side of himself than perhaps people have seen before,” Jones said. “He already has the extremists and the far right and people that have shall we say ‘disenfranchised lifestyles,’ but he needs more to be able to grab some of that middle ground. He is gaining – there’s no doubt about it. His credibility seems to be hardening up.”