By the time you read this, the 45th president of the United States will have been elected.

The polls dramatically tightened over the final two weeks of the campaign. The candidates were deadlocked nationally and tied in the key battleground states that decide who wins.

It will be fascinating to learn if the American people were actually willing to elect a person who is under multiple investigations to be president? In the closing days of the campaign, it became known that the Clinton Foundation has also been under federal investigation for more than a year.

We've never seen anything like this. We have seen thousands of emails dumped by WikiLeaks on a daily basis for the last month of the campaign, then the uncovering of 650,000 Huma Abedin/Clinton emails because of the criminal investigation of disgraced former Congressman Anthony Weiner for his alleged under-aged sexting scheme.

By my filing deadline, I had Donald Trump with 265 electoral votes. My analysis right before Election Day is that Trump would win all of the states Mitt Romney carried in 2012 and that he adds North Carolina (15).

That gave Trump a starting point (to the path of 270) of 206 electoral votes. I added Ohio (18), Florida (29), Iowa (6), Nevada (6) for a total of 265.

Things that I couldn't factor in. The late deciders. Who did the undecided voters ultimately select? In every election in American history, undecided voters choose the party out of power.

If Trump made it to 270 and beyond, he would have had to win one or more of the following: Colorado (9), New Hampshire (4), Pennsylvania (20), Michigan (16), Virginia (13) or Wisconsin (10).

Some of these states haven't chosen a Republican candidate for president in two or three decades.

In the closing days of the campaign, all trend lines were pointing in Trump's direction. Did the momentum continue through Election Day?

Something that also made an impression upon me in the closing days of this highly competitive campaign was that every dispassionate review of the election projected that Trump would win, including:

Stoney Brook University political science professor Helmut Norpoth, using a statistical model that has picked every winner (except President Kennedy) for 114 years; The United Kingdom betting markets (the same ones that correctly picked BREXIT); professor Allan Lightman, who has selected every presidential winner for the past 30 years; and the financial markets in America

(S & P 500) has conistently picked the the winner for 88 years; Alan Abramowitz's "Time for Change" model was done for the University of Virginia's Center for Politics, "Crystal Ball." It yielded a 51 percent to 48 percent Trump win.

CNBC reported in the final days of the campaign that Trump was more popular in 2016 than Barack Obama was in 2008. Their computer poured through millions of documents and concluded that Trump's popularity made him a lock to win on Nov. 8, 2016.

Another artificial intelligence system MoglA was developed by Sanjiv Rai. The system calculated 20 million data points to create election predictions. It was created for the 2004 election. It has correctly picked the winners of primary and general election each time since then.

These objective models all had Trump winning the presidency. This moved the needle with me, because each and every one of these examples removed human elements from the equation.

Every analyst, including me absolutely cares who wins whether they are willing to say so or not. These statistical models don't care who wins.

Even though Clinton's path to 270 has always appeared to be wider, my final analysis, using the late trend lines, coupled with undecided voters and a final look at the numerous Clinton scandals all lead me to predict that Trump will win the election.

If Trump wins, the Senate remains in Republican control. The House of Representatives is not in doubt. If Clinton wins, the Democrats only need 50 senators to win the majority.

Finally, here are my last two rationale for picking the next president: Kids voting for president have selected Trump. This has always mattered, because it directly factors in what they hear their parents talking about at home and teachers at school.

Lastly, the sale of Halloween masks has always been another foolproof predictor. This has predicted every actual winner since 1996. This year, it was a 3 to 1 advantage for Trump over Clinton.

In the end, human intelligence believes that Clinton will win. Artificial intelligence computes that Trump will win. Will the political soothsayers with beating hearts prevail or do the inanimate objects get it right, again?

I'm hitting the send button now to submit my column. Let's see how I did.

Harry Hurley is the president of Harry Hurley Consulting and Communications, LLC. He hosts the daily talk radio program "Hurley in the Morning" 6-10 a.m. weekdays on Townsquare Media, WPG Talk Radio 1450, where he also serves as the senior programming consultant.  Hurley has hosted various programs for local television and is the editor and publisher of his news and information website, www.harryhurley.com. Send comments to HarryHurley@aol.com