In 1967 a horse called Foinavon won the leading race in Britain, the Grand National, at 100/1. A melee at one fence stopped every other horse. Foinavon was so far behind he was able to avoid the panic and confusion.

In 1993, in California, the Breeders Cup classic went to 133/1 Arcangues, a horse which had shipped in from France and had never before run on a dirt track, the equivalent of diving off a highboard even if you may not be able to swim.

So don’t tell me Donald Trump has no chance of winning on November 8. His odds are only 10/1, a mere piffle compared to the prices of those equine long shot winners.

Here is the case for a longshot Trump win:

The polls are wrong.

Poll: Is the media biased against @realDonaldTrump? pic.twitter.com/9eI7ctx1a7

— FOX Business (@FoxBusiness) October 21, 2016

Possible. Indeed two polls, the Los Angeles Times one and Investors Daily one, show Trump level and up by two respectively. The Investor's Daily poll has been the most accurate of all in the past few election cycles according to Nate Silver of fivethirtyeight.com, the gold standard polling prediction firm.

Then there is the Rasmussen Poll, which today (Thursday) shows Trump up by three points, one of his biggest margins.

So when we say the polls are certain Hillary will win we should be absolutely clear that some polls, including one highly respected one, see her losing.

Read more: A Man in Full - exclusive interview with CIA Director John Brennan

How can this be? The best recent example to back this up is the last British election, which took place in 2015, not the Brexit polls.

An average of 11 different polls just before that election gave Labour a narrow 4 seat majority over the Tories, but it was clearly neck and neck.

New poll has Trump and Clinton neck-and-neck in Ohio https://t.co/g4pybSINcn pic.twitter.com/saWrCRaSUh

— CNN (@CNN) October 20, 2016

Only one polling company, Survation, picked up a huge late surge to the Tories but considered the poll fatally flawed and never even released it.

In the event the Tories galloped home with an overall majority gaining 330 seats when the polls had forecast almost sixty seats fewer.

There are common factors between that and the Trump/Clinton race. The “shy tory” vote were the unaccounted for Tory voters who, due to their party's poor reputation on liberal issues, would not admit to voting for them.

Then there was the “Lazy Labour” faction which saw Labour voters lulled into a sense of complacency because they were leading in the polls and were not bothered to vote.

It is easy to draw a similar line for this election in the US. Shy Trump voters thanks to the media's disdain for their man may be a huge factor. Equally, with pundits proclaiming the race over and won many Democrats may become lazy about turning out themselves.

Factor in the unknown unknowns such as a major terror atrocity, a vicious crime spree traced to illegal aliens, or a number of cop killings. Any of those could influence the race.

There is also clear evidence that polls are skewed towards Clinton at the moment. I’d bet all the tea in China that Trump will not lose Texas, equally that he will win Utah and likely Arizona. Talk of Dems winning those red states is just that, talk.

Great new poll shows me barely leading @Evan_McMullin, who I never met, by 1 point in #Utah. This is a disgrace! McMullin should RESIGN!! pic.twitter.com/WXnc3jO2CZ

— DonaId J. Trump (@realDenaldTrump) October 21, 2016

So there is still a race to be run despite the mainstream media calling it as over. They should listen to Yogi Berra and remember Foinavon and Arcangues.

Miracles do happen, longshots do win. Elections can "upset the apple tart" as former Irish leader Bertie Ahern once said hilariously.

It’s no time to be counting unhatched chickens.

Donald Trump could still do it? Could the polls be wrong. Flickr