Navigating a 24/7 instant news cycle is a fundamental challenge for gamblers - never more so than when politics is the market in question. Should we listen to the plethora of polls, even if they are often contradictory? Rumours sweeping the internet about imminent endorsements, that often fail to materialise or fail to make an impact? Or the hundreds of experts on TV, coming from different standpoints yet largely buying into the same narrative?
Failing to navigate it successfully can prove extremely expensive in the betting world, and that certainly proved the case on one of the most dramatic nights in the history of political betting. Super Tuesday, the centrepiece of the US election primary process, produced a spate of massive upsets, and 24 hours later, the instant narrative may be beginning to unravel.
First the headlines. As widely expected, the two favourites for the Republican and Democrat Nominations fared best, and maintained their market dominance. After seven wins, Donald Trump remains 1.3130/100 favourite for the Republicans. Hillary Clinton even shorter at just 1.061/18 for the Democrats after eight wins. So far as the Next President market is concerned, nothing much changed. Clinton remains 1.574/7 favourite, Trump next best at 4.216/5.
Those numbers led almost every channel to run predictable headlines about the winners, yet look beneath the surface and actually the results confounded many expectations. At least in the Republicans' case, they may signal a looming twist in the race.
As polls closed, Trump was overwhelming favourite, at virtually unbackable odds, for every state bar Texas, doubtless based on some very one-sided polls. That led some punters to look upon the night as a money-buying exercise. Big mistake.
In Alaska, home to Trump cheerleader Sarah Palin, the GOP front-runner was backed at 1.011/100 to win, equating to a 99% probability. Ted Cruz was matched at 100.099/1 before pulling off an historic upset - one of the biggest ever in these markets.
Similarly in Oklahoma, where Trump had been recorded 11 percentage points ahead by the RealClearPolitics average, he lost by 6% to Cruz after being matched down to just 1.031/33.
In Minnesota, Trump was matched down to 1.11/10 before finishing third, 15% behind easy winner Marco Rubio.
It could have been even worse for Trump backers. He was engaged in a neck-and-neck struggle with John Kasich before winning Vermont, where he had been recorded 15% up. Kasich had been rated fourth.
Ditto Virginia, where he won by 3% after being projected 15% ahead. In Arkansas, the odds flip-flopped many times in a tight race with Cruz.
On the Democrat side, Bernie Sanders didn't pull off any massive upsets, but he did win four states, defying the Clinton clean-sweep narrative. Given that few expected him to compete in the South anyway, there could be more to come in more favourable states.
Now, it would be easy to put the GOP upsets down to bad opinion polls, but that isn't an entirely satisfactory explanation. In the previous three contests in New Hampshire, South Carolina and Nevada, the polls had successfully forecast easy Trump wins, perhaps even under-estimating the scale of his support.
The question, therefore, is whether this relative underperformance reflects a dip in his wider support. Trump was recorded up to 49% in a national poll at the weekend, taken before his rivals finally upped the ante with serial attacks at last Thursday's CNN debate and beyond. Since then, he has come under relentless fire for failing to condemn the Ku Klux Klan and it's former leader David Duke, not releasing his tax returns and the ongoing controversy regarding Trump University.
Could it be that, as his rivals and the GOP establishment begin to unite and aim collective fire, that the Trump bandwagon is finally beginning to slow? Expect tonight's Fox News Debate, starring Trump's least favourite presenter Megyn Kelly, to be another firecracker.
Follow me on Twitter @paulmotty and at my website Political Gambler
For the latest on the US Election and all of our other politics markets, check out the new, improved Betfair Predicts, which forecasts outcomes based on the very latest Exchange data...