Hunter S Thompson can give us some sound betting advice

Hunter S Thompson liked to bet.

In a book on the 1972 US election he wrote: "I had never covered a presidential campaign before I got into this one, but I quickly got so hooked on it that I began betting on the outcome of each primary and, by combining aggressive ignorance with a natural instinct to mock the conventional wisdom, I managed to win all but two of the 50 or 60 bets I made between February and November."

He did the same thing with sports. In an article on the 1974 Super Bowl in Rolling Stone he wrote: "There is a definite, perverse kind of pleasure in beating the 'smart money' in sports, politics, or anything else and the formula for doing it seems dangerously simple: take the highest odds you can get against the conventional wisdom but never bet against your own instinct or the prevailing karma."

Thompson is no longer with us. He is best known for a book called Fear and Loathing in Las Vegas, which is not about gambling, and a magazine article called The Kentucky Derby is Decadent and Depraved, which is not about horse racing. "Unlike most of the others in the press box, we didn't give a hoot in hell what was happening on the track. We had come there to watch the real beasts perform."

But he loved to bet. And there is something to be said for his betting philosophy. Ignorance probably should not be backed, aggressively or otherwise. And even a good bet-selection method will usually fail more often than Thompson's did in 1972. But it is sensible to bet against the conventional wisdom - carefully, though, not indiscriminately.

You cannot reasonably hope to make money from bookmakers by betting with the conventional wisdom. By definition that is what nearly everyone else does. And they must lose. Otherwise how do bookmakers report huge profits every year?

When Marcus Rashford scored four goals in his first two games for Manchester United he was hailed as a world-class striker in the making. He might be. He might, though, have been a journeyman in the making. Or anything between.

How highly was Adnan Januzaj praised after his first games for Manchester United? More than one country pleaded with him to play for them. Now Januzaj is on loan to Sunderland, who are likely to be relegated and he does not look too good for them.

Federico Macheda scored winning goals for Manchester United on his first two substitute appearances. After loan spells with the likes of Queens Park Rangers, Doncaster and Nottingham Forest, he now plays in Serie B for Novara.

When a young player breaks into a top Premier League team the best future is predicted for him. That could materialise, but so could other things. Hardly anyone acknowledges the other possibilities.

It is the same with teams whose games have featured lots of goals or corners or bookings or anything.

I tend to bet that the splurge has ended. Often that is the least likely possibility, but sometimes it is more likely than the odds suggest. And that is because bookmakers set their odds knowing that most bettors will follow the conventional wisdom and assume that what they have just seen they will see again.

Most scientists think that way too. And more often than not they are wrong. John Ioannidis wrote a paper in 2005 for PLOS Medicine explaining Why Most Published Research Findings Are False.

Others added more evidence. In 2012 a biotech company called Amgen repeated 53 famous experiments and got the same results in six. A year earlier the pharmaceuticals company Bayer HealthCare had compared findings from 67 published trials with their own and got a match in no more than 17.

In 2015 the Center for Open Science repeated 100 experiments that had been reported in prestigious psychology journals. They got the same results in 39.

So the replication rates, as we might call them, were 11 per cent, 25 per cent and 39 per cent - all different, but all low. Most published science is wrong.

Scientists, though, do not make things up. Indeed they record numbers as carefully as any and more scrupulously than most. They have seen what they say they have seen. But in most cases what they saw is not what others who do the same thing will always see.

Hunter S Thompson would have snorted derisively, then said, probably in more colourful words: "I told you so."

Dundee may have made a big mistake by sacking boss Hartley

Dundee sacked manager Paul Hartley after seven defeats left them in the relegation playoff place in the Ladbrokes Scottish Premiership.

Managing director John Nelms said: "Unfortunately, the business we are in, sometimes change is necessary to achieve the goals we have set." Unfortunately for Mr Nelms, he may be wrong.

As Leonard Mlodinow explained in a book called The Drunkard's Walk: "Mathematical analysis of firings in all major sports has shown that those firings had, on average, no effect on team performance." Those findings have been replicated.

I added a scrap to the heap of evidence with my own investigation of the sacking of managers in the Premier League. On average teams got the same results afterwards as rivals in a similar predicament who did not sacrifice their manager.

For every Craig Shakespeare there is a Steve Agnew - which is not to say that Agnew is a worse coach, just that Shakespeare has got better results, so far.

Nelms said of Hartley: "I think somebody in the near future is going to get a good, hard-working manager." You could not make it up. A good, hard-working manager? That sounds like just what Dundee need.

Nelms said later that he was looking for a British manager with knowledge of the Scottish Premiership and Dundee. He employed Sky Sports pundit Neil McCann.

You might imagine he would have jumped at the chance to employ instead a manager who had won promotion to the Premiership then worked there for almost three seasons. But no. That is a description of the manager Dundee sacked.