Finding an Edge

Let's be very clear here, there is no mathematical edge to be had with fixed-odds bookmakers. No bookmaker will bet to a percentage remotely close to 100 percent, let alone less than it. But can you blame them? It would be a slow and painful form of suicide.

What never fails to amaze, however, is just how a dozen bookmaking firms can price up a 25-runner handicap hurdle on a Saturday afternoon, and each be perfectly in line with each other's prices when they are given to the Racing Post the evening before.

With so many runners, and so many differences of opinion from odds compilers, you would imagine the best possible percentage (using the biggest price offered about each horse) would occasionally breach the sub-100 percent barrier; remarkably, it never does.

That shows you just how difficult it is to find an edge. Occasionally, though, someone does the unthinkable and unearths a small margin play.

I'd guess the last time was 10 years ago, when the bookmakers were offering 100/1 about matches resulting in a 3-3 draw. Some boffin somewhere burned the midnight oil to discover that the true probability of a 3-3 draw in a Scottish Division Two game was nearer 70/1.

It was only a tiny edge, maybe, theoretically meaning a £10 stake on 70 such matches would result in a £1,000 return for the £700 outlay, but ingenious all the same. It was such a shame that there were so relatively few games in this division, and unusual, sizable bets on such an outcome soon set off alarm bells, leading to the rick being discovered by the bookmaking firms, who soon put their prices in line.

I'm also a big fan of John Carter and Paul Simons, who researched and studied scores at numerous championship golf courses, realising that at some, during a major tournament, the probability of a hole-in-one was long odds-on. For example, of the 38 major tournaments in 1990, 30 featured a hole-in-one.

Together, the following year, they drove the length and breadth of the British Isles, where they managed to find some independent firms that were prepared to offer them up to 100/1 about a hole-in-one at a major, and take the bets in singles, doubles, trebles, and upward.

As a result, they won a half-million pounds and closed down a lot of small, naive bookmaking businesses that would not and could not pay out.

So, how do you make a fortune at gambling through pure mathematics as opposed to the traditional betting in a belief that the price of your selection is bigger than the probability of it winning? The answer is simple: Bet against other punters through the medium that is a tote or pari-mutual system.

Now, contrary to popular belief, I'm not wholly convinced that "Mr Scoop 6" Harry Findlay, a guy I've known for the better part of 20 years now, is that far in front with the weekly Saturday afternoon [Scoop 6] totalisator bet on horse racing, despite winning it several times for sums up to and beyond £1 million.

But one man who tackled the tote pools successfully for many years and came out in front to the tune of an estimated £100 million to £200 million was Australian Alan Woods. I never met the guy or even knew of him until recently, when reading his obituary, sadly, but this is a fella you have to respect.

Born in Australia, at school he was good at maths, but at the university, he preferred to study the local poker machines and horse races before studying to be an actuary and working as an investment analyst for a merchant bank.

In 1972, Woods turned his mathematical attention to card-counting in blackjack. A string of successes in a Tasmanian casino encouraged him to graduate to the casinos of Las Vegas, before moving on to Europe and Asia. But after more than a decade, he gave up card-counting and turned his attention to horse racing in New Zealand, before moving to Hong Kong in 1984.

In Hong Kong, the same small pool of horses ran in the same type of races on the same two tracks, Happy Valley and Sha Tin, where thousands of largely ill-informed punters bet into enormous pari-mutuel pools. Their enormity meant that big punters could have big bets without decimating a horse's odds, something that could not happen at tracks in Europe and even largely bookmaker-less America.

Woods' mathematical brain joined forces with new business partner Bill Benter's innovative computerised betting systems to develop programmes that, in 1987, after three years of refinement, produced a modest profit of €100,000.

Betting against a public -- who were often doing their Hong Kong dollars following tips, backing lucky numbers, and studying the phases of the moon -- value abounded. Meanwhile, Woods and Benter continued to collect detailed data for every horse, and applied an increasingly sophisticated formula to calculate the probability of each horse winning the race being analysed.

The formula took account of a range of factors, such as distance, weight, last result, number of starts, and starting stall, with weights attached to each factor. This computer -- with software similar to what Wall Street's sophisticated hedge-fund analysts employ -- calculated the probability of a selection and then related it to the odds available, and "overlays" (horses with a markedly better chance of winning than the odds available about them indicated) were identified. Further calculations were made to identify selections for the exotic, big-pool jackpot bets, popular in Hong Kong.

The size of the bets? "The amount of money we bet has been limited only by the size of the pools," Woods was once quoted. In other words, he and his team were so confident of winning that they wagered as much as they possibly could before affecting their potential returns too dramatically.

Woods and Benter split up in 1987, but both subsequently thrived. The following year, Woods won the equivalent of £226,000, and the year after, £527,380, followed soon by £828,740 and £1.4 million. One day, he won £1.5 million, although on another day, he lost £1.65 million. Overall, the winnings massively exceeded the losses, and the scale of Woods' operation grew.

Woods wasn't to be found at the racetrack, though; he hadn't visited one since 1985. He therefore saw no individuals on the track, recognised no names, and cared little for the personalities. Racing was nothing more than a mathematical equation to be solved with all the players nothing more than a never-ending string of numbers.

It is somewhat rare for a man so immersed in a game not to attend its playing out, if only infrequently. But Woods had no interest in horses, as such, and felt no compulsion to attend their running. Such behaviour did not impact negatively on his punting. To the contrary, such detachment was an integral element of his success.

Unsurprisingly, a friend of Woods' once likened him to Howard Hughes, based on his reclusiveness. He could be found only in an office, organising his team's assault on the pools; it was a team of expert analysts as well as those putting bets on. The huge bets were eventually said to account for between 1 percent and 2 percent of Hong Kong's betting turnover, which is about £30 billion a year.

To put the enormity of the racing in Hong Kong into perspective, the tote's retention from one day's racing in the country equates to the cost of running and employing the country's police force for an entire year!

That is based on a tote rake of 18 percent, which is sizeable enough, but it makes Woods' computer system for selecting horses and betting permutations that were out of line to the probability of them prevailing all the more impressive.

Who said punting should be fun, anyway?