Australians are set to bet around $300 million on this year’s Melbourne Cup. For many people, the race that stops the nation is the one day of the year when they gamble.
If you try your luck this year, David Beirne is most likely the man you’ll bet against. He’s in charge of setting the odds for the Cup and hundreds of other races every year at TAB, the nation’s largest bookmaker.
While the third-generation bookie’s father and grandfather worked at the trackside betting ring, the 35-year-old runs his operation from a standing desk on a 29th floor office in the centre of Sydney.
From this vantage, instead of looking out over punters and turf, he looks out on a sea of computers and screens crowded with spreadsheets – it could be any other corporate office if not for the live sport and racing.
A team of about 100 “traders” take thousands of data points, crunched through quantitative analysis, and mix that with old-school bookmaking knowledge and intuition to decide how likely it is an athlete, team or animal will win a particular event. Then, they cash in on that insight.
“I used to enjoy doing puzzles when I was a kid and all this is a puzzle on a big scale,” says Beirne.
“We’re never always right and nor can we be – that’s what makes it exciting. It’s kind of the challenge.”
Here’s a rough guide to how bookmakers set the odds, and why (usually) they come out on top.
How do odds work?
Odds, presented as a dollar figure, reflect the bookmaker’s opinion of how likely it is something will happen. The higher the odds, the less likely; the less likely something is, the more they will pay out to someone who correctly bets it will happen.
Placing $1 on a winning horse at $50 sees the punter make $49 by risking $1 on an unlikely outcome. If a horse runs with odds of $2, it’s because bookies consider its victory highly likely and so will only give you back $2 for every $1 you bet (so the punter makes $1 profit).
Say, for example, a bookmaker thinks a horse has a 25 per cent chance of winning a race. Turning that probability into an odd is simple maths, like this:
You can take a horse’s odds and do this same equation in reverse to understand how likely bookies think it is to win.
(This is how it works for “fixed odds” betting. “Totes” or totalisator betting works differently, by pooling all the money and presenting odds prices according to how much of the total pool is bet on it.)
How do bookies figure out the probability of a horse winning in the first place?
Bookies give each horse a numerical “rating”, which represents where they think it will finish relative to other horses in a race.
They start with the horse’s previous races – how much it won or lost by, its starting price, and the overall time of the race. Much of that can be quickly crunched by computer analysis.
On top of that, bookies consider dozens of other variables. The barrier draw, how much weight the horse will carry as a handicap, the jockey, the track condition, the race class and even the weather forecast.
“There’s no exact science to it,” says Beirne. “We’ve worked on trying to turn the art into more of a science – we blend the two together.”
Once a horse has a rating, the bookmaker can decide the probability of it winning and offer punters an initial price.
Bookies open betting on some horses for the next year’s race almost as soon as the Melbourne Cup finishes, and the odds change from there, sometimes quite dramatically, right up to the moments before the barriers open again on Cup Day.
One of this year’s favourites, Surprise Baby, opened at $26 immediately after last year’s cup and over the past year has narrowed to $9 heading into Tuesday’s race.
Russian Camelot opened at $201 on November 7 last year, was cut from $26 to $11 after it won the South Australian Derby on May 9. It’s now a $13 chance.
And TAB cut its odds for Cox Plate winner Sir Dragonet from $51 to $18 after it was bought by an Australia ownership group on August 18.
Beirne describes this as “course correcting” – as new pieces of information come to his team’s attention, each horses’ odds change accordingly.
What do bookies add to algorithms?
A computer misses the full story of how a horse is likely to perform.
Only a human mind can evaluate its barrier trial, test the sentiment in the market, and consider its bloodline and who trained it.
A horse also might have had some bad luck in recent races – knocked slightly off course, for example – which means it should have shorter odds than just its recent places would suggest.
“Those things matter a lot,” says Beirne. “We are talking sometimes milliseconds between first and second.”
Bookies also keep a close eye on each others’ odds, because a large difference can indicate they have rated the horse incorrectly. And because they need to keep their prices competitive to attract punters.
Beirne says most of these elements of the trade would be familiar to his father and grandfather from when they were in the job.
“It hasn’t really changed – it’s just the technology to support it has,” he says.
How do punters affect odds?
A large amount of money flowing into a certain horse can be a wake-up call that a bookie has incorrectly priced a horse or missed a crucial piece of information.
A rush of bets for an international horse usually indicates the horse’s owner has decided to bring it across for the Cup – the news has leaked out.
“Money talks in this business,” Beirne says. “We’re not so egotistical to think our original prices are always right.
“Customers can be very smart – it’s our opinion versus theirs. We have to then play a bit of cat and mouse with them to work out what they’re doing.”
But even if he is convinced he has the price right, Beirne says he needs to exercise “risk management”, and adjust odds to stem the flow of cash towards a certain horse if his potential loss is getting too high if it wins.
Customers behave in such unique ways that it is sometimes even funny to us. But that psychological element is also a consideration …
Sometimes the odds are also influenced by punters’ quirky behaviour, inspired by superstition or bright-eyed optimism that they can back an unlikely winner.
“Customers behave in such unique ways that it is sometimes even funny to us,” he says. “But that psychological element is also a consideration into some of our pricing and how we manage risk.”
At 2018’s Melbourne Cup for instance, TAB’s biggest “stand” – or how much it could potentially lose – was on Who Shot Thebarman, a $41 chance that Beirne thinks once-a-year punters backed simply because they liked the name.
Likewise, horses with little chance of winning attract an irrational amount of money because punters want to bet small to win big.
For that reason, Michelle Payne’s fairytale Cup win on Prince of Penzance in 2015 was actually a poor result for Beirne, even though it was a very unlikely outcome. The horse was a rank outsider, priced at $101.
While a win for the favourite might have punters cheering and an underdog’s victory makes for a great story, they are the worst outcomes for Beirne.
He admits this makes him not a great person to watch sport with – always cheering for draws in soccer and unpopular mid-field runners in racing.
How do bookies calculate their profit?
Bookies don’t want to set odds that actually reflect what they think will occur – if they did that, they wouldn’t make any money.
Take, for example a two-horse race, where there is a 70 per cent chance of one horse winning and a 30 per cent chance the other will win.
If those punters back each horse in proportion to the bookmakers’ “implied probability”, the amount the bookie has to pay out in winnings will be equal to how much they receive in bets.
Instead, bookies tweak the odds in their favour so that the total implied probability for the entire field is more than 100 per cent, with every point over 100 per cent (the ‘over-round’) representing their theoretical profit margin (usually between 115 and 125).
The maths looks like this:
So, in a perfect scenario, if the over-round on a race is 115 per cent, the bookmaker expects to keep $15 of every $115 they receive in bets.
But that doesn’t mean they can’t lose. In practice, it is impossible to “balance” a book perfectly and stand to pay out the same amount for each horse. And the bookie doesn’t always get their predictions correct.
“That margin is only as good as us being able to understand that the true price was right in the first place,” says Beirne.
What are bookies thinking during a race?
Bookies will continue tweaking odds right up to when the race starts on Tuesday and the barriers open.
After that, Beirne says he’s “kind of like everyone else” with a dose of nerves and a flutter as the horses round the top of the straight.
Where he’s different from most racegoers, though, is where his attention goes – not to the front of the pack but around the course, scanning for the risks to the $115 million or so in Cup bets he will have accepted.
“My eyes are usually less so on the horses that are really good results for us,” he says.
“They’re always following the horses that are disasters for us.”
Note: A version of this story was first published in November 2019. The story was updated in November 2020 to reflect the horses and odds in the 2020 Melbourne Cup.
Business reporter at The Age and Sydney Morning Herald.