Arbitrage: My $45k bet that couldn’t lose


For an arbitrage punter, it's essential that the numbers stack up... For an arbitrage punter, it's essential that the numbers stack up...

Once one of the biggest arbitrage punters in Australia, Luc Pettett - CEO of - shares his personal pro punting journey. From building an automated betting program, to (almost!) risk-free $45,000 wagers, to ultimately being shut down by the corporate bookmakers, Luc’s story provides a unique insight into the razor-thin opportunism of arbitrage betting.

Three years ago I was the biggest arbitrage punter in Australia, having built a ‘bot’ to exclusively place my bets.

At its peak, my automated computer program was in 24-hour operation, placing about 700 bets a day and using six servers hosted offshore.

And on most days - as is the nature of arbitrage betting - I could not lose on every single race or sporting contest that the bot chose to bet on.

My bot wouldn’t get drunk and reckless, bet with its heart or start chasing losses. It would simply do what it was told - all day every day - by honing in on any betting market it considered to be potentially profitable.

This is the beauty of a bot: entirely automated, largely risk-free, and capable of placing bets with any of the major Australian bookmakers or TABs in less than 1.2 seconds.

Turning a profit

Once my system was up and running, it didn’t take long for the profits to start trickling in (when you’re an arber, it never pours but it almost always rains), to the tune of $400 a day; $2800 a week.

"at one point I even had the ceo of a well-known Australian BookIE call me UP AND OFFER TO purchase the bot, just to get it out of 'enemy' hands."

Luc Pettett

I politely declined.

Not all of the bookmakers would have felt this way, however.

When you are arbing, you should theoretically be taking an equal edge off - or be equally in profit with - all bookmakers (assuming you are betting equally between them).

In practice however, this just does not come to pass.

After a while, I did in fact observe considerable imbalances in profitability across my various accounts. Overall I would be winning (I wouldn’t have bothered going to such trouble otherwise), but I was anything but a winner in the eyes of some bookmakers.

And they of course would be right: my betting account would be significantly in the black with some bookmakers, yet consistently in the red with others. Lose with one bookmaker; get that loss back plus collect a little more with the other.

What all this boils down to is that the traders at some particular bookmakers were framing markets that favoured my position - whether by making mistakes on a winning team or offering inflated odds on a horse - on a consistent basis.

With one bookmaker, therefore, I might be identified as a professional punter (and would have trouble keeping my account open as a consequence); yet with another, I was identified as a must-keep resource that bet big, but lost often.

The corporate bookmakers in the latter category hence saw me as a ‘valuable’ client and I’d end up receiving a barrage of VIP offerings: free flights, grand final tickets and an array of other incentives in order to retain my business. It was a strange time - here I was making a very solid living out of gambling, yet some in the industry were treating me like the proverbial ‘Golden goose’!

The bot bets big!

During one of my biggest weeks I recall I was in Melbourne for a VIP event. Black Caviar was due to race the following day and the bookies were heavily promoting it.

One major corporate bookmaker had invited me down to attend an AFL game and sorted me out with some accommodation at Crown Casino. We had a blast enjoying the footy and hospitality at the ‘G, following on with some drinks at a CBD bar.

The next morning I was taking things quite leisurely - as you do after a big night - until I was surprised to discover that while I was asleep, my betting bot had gone into overdrive. Unbeknownst to its slumbering creator, the bot had placed over $45,000 of wagers on Black Caviar, having capitalised on some especially competitive odds being offered up by one online bookie ($1.15).

Such a suddenly large bet would normally set off alarm bells in the conservative world of arbitrage, but upon further inspection, it became apparent that the bot had stayed true to its strict programming guidelines.

While it was ‘loading-up’ on Black Caviar, the emotionless, in this case wholly ‘un-Australian’ bot was also placing lay bets against the champion sprinter - as well as win bets on the horses that made up the rest of the field - eventually forming a scenario whereby I couldn’t possibly lose.

I would win that $45,000 back regardless of where Black Caviar finished, plus of course my tidy arb profit. Undoubtedly, I wanted the mighty mare to win. But in terms of my betting bank, it didn’t actually matter if the unthinkable happened and ‘Nelly’ tasted defeat for the first time.

(Without wanting to kill the suspense, Black Caviar won!)

The theory

In its simplest form, arbitrage betting involves backing multiple outcomes in an event and securing a profit regardless of the result. This is achieved by comparing betting odds amongst bookmakers and reacting to the top odds when they reach a certain point.

Any individual bookmaker always has a statistical advantage over you (on average it’s about a 113% return back to the bookmaker) but when you compare them against one-another, you can often find a scenario where the numbers are in your favour (eg. 99% return to the bookmaker).

Here is a basic scenario:

Manly $1.92
Cronulla $1.92
Return to the bookmaker 104%

And now let’s introduce some competitive odds from a few different bookmakers.

  Bookmaker A Bookmaker B Bookmaker C
Manly $1.85 $2.00 $1.95
Cronulla $2.05 $1.90 $1.75
Return to the bookmaker 103% 103% 108%

Finally, let’s just bring together the best price on offer, because it doesn’t make sense to take $1.85 with Bookmaker A when $2.00 is available with Bookmaker B.

  Best Odds
Manly $2.00 (Bookmaker B)
Cronulla $2.05 (Bookmaker A)
Return to the bookmaker 98%

So, we’ve found one! You will have a 2% edge here, regardless of the result of the game.

You could cheer for either team, it doesn’t matter. In fact, there’s no need to emotionally invest in the result of the game at all. Just check your account for the profits when the final siren sounds.

For the above scenario, my suggested bet would be as follows:

- $506 on the Manly at odds of $2.00 with Bookmaker B = collect $1,012 if Manly win.
- $494 on the Cronulla at odds of $2.05 with Bookmaker A = collect $1,012 if Cronulla win.

You’ve spent $1,000, so that’s a profit of $12, regardless of the result. The profit is small, right?

Well to put it into perspective: you’ve invested $1,000 and made back a risk free return of 1.2% in a couple of hours. The highest interest rate on offer right now with the banks is roughly 4% in an entire year. That’s a time difference well worth exploiting, especially when you consider the accumulative nature of arb betting as opposed to fixed interest on cash with a bank.

If you do this over and over again, growing your profits game after game, race after race (there is a race every few minutes) as the bot keeps increasing your stake, it adds up quicker than you would believe - especially when you are putting virtually zero labour and time into the entire process!

Programming my bot

Before you can spot any arbitrage opportunity, you need to make sure you are bringing in the correct odds from every bookmaker.

My system was built to 'scrape' bookmaker's websites, pulling in the fixed odds on practically any market I had programmed it to cover. It would even cover overnight UK races, NRL, AFL, NFL, Eastern European women’s handball - you get the drift.

Once the pricing was brought into the system, the bot would be configured to only fire off a bet once a certain threshold was reached.

I wouldn't place a bet if the percentage was a profit but that profit was too small, and often on a Saturday I would make this profit threshold a fair bit higher: there is no point in placing a bet to win $12, when you have $1,000 tied up for 6 hours before the result of the race is finished.

The bot would prioritise markets that were more immediate, because it knew that it would receive balance back quicker than from those markets that wouldn’t be settled in hours or days.

If my configured threshold (say, 2% profit) was met, the bot would then push through its suggested bet into a queue, ready to commence its bet placement phase.

Bet placement was lightning fast. It could sometimes place bets in 200 milliseconds.

To place bets, the bot was programmed to scrape the bookmaker website.

"The bot would mimic a login, build up a betslip and confirm the bet - reading all messages off the screen along the way - just to ensure the bet was successful. "

Luc Pettett

Believe it or not, in my early days, the bet placement wasn't automated. This meant that throughout the day, the system would actually make an alarm sound notifying me of an arbing opportunity, and informing me of which bets needed to be placed, and where.

My job was to race into my home office, place the bets manually and disengage the alarm! After the alarm interrupted a few quiet dinners with my wife - often for profits as skinny as $5 - it didn’t take me long to realise that my current method wasn't very sustainable.

The risks

Nothing - and I mean nothing - is entirely risk-free when it comes to punting. The liability in this situation though is pretty low, so if a system is built correctly, the risk is negligible.

Below are a few dangerous situations I encountered when building and utilising the bot. Some were very frustrating experiences (losing some serious cash), whilst others turned out for the best regardless (for example the team I was over-exposed on ended up winning anyway, so the outcome was still positive!).

  1. The bookmakers do have the right (according to their terms and conditions) to cancel your bet due to ‘incorrect odds’ being published. This can cause havoc (leaving you over-exposed on one team only).

  2. When a gun player is a late withdrawal in a footy game or a key runner is scratched at the barriers, it forces serious change in the betting markets. Some bookmakers get this information sooner than others, resulting in a very complex and confusing situation for your bot to digest and safely negotiate.

  3. You haven’t prepared for your betting balance to hit zero, meaning you can’t get one side of your bet on. This is the worst situation you could be in because you are just gambling on one side of an outcome. You may as well turn off your bot.

  4. You are out of action for a few days while you shuffle money between accounts (payment processing via POLi really helps in this situation). Oh, and don’t even think about using your credit card - you lose your edge completely if the bookie charges you fees!

  5. You get detected by a bookmaker as being an arber and they permanently close down your betting account.


Why did I stop?

Like all punting opportunities, it’s a moving target. Eventually my priorities changed, but more to the point - I was detected.

Like most websites, scraping the bookies is against their terms. Eventually, their technology became better and better at closing me down. On one occasion, I used my brother-in-law’s betting account and it was closed just 4 hours later. Hardly worth the $60 or so had I made in profit!

So you want to be an arber?

It’s possible to still 'get on' as an arber these days. It’s just harder. The bookmakers are getting more and more advanced at watching their books for this sort of automated activity.

On the other hand, individual arbitrage opportunities actually arise more often in the current punting landscape. This is because the corporate bookmakers and TABs are becoming far more competitive with their pricing, and odds comparison services are becoming the norm.

A bookmaker today can ‘fingerprint’ your computer, taking in up to 400 different sensors to ensure you are the same person (it’s not as simple as using an IP address anymore!). They can almost certainly detect when you are betting automatically (by monitoring time between bets), and they can detect when you are placing bets using a calculation (who would so frequently place bets the nature of $494?!).

"Arbitrage betting isn’t for the faint-hearted or impatient. "

Luc Pettett

I’ve got no doubt there would be possibly hundreds of Aussie punters today doing as I did, but they undoubtedly face numerous challenges.

My creation of an effective arbitrage system was by no means the result of sheer luck. Rather, it came about primarily due to two factors: meticulous honing of my program over many years, and a burning personal desire to beat the bookies!


Articles View Hits
Follow James Conway
@fithorses on Twitter


SMS Fitter Horses Pack

Accept the Terms and Conditions: