If you’ve ever been tempted to gamble and you’ve seen the way that casinos and bookmakers stack the odds against you. One of the most clear examples is roulette that has 36 black and red numbers as well as the green numbers 0 and (in the U.S.) 03. There are 38 options for you to choose from. If you are betting on black or red, the odds of choosing right are 18/38. Likewise, an acceptable payout for a stake of $1 is $2.111. The house, however, pays just $2 and retains the difference. So the house is guaranteed an income.
Similar biases are evident in the bookmakers’ odds for soccer, horse races, and all other sporting events. Bookmakers always make sure the odds are favorable to them. But determining the odds for these games are more difficult than for roulette, as the calculations are trickier.
And that raises a tantalizing possibility. Is it possible to come up with an improved method of calculating the odds and be able to beat bookies?
Today we get an answer thanks to the work of Lisandro Kaunitz from the University of Tokyo and a couple of other friends, who have found a way to continuously earn money through the betting market online for soccer.
However, their work has a serious restriction. Kaunitz and co. say that when the bookmakers realised of this achievement, they stopped the researchers from making further bets.
Gamblers have for a long time played with ways to beat odds, but success is rare. This is because bookmakers are able to come up with precise odds. They typically employ teams of statisticians who study historical data for a sport like soccer . They then design sophisticated models to determine the most appropriate odds for every game.
Kaunitz and co say that, so far as they are aware, no one has ever been able to defeat this system with better statistical algorithms.
Yet, despite this method of operation however, there’s a weakness in the method bookmakers operate. It has to do with the way they place bets in order to guard against the risk of big payouts.
For example, when two teams play a game of soccer, bookmakers such as 유로88 determine the odds of each team achieving winning, losing, or draw. There are times when large numbers of gamblers are betting on an outcome for reasons not directly related to the odds. That team might be more popular than expected, for example. In this case the bookmaker is scheduled to pay a huge amount if that event occurs.
Bookmakers are able to protect their bets by offering odds that are more favorable on the opposite outcome. This is how they are able to attract bets that compensate at least a portion of possible losses.
Kaunitz and co. say that this method also provides the possibility for anyone to spot it. The technique that researchers have perfected is an algorithm that can consistently identify odds favoring the punter rather than the bookie.
Their method is straightforward. They begin by thinking that bookies themselves are proficient in determining odds and the rates they provide reflect the probability of winning, losing or loss, as well as their own margin.
In that scenario, a good measure of these probabilities is to use a simple average of odds offered by all the bookies–a kind of wisdom shared by the crowd. This yields the average odds which Kaunitz and co say is a remarkably accurate reflection of the real odds.
It’s then a simple task to look over all the odds that are offered and find out the outliers. Kaunitz and co . then work to determine how good the outlying odds are. If they’re high enough that the bet will make sense, at a minimum in the long term.
This is exactly what Kaunitz and co . have created. They designed a Web crawler that collected odds offered by betting companies on soccer games around the world. They computed the odds average and uncovered any outliers and then determined if a bet would favor them or not.
Before committing any real money, the researchers checked the concept with 10 years of historical data on closing odds and the results of 479,440 soccer matches between 2005 to 2015. The model paid out 44 percent of the time and produced a return of 3.5 percent over the course of 10 years. “For an imaginary stake of $50 for each bet, this is an equivalent profit of $98,865 for 56,435 bets,” they say.
An important question is whether this could have been just luck. Could they have been lucky? The team then evaluated their results against two thousand simulations where they made bets on the identical games. In that scenario, the wagers were returned 39 percent of the times, yielding a return of -3.2 percent and that’s equivalent to a loss of $93,000.
The team was able to assess the probability that their first success was a fluke. “The possibility of achieving more than $98,865 for 56,435 wagers with a random bet strategy is less that one in a billion,” they state.
That was a good thing, because it gave Kaunitz and co good reason to believe their approach would be effective in real life, but there was an issue. Normal punters are not able to always place bets on odds of the game, which can vary significantly from the odds given prior to the game.
So Kaunitz and co decided to recreate this. “We have decided to run an actual simulation by placing bets with odds ranging from 1 to 5 minutes prior to the start of each match,” they say.
The way that odds change in the run-up to games is not publicly accessible, so the team created a bot that collected these odds from betting websites around the world from September of 2015 through February 2016. Then , they tested their algorithm in this data set.
The results were better. Their bets returned 47.6 percent the times. They also earned a 9.9 percentage return. “If each bet was $50, our plan would have generated $34,932 in profit over 6,994 bets” they say.
Incredibly, a random betting strategy using the same data resulted in a profit of 0.2 percent and an earnings of $825. It could be a result of the fierce competition among betting firms online that provide better odds to draw punters under an unusual loss-leader policy.
The team then tried an approach that uses a method known as “paper trading” using which, the team place fictitious bets using real-time data , instead of historical. This is vital because it allows them to verify whether the odds they quote are available through an on-line bookmaker.
In fact, they found that in 30 percent of times, the odds had changed by the time they attempted to determine the odds on the internet. In these instances, they discarded the bet.
However, the strategy was profitable. In the course of three months paper trading the bets returned with a profit of 5.5 percent, making $1,128.50 for 407 bets worth $50.
“At this point we decided to place bets with actual money.” Kaunitz and co. Kaunitz and Co.
Then they repeated the strategy over the course of five months, using identical methods, except that a human operator would actually make a bet of $50 online after analyzing the odds. In the five months the bets were paid back 47.2 percent of the time. Additionally, they earned $957.50 on 265 bets. That’s a staggering yield in the amount of 8.5 percent.
Eagle-eyed readers will notice that the number of bets placed was considerably lower than during the period of paper trading. “The reason for this is because we didn’t have a dedicated operator who bet on all the available opportunities 24/7 and consequently we missed many of the bets that appeared,” they say.
However, the lower number of bets did not matter. “Our paper-based trading and actual betting activity confirmed the profitability of the strategy” declare Kaunitz and co.
That’s a smart strategy and an interesting result. Kaunitz and co found an Achilles’ heel within the gambling industry and capitalized on it for their own gain.
However, their tale comes with a sting. “Although we played according to the rules of the betting industry but a few months later, we began to place bets through actual bookmakers started to significantly limit the amount of money we could deposit,” say the team.
The bookies usually limited the stakes they could bet or suggested a “manual examination” of the bet before taking it. In such a case it was impossible for the team to make their bets.
If the bookmakers had chosen the bets to question at random the outcome shouldn’t have had any effect on how profitable the strategy was. However, Kaunitz and colleagues argue that it’s unlikely, and the actions taken by the bookmakers could have had a significant impact on them. “Under the circumstances, we might not maintain your betting method,” they say.
Kaunitz and co are clearly unsatisfied: “The sports betting industry is free to advertise and provide odds to their clients, but those clients are expected to lose money and should they be successful they may be banned from betting.”
The team points out the fact that this type of activity could be considered illegal. “Advertising products or services with the intention not to sell them , as described, or advertising products or services with no intent to fulfill a demand that is reasonably anticipated but in order to induce the customer to purchase another product (a technique often referred to as ‘bait’ or ‘bait and switch’ marketing) is considered as false advertising and carries pecuniary sanctions in the U.K., Australia, and the United States of America,” they say.
And they call on governments to ensure that they properly regulate the gaming business and to avoid this kind of practice from happening again.
The success of this method is not yet clear. But the results are intriguing nonetheless.
Global Business Article Submission