Imri Sofer

# The 4% | Why Sports Betting Is Broken And How To Fix It

Updated: Apr 21, 2022

Sports betting is a predatory industry designed for players to lose their money by charging exponentially growing fees.

Although not evident at first sight

**–**you*are*playing against the house.As a legitimate player, you end up paying for your own risk, other players' betting manipulations, and greedy profit margins.

Using cutting-edge technologies, ingenuity, and integrity, Betsquare has created the world's fairest sportsbook.

As a young adult, I used to run to the local kiosk and place $2 bets on a streak of outcomes. Getting them all right meant that I would multiply my money 10 times over! The thought of the potential win, even if only in the form of a number on a paper receipt, was exhilarating. Little did I know, the state lottery, which has a monopoly over gambling, charged a 25% commission on these bets.

### Sports betting = Roulette

At first glance, sports betting seems to be a fair competition where one gets to apply sports knowledge to pick the winning outcome for a chance to quickly multiply one’s money. In reality, this game is designed exactly __like roulette__.

Betting on red or black, most people understand that the odds are stacked against them – betting your money on red doubles your money if the ball lands on a red tile, and loses it if the ball lands on either a black tile or on the single green tile.

The gaming operator (or *the house*) always has the advantage of having slightly better odds than the player. The same principle, better known as *the house edge*, applies to sports betting as well.

### Margin - A Clever Way To Take Your Money

Let’s look at the simplest sports bet – picking a straight winner, Team A vs Team B (which I will refer to as **A vs B**). Assuming that both teams are completely equal in ability and play in a neutral venue, and that the game cannot end in a tie, each team has a 50% win probability, similar to a fair coin toss – heads or tails; Team A or Team B. This scenario is called *even money **—* a completely fair situation where both the player and the house have an equal opportunity to win. For every dollar that we’ll invest, we’ll either double our money or lose it all.

While seemingly fair, even money bets are not feasible in the real world. The operator has many big-budget expenses other than paying winning tickets, such as salaries, overhead, and above all - marketing. Finally, in order to have an incentive to operate the game at all, the operator has to see a clear path to profit and a safety net against a potential bankruptcy.

To mitigate these expenses and ensure profitability over the long term, the house must charge a commission for each bet made on its book – this concept is better known as the *margin*. If you’ve never seen such an item on your betting slips, you’re not alone. Sportsbooks are carefully hiding this crucial information within the odds that they are offering.

Most competitive sportsbooks are charging somewhere around 4% margin as a standard, not so bad considering how much the state lottery charged me for my bets all those years ago!
Here comes the first caveat about margins **–** it is taken out of both sides of the equation, whether you win or lose. But it's done in a surreptitious way **– **by lowering the odds for both sides by double the margin; in case you've picked the correct outcome, you'd end up paying the commission for both you and for the losing side.

Let's apply margins to our A vs B contest and assume a *stake *or a* wager **— *the amount we are willing to risk — of $100. Winning our bet will net us our initial investment ($100) plus the expected profit (~$92). On the other hand, losing the bet will cost us our initial investment ($100). We're risking $100 to win $92 on equally able teams!

The margin iscompoundedevery time you place a bet. Whenever you see this word you should either go all-in or run for the hills.

### It’s Only 4%, Who Cares?

Well it’s not, and you should care.

The 4% margin is *compounded* every time you place a bet. Whenever you see this word you should either go all-in or run for the hills. In this case, it’s the latter. When an interest or a commission compounds, it means that it grows exponentially instead of linearly.
Let’s calculate our final balance starting with a *bankroll* — our overall betting budget — of $100 after a series of 4 alternating winning and losing bets of $100 each. For the sake of simplicity we’ll round the output to the nearest integer.

```
Initial balance = $100
1) Win (+$92) = $192
2) Loss (-$100) = $92
3) Win (+$85) = $177
4) Loss (-$100) = $77
```

Even though we guessed as many correct scores as we did incorrectly, we lost **23%** of our bankroll in only 4 bets! Based on this logic, the more we bet, the more we lose, at an ever increasing rate. Compound interest - the silent killer.

### Know Your Math

This is the moment that your 8th grade math teacher has been waiting for all these years - the big “I told you so” - math *can* make you money.

Successful sportsbooks specialize in analyzing risk or more precisely an *expected value (EV) **—* how much they expect to win of a certain contest given a normal distribution of bets. At first glance, __the math__ behind calculating EV seems hairy:

```
100 - the arithmetic series of ((1/decimal odds)*100) of all possible outcomes.
```

Stripping it down, it’s a pretty simple concept. Combining the offered odds of all possible outcomes generates a smaller winnings pool than there should be - picking the wrong outcome, you still lose all of your money, but picking the right outcome you win less than you actually deserve. In the A vs B example, if you had split your $100 stake evenly across both possible outcomes, $50 on Team A winning and $50 on team B winning, you are guaranteed to lose the margin!

Bottom line, you end up paying a monstrous fee to cover for imprecise risk models, foul-play insurance, and your bookie’s expensive taste of wine.

### Paying For Someone Else’s Sins

We have established that the house always wins given that the bets are evenly split between outcomes. But there’s a far more common scenario: what happens when they’re not?

More often than not, the bets distribution will be skewed towards one of the outcomes. In that case, the house can either win big or lose big, and if there’s one thing that operators hate, it’s *variance* (or uncertainty). In statistics, however minuscule the probability of an outcome is, it might still occur. At the end of the day, operators have high fixed-cost obligations - having a long enough losing streak could mean a sudden bankruptcy.

With the emergence of real-time betting, decisions of accepting or rejecting bets must be done in a span of only a few seconds. Sports betting operations had to become more efficient at analyzing and controlling the variance. To do so, they came up with a pipeline and a set of tools to provide them with an additional advantage. In traditional sportsbooks, the decision-making process usually goes as follows.

The process begins with pattern recognition algorithms, designed to flag potentially losing transactions by taking into account the stake, the player’s reputation, and the contest’s risk coefficient (a globally broadcast NFL game is much less likely to be compromised than an obscure __Ukrainian table tennis__ match is).

The flagged transactions are then passed on to a team of dedicated sports experts (also known as *traders),* to be approved manually. While some transactions might be irrevocably denied, others are assigned revised, higher margins and sent back to users for approval.

In addition, traders assign flagged customers __derogatory marks__ that will permanently and negatively affect customers' odds or, in some cases, even completely limit them from playing.

Last, sportsbooks design their terms and conditions so that even when customers win, these wins will be strictly limited.

However elaborate the operator’s risk models and restriction schemes are, they are still vulnerable to manipulation that costs bookmakers — hold tight — north of a __trillion dollars a year__! There are many ways for players to gain an advantage over the house by using a cocktail of mathematics, statistics, high-frequency information trading, social engineering, and promo manipulation.

To cover for skewed betting distribution and bet manipulation, sportsbooks pad their margins, heavily. While most mainstream operators are able to predict the variance with a margin of error of ±0.5%, they keep charging customers puffed up margins. Tracing the missing pieces of the puzzle you’d find that 1% goes to a safety net fund and that the remaining 2.5% covers the operational costs and an extra wide margin of profit. Bottom line, you end up paying a monstrous fee to cover for imprecise risk models, foul-play insurance, and your bookie’s expensive taste in wine. In fact, as a casual player, betting with 4% margin over time gives you less win percentages than the number of fingers you use to scroll down this article.

For as long as sports betting operators define their business model in terms of **margin = risk + profit**, the rules of the game will remain the same - when the player loses money, the house wins money, and vice versa. House vs. player; eat or be eaten.

Same horse, different saddle... [the house] keeps charging the players a high fixed margin, masked as “transaction fees”.

### No Horse In The Race

In the past few decades new technologies have emerged in the industry in an attempt to shuffle things up. They all targeted variance as their main enemy. Neutralize risk - ensure profit.

One of the most hyped technologies is peer-to-peer sports betting. The idea was introduced to the market back in the early 2000’s by an innovative company called Betfair. Their concept was groundbreaking - treat sports betting contests like financial markets. Every contest has people willing to buy and sell bets at different prices across a *book* — a ledger that summarizes the supply and demand for the specific market. No more should the player battle an omnipotent house for a slim chance to make gains; one could now compete against one’s peers while the house only takes a fixed cut of the transactions; a *razor blade *approach.

While the idea seems like a silver bullet, it is fundamentally flawed. Ignoring the liquidity problem (where some markets don’t have enough supply and demand to offer you attractive *long* (buy) and *short* (sell) prices, and these gaps are filled by market makers with superior knowledge, tools, and funding), this solution merely metamorphoses sportsbooks from roulette game operators to poker game operators. Let me explain.

In poker, the house usually takes a fixed cut out of every hand; this commission is called a *rake*. The poker operator doesn’t really care who ends up winning the pot (not an entirely accurate statement, but that’s a subject for another article), but rather that it is able to cut the largest possible commission from the pot. In much the same way, Betfair doesn't really mind who ends up winning but rather how much money was bet on each competition.

Let’s return to our A vs B example. We’re smarter than cavemen now and understand how to judge our expected value better.. right? (repeat after me: low margins gooood, high margins baaaaad.) Playing on the messianic Betfair, which charges a 2.5% transaction fee, we —again — placed 4 bets of $100 each that resulted in the same chain of alternating wins and losses.

```
Initial Bankroll = $100
1) Win (+$95) = $195
2) Loss (-$100) = $95
3) Win (+$85) = $185
4) Loss (-$100) = $85
```

Does this chart look familiar? Same horse, different saddle. While the house eliminates its risk by discounting itself out of the equation, it keeps charging the players a high fixed margin, masked as “transaction fees”. Only a fraction of the savings goes back to the players. Not so fair after all...

As long as the risk is acceptable, we won’t charge a commission. As the risk increases, we’ll charge a gradually increasing commission.

### The Fairest Sportsbook Doesn’t Exis-

It seems as if we’re at an impasse. Perhaps sports betting is an inherently flawed game; maybe players are destined to fight the house in a battle to the death…

In order to break the existing paradigm, we need to find a win-win situation for both the house and the players: a feasible way to lower the margin to the minimum that will incentivize operators to accept bets while allowing the players a fair chance at winning.

When this problem came across my desk back in 2018, I was looking for a new fair-tech project to start in a predatory industry that could be disrupted with cutting-edge technology, ingenuity, and integrity. Sports betting seemed like the perfect candidate.

Shortly after that, __Betsquare - the smart sportsbook__, was born. Over the past two years, we have engineered a fair peer-to-peer sports betting game. We’ve designed our platform so that for each individual contest, the distribution of bets will converge into the **minimal margin relative to the expected risk**, offering players a fair chance to win while eliminating unnecessary risks and maintaining steady profitability. Let’s unpack this bombastic headline, step by step.

### Zero Margin Sports Betting

How can a sportsbook be the fairest in the world? First and foremost, offer the best odds in the world, or reciprocally, the lowest margins possible. Easier said than done.

Exploring potential ways of cutting margins to a minimum, we came up with a wacky idea - what if we completely got rid of margins - offering zero margin sports betting?! The infeasibility of even money bets indicated this would be impossible. Since we couldn't really eliminate most of the fixed costs, and didn’t plan on starting a nonprofit venture, we had to find a more sustainable solution.

Shortly after, we reached our eureka moment! Instead of charging fixed "one size fits all" margins like other sportsbooks, our platform could adjust *a dynamic margin* based on the contest's macro trend - so that aggregating all betting activity will eventually converge into a global margin that will reflect the platform's total perceived risk from this contest.

We’ll initiate the competition by establishing a *tissue price* — a fair price based on the true ability of the teams — and determine a *risk threshold *— the distance between the total amount to be bet on each possible outcome — that the platform can safely cover. As long as the betting distribution is within the acceptable risk threshold, players will be able to enjoy a free-to-play, zero margin zone. The farther the bet moves the dial away from the equilibrium point of acceptable risk, the more margin the bet will incur.

__Geek-to-Human translator says:__ we’re looking at each sport contest as a whole and adjusting the commissions based on the perceived risk. As long as the risk is acceptable, we won’t charge a commission. As the risk gradually increases, we’ll charge a gradually increasing commission.

While most players will enjoy 0% commission and some will pay more, they’ll all end up better off.

### Game Theory To The Rescue

To maintain such a system in balance, we have to incentivize players to bet on the opposite trend at any current moment. To solve this issue we employ an idea from game theory - players will always look for the best deal.

Whenever the bet distribution becomes increasingly skewed towards a specific outcome, the platform will present more attractive prices for other outcomes and vice versa - an incentive-based balancing system.

Let’s design a multiplayer game with 3 participants for our ol’ reliable A vs B contest. We’ll set up a risk threshold of $150.

```
1) Player 1 bets $100 on team A and enjoys a 0% margin on his entire bet. The risk variance is now $100 towards Team A.
2) Player 2 bets $100 on Team A as well and enjoys 0% margin on the first $50. The distance between the pots of each outcome is now exactly $150, which is equal to our maximum risk threshold. Every bet that goes towards team A will now accrue a higher margin. Hence, for the next $50, Player 2 will pay a 0.5% margin. The risk variance is now at $200.
3) Player 3 bets $100 on team B. This will bring the risk variance down to $100. Since the risk variance was leaning towards Team A before the bet was placed, player 3 will pay a flat 0% margin for her entire bet
```

### You Get What You Deserve

While establishing a free-for-all, zero commission sports betting is a compelling idea; it opens the platform for *arbitraging* — exploiting the shifting margins in order to make a profit on the expense of the rest of the players.

To solve that problem, we have introduced a reputation system. Using deep learning of players’ interactions over time, our platform assigns personalized risk coefficients to players. For most casual players, the risk coefficient will be insignificant and won’t affect their end-pricing. For the outliers who will be flagged by the system, the risk coefficient will add a cushion to their margin equal to the predicted level of risk associated with their profile.

While most players will enjoy 0% commission and some will pay more, they’ll all end up better off, since our platform always looks to charge the least margin possible in order to cover the risk.

### Show Me The Money!

In breaking the house-vs-player paradigm, we’ve successfully eliminated the risk factor and passed the savings on to the player. The last question that’s left open is: how are we planning to pay for our fixed costs and, heaven forbid, make a profit?

Our platform is designed to aggregate all risk instances to converge into the minimal margin possible that covers the predicted variance. On top of that, we are adding a 0.5% premium that should cover our fixed expenses and a fair profit margin.

### What’s Next?

Over the next few years we’ll gradually release more parts of this ecosystem, along with other cool technologies.

Our first product, __Robet__, was released at the beginning of 2020 and is live in more than 150 countries, serving tens of thousands of users. Powered by __Linearity²__, Robet has introduced a limited concept where players are able to bet with margins as low as 0.25%, only paying fixed exchange fees when they deposit and withdraw their funds. This pricing system shifts the commissions from exponential to linear - making it significantly fairer to the players.

## Comments