One of the most common questions amongst players is ‘have I been cheated?’ This is perfectly understandable. Many of the most popular games offered by casinos are ‘high variance’. This means that there are rare big payouts balanced out by large numbers of smaller losses. Some players are lucky and get more than their fair share of the big wins. Others are unlucky and see more than their fair share of losses. When you are the one experiencing the losses it is perfectly understandable that you may start to feel that the results cannot be natural and as such you must have been cheated. That is human nature.

The simple reality is that the human brain is ill-equipped to assess risk vs reward decisions relating to games of chance that have such a high level of variance. Part of what has made us the most successful species on the planet relates to the ability of our minds to identify patterns. We are instinctually programmed to look for patterns. But when dealing with outcomes that have such significant variance, this programming malfunctions. It looks for patterns in small data sets even though mathematically no conclusions can be drawn until we have huge data sets.

While it is understandable that when a player loses enough that they are emotionally distressed by the loss they start questioning whether the results of their interactions have been fair, not every case can be dismissed so easily. There have been identified cases of gaming softwares that have functioned in a non-random/unfair fashion. So what can we conclude about the fairness of online gambling games? This article will look to offer answers to that question.

The first matter we should address is the definition of “fair”.

In mathematical terms, ‘fair’ is used to describe an event where all outcomes have an equal probability of occurring.

Mathworld – “A game which is not biased toward any player.”

Wikipedia – Fair Coin – “A fair coin, when tossed, should have an equal chance of landing either side up”

Wikipedia – Dice – “Uniform fair dice” are dice where all faces have equal probability of outcome due to the symmetry of the die”

Under the above definition, there are next to no transactions in a casino that would be considered “fair”. Every single game, every bet offered, is offered with a built-in house edge meaning that the chances of winning are not equal for both parties.

It is also important to clarify the difference between the number of possible outcomes and the probability of those outcomes occurring. During my time as a teacher, I repeatedly encountered students who considered the lottery a “50/50” proposition. They came to this conclusion because there are two possible outcomes – you either win or you lose. For most people, and I have no doubt those students as they matured and gained more experience with the world, it is readily apparent that the lottery is not a 50/50 proposition as, while there are two potential outcomes, the chances of losing are significantly higher than that of winning.

So if casino games are not mathematically fair, and this should be reasonably accepted by all going into any gaming engagement, then we have to find a different definition of “fair”.

When considering gambling games – that as described above do not meet the standards for mathematical fairness – the standard set by most regulatory authorities in terms of fairness are three-fold:

i) Where the game replicates a physical device, the results of the game must conform to the probabilities that would naturally occur using this device. This means that if the game shows you playing cards (like Blackjack or Baccarat) the outcomes of the game have to be the same as playing with real cards. So the probability of drawing any given card is always 1 in 52. The same applies to dice games like Craps or Roulette wheels.

Below you can find the relevant term from the UKGC’s Remote Technical Standards:

‍RTS 7C. a
‍“Where a virtual event simulates a physical device, the theoretical game probabilities should match the probabilities of the real device (for example, the probability of a coin landing heads must be 0.5 every time the coin is tossed).”

ii) For games that are not based on a physical device, most commonly slot games, the operator has to provide clear information on the rules of the game, how you win and the cost of the game (house edge). This is covered in the following inclusion in the UKGC’s Remote Technical Standards:

‍RTS 3C
‍“For each virtual event, game (including bingo), or lottery, information that may reasonably be expected to enable the customer to make an informed decision about his or her chances of winning must be easily available before the customer commits to gamble. Information must include:
i. a description of the way the game works and the way in which winners are determined and prizes allocated
ii. house edge (or margin)
iii. the return to player (RTP) percentage oriv. the probability (likelihood) of winning events occurring.”

iii) Games cannot be “adaptive” to the bet you place. In simpler language, the game cannot change the likelihood of an outcome based on how you chose to bet. An example of an adaptive game would be a Roulette game where if the player bets on red, the game changes the probabilities of the outcomes so that black is now more likely to occur or vice versa. A game designed in this way would be considered ‘automatically adaptive’.

Below you can find the relevant term from the UKGC’s Remote Technical Standards:

‍RTS 7A. d
‍“Restricting adaptive behaviour prohibits automatic or manual interventions that change the probabilities of game outcomes occurring during play.”

The requirement also prohibits “manual interventions” – i.e. someone changing the results of a game to prevent a win.

Another example of a restricted game design would be a slots game that makes the top payouts less likely if the player increases their stake. It should however be acknowledged that slots games can change the likelihood of outcomes based on player actions if they are clear in the game rules regarding the consequences of the action. An example of this would be ‘Bonus Buy’ games, where players can choose to place a larger bet to trigger the bonus feature automatically rather than playing the main game hoping to trigger the bonus feature. Many of these types of games will provide the player better odds when placing these larger bets.

So when gambling we all agree that ‘fairness’ has a different definition, one that includes an acceptance that the games are in fact weighted in favor of the operator and instead focuses on the games not being deceptive or misleading in nature and precludes operators designing games or changing results to disadvantage the player in an adaptive fashion.

So, we know that reputable regulators will look to put in place standards that require gambling software developers to provide games that operate in a fair fashion. What else is required? A Random Number Generator (RNG).

An RNG is a piece of computer coding that, as the name suggests, gives a random result. It is the part of the game that introduces the element of chance. To again look to the UKGC’s Remote Technical Standards we can see that the requirements are as follows:

“Random number generation and game results must be ‘acceptably random’. Acceptably random here means that it is possible to demonstrate to a high degree of confidence that the output of the RNG, game, lottery and virtual event outcomes are random through, for example, statistical analysis using generally accepted tests and methods of analysis. Adaptive behaviour (ie a compensated game) is not permitted.

Where lotteries use the outcome of other events external to the lottery, to determine the result of the lottery the outcome must be unpredictable and externally verifiable.

‍RTS implementation guidance 7A
‍RNGs should be capable of demonstrating the following qualities:
i. the output from the RNG is uniformly distributed over the entire output range and game, lottery, or virtual event outcomes are distributed in accordance with the expected or theoretical probabilities
ii. the output of the RNG, game, lottery, and virtual event outcomes should be unpredictable, for example, for a software RNG it should be computationally infeasible to predict what the next number will be without complete knowledge of the algorithm and seed value
iii. random number generation does not reproduce the same output stream (cycle), and that two instances of a RNG do not produce the same stream as each other (synchronise)
iv. any forms of seeding and re-seeding used do not introduce predictability
v. any scaling applied to the output of the random number generator maintains the qualities as detailed”

Between the standards that regulators put in place restricting how software providers can design games, and those in place to ensure the correct function of the RNG, where adhered to these ensure that games operate in a fair fashion.

So we know that any serious regulatory agency will look to put in place robust standards, but how do they enforce them? The only way to tell whether an operator is compliant with these standards is to engage in some fairly high-end mathematical testing of the games on offer which likely exceeds the skill or capacity of the regulator to manage. As such, regulators require licensees to engage with independent Testing Labs.Testing Labs look to check that gambling games are offering results that are both random and in line with the rules as detailed for the game. They do this by sampling the data set of game results and testing the RNGs.While Testing Lab certificates are a strong indication of a fair game, there are limitations to the effectiveness of this strategy that we will discuss below.

All an RNG does is return a random number within a set range. The probability of any number occurring should be exactly the same as any other number occurring. Beyond creating the random number, RNGs do not do anything. And this can be abused.

Let us use the example of a game intended to return the same results as a 6 sided die. So every result should have a 1 in 6 chance of occurring. This game uses a perfect RNG with an output range of between 1 and 30 (so every number between 1 and 30 has an equal chance of occurring). A fair game designer could use the results as follows:

RNG DICE
1-5 = 1
6-10 = 2
11-15 = 3
16-20 = 4
21-25 = 5
26-30 = 6

Every result on the dice has an equal chance of being shown. The game accurately represents a 6 sided dice.However, if the game designer instead chose to use the RNG results as follows:

RNG DICE
1-15 = 1
16-18 = 2
19-21 = 3
22-24 = 4
25-27 = 5
28-30 = 6

The game no longer accurately represents a fair 6 sided dice. 1 will be rolled half of the time!

So despite the fact that the RNG is entirely fair and would pass the tests required of it, by misusing the results of the RNG the game can still be unfair.

The point in the above is not that RNGs certification is worthless. In fact, you cannot have a fair game – full stop – without a fair RNG. For that reason RNG certification is an essential factor in demonstrating that a game is fair. It is however only part of the demonstrating that a game runs fairly and by itself is meaningless.

The other part of proving that a game runs fairly – after you have demonstrated that the RNG is truly random – is testing the results that the game returns to ensure that they comply with the rules that the game advertises.

The problem with Game Sampling is that it is a LOT of work. It is time-consuming and takes a lot of computer resources. Given that – at a rough estimate – there are around 6k slots games online today (probably a lot more), all of which required initial testing for the software provider, and thousands of online operators who require ongoing testing to ensure that the games are continuing to function correctly, there is more work on this front than we could ever conceivably have the resources to manage.

As such Game Sampling for operators tends not to be ‘test all games every month’. The Testing Lab samples a selection of randomly selected games and assesses the results that these games have returned over that month to confirm that they fall within the mathematically expected range each month. This does leave gaps. It is not complete testing every month.

Not so fast! The purpose of this article is to take a broad and measured view of this topic. Not to give either false reassurances or alarmist rhetoric. So far we have looked at the technical measures and safeguards that have been put in place to ensure that the games being offered are fair. While it is reasonable to conclude that these measures are not absolutely foolproof, consideration does also need to be given to the incentives vs the rewards that would be at play if a cheating game was to be put out in the modern market.

We regularly see players submitting complaints about game fairness. These normally occur after an interruption in play (connection issue) has interrupted a round or after a player experiences a loss that is larger than they are comfortable with. In both situations some players will conclude that the casino has intervened or tampered with their game causing them to lose. This is a fundamental misconception.

Casino operators cannot control the games they offer. They buy games in from software providers. No software provider in a regulated market is going to give any of their casino clients sufficient access to their games to allow them to manipulate the results unfairly. Why? Because it is not just the casino operator who has to hold a license. The software provider has to be licensed as well.

Consider NetEnt, Playtech or Microgaming. These are huge organisations providing their games to thousands of different gambling operators. If they allowed casino operators the option to manipulate their games unfairly, sooner or later some greedy casino operator is going to get caught. When they get caught, it is not just the casino operator that stands to lose their license. The software provider would also face losing their license for breaching the technical standards requiring games to perform in a fair manner. The downside risk for the software provider in allowing a casino client to increase their profit by cheating would be the software provider potentially losing their license altogether and as such losing all of their business.

Logically it makes no sense what-so-ever for a software provider to consider letting any 3rd party have control of their games. That being the case, unless you are playing a game that has been developed by the casino operator – commonly referred to as proprietary software – it is extremely unlikely that any reputable software provider would allow the casino operator to interfere with the game in any way, shape or form.

Even if a software provider were to consider taking this risk, the level of conspiracy that it would take to achieve this would effectively ensure that they got caught. How many programmers, designers, managers or sales people in the software provider’s employ would need to be aware of the practice? How about at the casino end? How many employees would need to be aware of what was going on to allow the casino to effectively manipulate games? How many different casino clients would the software provider have to offer this functionality to in order to justify the risk to their own license?

The reality here is that it would only take one disgruntled employee to compile some evidence and then release it and huge companies would lose their licenses. And given the large number of people that would need to be in the loop, someone eventually doing this would be inevitable.

Not so fast! Yes, there are many very sound reasons that it is extremely unlikely you have been cheated as long as you have played with a well-regulated operator. However, the entire system only functions to prevent cheating casinos where there is a robust regulator acting to oversee the market. The incentive that prevents cheating casinos/software providers is that they could lose access to markets, face crippling reputational damage and potentially face prosecution if caught. Where no regulator exists or only a weak regulator licenses the casino/software provider, there is no barrier what-so-ever to the software providers or operators cheating.

So the key here is to stick with well-recognized regulators that carry clearly published licensing standards. The regulators that we feel reasonably fall into this category are as follows:

United Kingdom Gambling Commission

Malta Gambling Authority

Alderney Gambling Control Commission

Isle of Man Gambling Supervision Commission‍

Gibraltar Gambling Commission

Spelinspektionen

This list is not exhaustive. There are other local licensing systems (national licenses that solely apply to the function of the gambling operator within your country) that are likely credible as well. However, we would strongly discourage players from putting any faith or confidence in other international type licenses originating in small island states not listed above. Sadly questionable licensing bodies are some of the most prevalent “licenses” in the market.

Sadly, in the last few years, there has been a rapidly growing trend of entirely unlicensed operators popping-up and distributing fake versions of slot game titles from other popular providers.

These fake games look identical to games distributed by Net Entertainment, Playtech and Novomatic (and potentially many others) however careful examination of the games and in some cases direct confirmation from the software providers have demonstrated these to be fakes distributed by illegal developers.

These games are especially dangerous. They are dressed up to look like games from software providers that you know and can be reasonably concluded to be trustworthy. However, just because they look the same does not mean the manner in which these games function is anywhere even reasonably close to the genuine article. It would be entirely possible – and in our opinion highly likely – that the top payouts of these fake games have been disabled entirely.

As there are no regulators to require testing certificates, there are no testing certificates. There is no risk of losing licenses, access to markets or clients. The casinos providing these games know they are fakes. There is nothing to stop the groups distributing these games creating them specifically to cheat players. And why wouldn’t they? They are working with unlicensed operators who no doubt would love the additional means to increase their profit and reduce their risk of having big winners.

If you have played at a casino offering fake games you fall into a category of players that we would consider to be likely to have been cheated while playing online.

There is an easy way to protect yourself from exposure to these fake games and that is to stick to playing with operators licensed by the regulators mentioned above.

So, have any online casinos ever been caught cheating? Yes, and we will detail some of the instances below:

Amigotech – In 2015 a player approached us after having played 560 hands of 2 Ways Royal Video Poker and having never received a single hand better than two pair. When we reviewed the game logs we established that there is a greater chance of one person playing the UK lottery 3 weeks in a row and winning the jackpot each week than a player receiving no hand better than two pair in 560 hands of the game. This was a clear-cut case of a non-random game and it certainly appears likely that the game was programmed specifically to not deliver a payout of higher than two pairs.

During this analysis, it also appeared very likely that the Double feature offered at the end of this game was delivering non-random results strongly favoring the operator.

WinADay/Slotsland – In 2012 we identified problems with the multi-week Roulette games offered by WinADay and Slotsland casinos. Effectively when the player bet on a single number they would never win. It should be noted that when alerted to this issue the operator did refund the players we were aware of who were impacted.

SkillOnNet – In 2012, after reports of non-random behavior in SkillOnNet’s Double feature in their Video Poker games posted on a major forum, we conducted an investigation. Our results corroborated those of the forum members demonstrating that this game was indeed non-random and did not comply with the probabilities that would naturally occur based on the cards displayed. It should be noted that SkillOnNet have never acknowledged that there was an issue at all, but also that subsequent testing has never encountered a problem again. In our opinion, this strongly suggests that the game was adjusted after we brought the matter to the software provider’s attention. We cannot state with absolute certainty that this was an intentional rather than unintentional issue, but the manner in which it was addressed leaves us no confidence in operators run by this license or distributing SkillOnNet software. We do not currently consider it likely that there are ongoing issues with SkillOnNet’s games.

Absolute Poker/Ultimate Bet – In the second half of 2008 major headlines were created when a player – NioNio – at Ultimate Bet was mathematically shown to be winning at a rate many orders of magnitude above the most successful poker players to ever have played. It was later identified that this user has ‘super user’ privileges that allowed them to see other players’ hole cards during the game. Some parties have made strong implications that this user was not acting alone and that the ownership of Ultimate Bet were actively assisting this user. These two sites closed shortly after this information came to light.

Rival – In 2015 it came to light that Rival casino software has been sharing a Google Analytics code (how Google tracks traffic it sends to websites to report information about that traffic to the website owners) with CertainKey.com, a company offering Testing Lab certificates. This strongly suggests that there were ownership ties between Rival software and the Testing Lab that were supposedly engaging independent 3rd party testing for them. Add to this that there appeared to be ownership connections between the software provider and a number of casino operators running on their software platform and it is not hard to see the toxic potential that could play out with an incestuous relationship between software provider, testing lab, and casino. While this is NOT directly a confirmed cheating issue, the implications of these relationships are that cheating could easily have been facilitated.

Finsoft – In 2012 Finsoft was identified distributing card games that did not comply with the probabilities of real cards. Alongside this, the free versions of these games were demonstrated to deliver higher return rates to players than their real play counterparts. Many of the biggest providers of the day were carrying these games, but sadly at this point regulatory standards had not evolved to their current levels and the regulator responsible chose to ignore the issue.

Cheating games is not the only way that a casino can disadvantage players in order to increase their advantage. By playing around with the terms and conditions that they offer, operators can create situations that allow them to profit from term breaches. Below we will discuss some of the different ways this happens:

Max bet terms – Across the industry almost every casino operator will restrict the maximum bet size allowed while players are using bonuses. In fairness, it would not be financially sustainable to offer bonuses at all without these restrictions. However, these terms are just that, terms. There are very few operators taking steps to enforce the maximum allowed bet automatically.

Why is this a problem? When a player doesn’t read the terms properly, or forgets about this restriction and ends up betting more than the maximum allowed bet, if they lose their balance, the operator doesn’t spot the term violation and keeps the deposit. If they win, the operator spots the term violation and refuses to payout. This creates a win/push scenario for the operator and a lose/push scenario for the player.

Maximum bet violations account for a significant proportion of the complaints our service manages. We recently published an article on this matter highlighting that in this day and age there is no good reason that maximum bet terms should not be enforced automatically.

Inconsistent enforcement of terms – Less common, but another problematic practice that we have seen crop up repeatedly when managing complaints are situations where the operator will enforce a term only where it benefits them to do so. This is commonly combined with the maximum bet term discussed above, where multiple bonuses have been claimed in close succession, the maximum bet term has been violated during the play of each bonus but the operator only looks to enforce the maximum bet term on the final bonus to void play where the win occurred. All the previous bonuses, with an identical term violation, are left to be considered valid because the player lost their deposit.

Tricky terms – Some operators are still employing term construction in an effort to catch players out. The classic example of that is the maximum bet term that is worded “greater than or equal to”. This creates an issue as rather than having a clear maximum bet term of €5, as most of the industry do, with the inclusion of the words “or equal to” they have created a maximum allowed bet of €4.99. The extra €0.01 decrease in allowed bet makes no practical difference to the operator in terms of protecting against bonus abuse and a player scanning the bonus terms who already has an idea of what they are looking for can be easily caught out, seeing the figure of €5 and ending up getting their win voided because of a €0.01 overbet.

Max bet on non-bonus play – This practice is widespread with one particular group of rogue operators licensed in one of the aforementioned island states. They build in a max bet term to their general terms and condition. Why is this different to a max bet term in bonus terms and conditions? When you accept any promotional offer, you naturally expect to be subjected to some restrictions. It comes with the territory. When you play without bonuses you expect to be able to play without additional restrictions. There are wagering limits built into every game, but they don’t apply. You need to read the small print to know what the real wagering limits are.

To make matters worse, this group of operators has been known to not only seize the winnings, not just confiscate the wagered funds, but to also seize your entire balance. Imagine walking into a casino, going to a Blackjack table, buying in for €100, placing a €50 bet, winning your bet then the dealer refusing to pay your winnings, taking your bet as well and then reaching over and taking the other €50 in chips. They then tell you that you should have read the small print at the door. Does that sound acceptable to you?

License wide self-exclusions – Another fertile area for weakly licensed operators to exploit is gambling addiction. The scam works like this, you self-exclude at property A. A few months later you decide to sign-up to property B. You play and you are allowed to lose for as long as you like. The longer the better! If you ever win and try to withdraw however your win will get seized. Property A and Property B are owned by the same people. They don’t tell you about the relationship on their website. They don’t tell you about it when you self-exclude either. But the rationale behind this policy is that they are “protecting vulnerable players”. The reality is that the only thing they are protecting vulnerable players from is winning. They are more than happy for those vulnerable players to inflict further damage on themselves as long as the operator is profiting out of the transaction.

Vague restricted countries – Another weak license ploy is to build into terms and conditions a restriction on ‘any player from a country where online gambling is illegal’. They then go ahead and accept players from everywhere and anywhere, often building sites tailored to that specific country with geotargeting and language customization, and if a player from that country hits a big win they refuse to payout. It doesn’t matter that they have a large number of other players that they are happily taking deposits from located in the same country. The restriction is not really on your country. It is on being from your country and winning.

Prohibitions on self-exclude players – This particular policy involves the operator putting a general restriction on‘ gambling addicts’. The restriction has nothing to do with self-exclusion on their property. It simply means that if you have a big win, before you get paid the operator will go searching for you online. If they find anything you have posted that suggests that you have had addiction issues, or have self-excluded at any other operator ever, this will then become grounds for the operator to refuse payment. It is yet another shocking example of how operators in weakly licensed jurisdictions construct their policies not to protect vulnerable players, but specifically to victimize and exploit them.