Evolution's Monopoly - stats are off

[Stevej]

Evolution Gaming's Monopoly has a dice feature in the bonus. The stats are off, showing that the dice results are either pre-recorded or controlled some other way. Even though I enjoy playing few of their games, I don't like them calling themselves a live casino and claiming these are "live games".

Look at the image attached. The results are from the last month. I could see some volatility happening here but this is way too off - every single doubles are way lower than the other results. For those who does not know, the doubles will add you an extra dice throw in the bonus game.

This current bonus feature is rigged.

@Evolution Gaming Official 

 

[780xFTW]

other outcomes need to be divided by 2 to factor in the single dice results.

nonetheless the double results are too low and it for sure looks rigged cause its all of em

 

[Stevej]

other outcomes need to be divided by 2 to factor in the single dice results.

nonetheless the double results are too low and it for sure looks rigged cause its all of em

Yes, exactly. Doubles have 1/36 odds, other 2/36 = 1/18.

 

[dagreenblazer]

My personal opinion, i would just call that Variance.

1 month/7566 spins is a very small sample size when it comes to rolling 2 x 6-sided dice - even within that sample doubles are only running at just over 50% under frequency - this may look large on a small scale, but if you sized up to a billion rolls, then what currently looks a visually dramatic spike would soon flatten back out to average.

The other reason why i disagree Monopoly specifically is rigged is because extra rolls don't necessarily equal extra profit - it's perfectly possible to reach Park Lane & Mayfair within 4 normal rolls, so hence if it was rigged i would expect to see 11 & non-double 10 being restricted too, yet these 2 rolls happen the same as all other non-doubles.

Plus the main reason it's not rigged, is because they don't need to - the game design itself is such a marathon of different hoops to jump through before it ever pays big that casinos make money purely on people chasing a near impossible dream. The board layout we're having to roll the dice for is the biggest problem because so few squares actually pay decent; the actual dice themselves don't need to be rigged because there's so little to aim for anyway. 

.

 

[Stevej]

Plus the main reason it's not rigged, is because they don't need to..

I was waiting for this comment. 

But yeah anyway, how many months with these levels would you see this as acceptable variance? 

And you can also change the timeframe inside the month - in each timeframe these doubles are lower. You can see 1 hour, 3 hours, 6 hours etc. All doubles - not just one or two - but all of them are lower in each timeframe. Stats are from tracksino.

Just to clarify. Odds for each double is 1/36 = 2.8%

This month: 1x1 = 1.15%, 2x2 = 1.15%, 3x3 = 1.37%, 4x4 = 0.9%, 5x5 = 0.96%, 6x6 = 0.85%.

They are off, a lot! I understand that getting doubles won't give you huge benefit in profits, but that was not my point. I'm more interested if they have real dices there or not.

 

[IguanaBob]

Normally I wouldn't buy into this, but looking at that, all doubles are lower than 2.77%, (some a lot more) and all non-doubles are higher than 5.54%. That level of discrepancy after nearly 8000 rolls looks mighty suspicious.

 

[Stevej]

Normally I wouldn't buy into this, but looking at that, all doubles are lower than 2.77%, (some a lot more) and all non-doubles are higher than 5.54%. That level of discrepancy after nearly 8000 rolls looks mighty suspicious.

Yeah, I agree. And just to underline the fact, that if you try to seek any other timeframe inside that month (at least those timeframes available in tracksino) you can see that those levels stay quite stable in all timeframes, no big changes to up or down. There isn't any variance inside that month.

 

[Blacksheep]

I would guess the bonus game is a pre-recorded video. It could never be that off with real dices, especially when it's all the doubles across all timeframes. A tiny 'design fault' here by Evolution.

Regardless, nobody gives a shit and stupid people continue playing.

 

[IguanaBob]

There are 482 doubles out of 7566

Ignoring the fact that the results also scale to smaller values, just doing a binomial calculation  the odds of getting 482 or less out of 7566 when you have  a 1/6 chance is less likely than winning a 49 ball 6 choice lottery (14 million to one) 22 times in a row, buying only one ticket each time.

I remember when Monopoly was new and all the hosts were insisting it was live but would not be cajoled by the chat, (not me), to do something like get on camera and give a thumbs up whilst the dice roll. I never thought it was suspicious but I did wonder why they always refused as it seemed pretty tame and easy to do.

 

[dagreenblazer]

I don't want to sound like I'm defending Monopoly, because i fully agree it's 99% of the time a poor value game. I've played it for years, though recently i haven't bothered so much cause payouts been rubbish. The reason i come back to it though is because it's a fun, easy game plus i like how there's only 2 bonus sections to have to cover and how everyone gets paid the same, none of this pick your colour BS.

Firstly, i still believe 1 month sample size is way too small in terms of analysis.

Secondly, when dice are being rolled you can clearly see the presenter/studio behind the dice - admittedly it's a small little window and often presenter doesn't move across it, but it is always 100% live - the dice are never pre-recorded. 

With regards to dice being fixed to not roll doubles, do you not realise the extra work that would be required to get 2 live dice not to roll doubles?? If they wanted to fix dice, why even have them live, why not just have digital dice they can easy manipulate?? The further you go down the rabbit hole of magnets, brakes & motors, etc the further you get from the actual truth. 

The real truth is as i said before - to hit big in Monopoly, the game design and subsequent mathematical hoops you need to jump through mean that achieving a big win is an extremely long sequence of separate events that all have to happen in a specific order;

For example.......1 in 27 odds of Chance lands, Chance does whatever odds of revealing 10x Multi, Wheel respin lands 1 in 54 odds of 4 rolls, all 4 dice then roll the correct sequence of numbers to both avoid tax & jail AND if necessary roll doubles to reach the 2 out 40 possible squares which pay big (Park Lane & Mayfair). The slightest deviation at any point and big win is gone.

The steps you have to complete to hit a big win in Monopoly are HUGE!!

Casinos are more than happy for you to believe in fixing dice, motors & brakes because then it distracts you from both the fact you're losing full stop AND from the real reason of why you're losing.

Scams, fixes & cheating, etc get people emotional and emotional people spend more & chase more losses, behaviours that casinos love. No-one really gets emotional over plain old boring Maths though, yet the Maths of the poor value game design is 100% how they steal your money and is where the real scam actually lives.

 

[IguanaBob]

Firstly, i still believe 1 month sample size is way too small in terms of analysis.

Assuming the values in the OP are correct how do you explain the binomial analysis in my last post? Getting that many doubles or less is less likely than winning the original National Lottery setup 22 times on the trot. That is not a small sample size, that is virtually impossible. I haven't verified the values in the OP but if they're true then there is something very wrong.

 

[IguanaBob]

Can anyone verify that the tracksino values are in fact accurate? The most logical explanation would be that they are not, it makes more sense than Evolution deliberately fudging the dice rolls. If the tracksino values are correct though there's no way that could happen by random chance.

 

[Stevej]

Can anyone verify that the tracksino values are in fact accurate? The most logical explanation would be that they are not, it makes more sense than Evolution deliberately fudging the dice rolls. If the tracksino values are correct though there's no way that could happen by random chance.

I don't know if they are accurate, but I don't see any reason they would not be. Even if they "missed" some dice throws, why does it only reflect to those doubles being lower? Maybe tracksino wants to manipulate the results, hehe.  Just kidding. 

Like you explained in your previous post, this scenario is impossible, and like today if you check last 24 hours you can see those doubles are again way too low. Whatever timeframe you look at in the last month, they are always lower. I'm sure this will keep happening every day until they notice this issue.

I don't know what's the "evolution magic" in this bonus game, but it's not real analog dices. Probably it's still legal, though. One thing is sure, this (bonus) game is rigged, unfortunately.

 

[dagreenblazer]

Assuming the values in the OP are correct how do you explain the binomial analysis in my last post? Getting that many doubles or less is less likely than winning the original National Lottery setup 22 times on the trot. That is not a small sample size, that is virtually impossible. I haven't verified the values in the OP but if they're true then there is something very wrong.

To be honest, i don't even know how you've extrapolated that answer - the only thing i can think is do you realise those probability results are displayed with Negative Powers?? It's 10 to the Minus Power of 160. 

Even disregarding sample size argument and everything else, to believe that 482 doubles out of 7566 rolls is so far under frequency of Expected Values that it's comparable to winning the lottery 22 times in a row is just ludicrous.

 

[Stevej]

Firstly, i still believe 1 month sample size is way too small in terms of analysis.

Secondly, when dice are being rolled you can clearly see the presenter/studio behind the dice - admittedly it's a small little window and often presenter doesn't move across it, but it is always 100% live - the dice are never pre-recorded. 

I don't think you understand maths so well, and not being really precise about the details.

Have you ever seen any transparent video running and people "behind that video"?  Although, I don't know if it's a video, but that's technically possible.

 

[IguanaBob]

Even disregarding sample size argument and everything else, to believe that 482 doubles out of 7566 rolls is so far under frequency of Expected Values that it's comparable to winning the lottery 22 times in a row is just ludicrous

No it's simple mathematics. 1/6 is a very low variance event. After 7500 attempts the value should be close to 1259. The sheer magnitude of the gulf between 1259 and 482 after so many attempts can be calculated as a decimal with 159 trailing 0s. That is just how far the gulf is. By contrast a 14 million to 1 event  only has 6 trailing 0s

There are only 2 logical explanations

Tracksino is getting it wrong

or

Evolution are fudging the rolls

The other explanation: Tracksino are right and Evolution are fair, is so beyond the realms of likelihood that it can be ignored.

 

[dagreenblazer]

I don't think you understand maths so well, and not being really precise about the details.

Have you ever seen any transparent video running and people "behind that video"?  Although, I don't know if it's a video, but that's technically possible.

If you think 1 month sample size of an independent value that has already been running for 5 years or so is a quality sample size then crack on.

If you think modelling that 1 months results post hypothese and being shocked it doesn't correlate with Expected Values which are based against Infinity is a sound platform to judge whether a game operator is cheating, then again feel free to believe.

You guys clearly only think we lose because game operators fix the games & it must be a conspiracy - you talk like if only the cheating stopped then all players could suddenly be winners, and obviously these 95% RTP games would definitely pay 200% if only the fixing ended lol!

It's just sad to see - it's like falling for the distraction technique while the pickpockets slipping your wallet away. 

No it's simple mathematics. 1/6 is a very low variance event. After 7500 attempts the value should be close to 1259. The sheer magnitude of the gulf after so many attempts can be calculated as a decimal with 156 trailing 0s. That is just how far the gulf is. By contrast a 14 million to 1 event  only has 6 trailing 0s

Bro, make the step up from simple mathematics and google Negative Powers - those zeros aren't trailing mate......

 

[Stevej]

You can calculate the scenario via python script just simply:

from scipy.stats import binom

# Parameters
n = 7515  # number of trials
k = 479  # number of successes
p = 1/6  # probability of success on a single trial

# Calculate cumulative probability
cumulative_prob = binom.cdf(k, n, p)

print("The cumulative probability of getting 479 or fewer doubles in 7,515 rolls is:", cumulative_prob)
 

Result:
The cumulative probability of getting 479 or fewer doubles in 7,515 rolls is: 3.489497726420227e-158

And just to remind that if you dive inside that 7515 trials and create sub-sets of it, the doubles seem to be lower in each sub-set

I've checked these: 117 / 1,892 (6.18%),  43 / 809 (5.32%), 19 / 265 (7.17%), 11 / 133 (8.27%), 4 / 74 (5.41%)

Last month

 

[IguanaBob]

Bro, make the step up from simple mathematics and google Negative Powers - those zeros aren't trailing mate......

What do you think *10^-159 means? I'll tell you what it means, it means 0 followed by a decimal then 158 zeros then the vaiue

Look mate, I'm not a conspiracy theorist but the math don't lie. Either the tracksino data is wrong or the Evolution rolls are fudged. 7500 is a huge sample size when talking about a 1/6 chance. Go to google, put in "binomial calculator", take your pick of which one you want to use  and put in 7566, 482, 1/6 in the relevant fields.

I'm still thinking the problem is probably with tracksino, but the idea that the listed values don't represent a hyperunrealistic gulf is ludicrous.