16 replies to this topic

[GaryGoldman74]

Hi all just very confused as to why the B3 machines are considered "random" when they still display a % payout on them?! In theory if they are truly random does this mean premises can actually lose cash if they were to pay big a lot of the time to the player "randomly" (very unlikely I understand but this is what confuses me big time about the whole "random") Or do they just work to the % shown on them? I come across loads of them stating "92 % payout" but then read "B3 machines are random" Sorry random post but really bugging me, surely if "random" then premises could in fact not make any financial gain if the machine randomly keeps paying (again I understand so unlikely). Many many thanks everyone who can help with this! Cheers

[fruitman69]

This has been asked and explained countless times.

 

Yes indeed a random machine can lose money over a shorter time but the % displayed is the % that the machine should theoretically hit over many thousands of plays.

 

 

Consider me throwing a die,   outcomes 1-6  randomly  yes?  assuming it don't have a bias and I throw it in a non bias way.

 

 

Say I charge you  £2  a throw  and use the paytable below.

 

1 = 0 - lose

2 = 0 - lose

3 = 0 - lose

4 = £2.00 - money back

5 = £3.30 - £1.30 profit

6 = £6 - £3 profit

 

Although this game would be random  I can declare that it does actually pay out at least 94%  average.

 

Eg for every £12 wagered the expected long term average return is £11.30  ( actual  94.17%  to two decimal places)

 

 

 

For a more volatile game you could use

 

1-5  LOSE

6 Pays  £11.30

 

Same %  but  totally different game volatility.

 

 

I hope that helps explain how a "random" game can still have a % payout.

Edited by fruitman69, 07 May 2015 - 12:01 PM.

[GaryGoldman74]

Cheers bud I think I understand haha. So let's say a £500 B3 has 1000 numbers, 1-999 no jackpot then of course the remaining number 1000 is jackpot, then are the "1000 numbers" started again, surely? So it could pay the jackpot on number 200 then no jackpot for 800 more numbers and then the numbers are started again? So there is only one number out of the 1000 that is the jackpot and once it's won the jackpot cannot be paid until all 1000 numbers have been used and started again? So even tho it is "random" the "random number generator has to pick the jackpot number once out of those 1000 numbers and thus the % is met? Sorry and thank you!

[nails]

i prefer the bingo balls in a bag explanation. lots of losing balls, lots of 50p win balls etc and 1 £500 ball. all in though their are hundreds and thousands of balls, so yes the jackpot is available every spin. to make it worse, barcrest machines are compensated, which basically means if you have a large win - it put a few thousand more balls in!

[GaryGoldman74]

i prefer the bingo balls in a bag explanation. lots of losing balls, lots of 50p win balls etc and 1 £500 ball. all in though their are hundreds and thousands of balls, so yes the jackpot is available every spin. to make it worse, barcrest machines are compensated, which basically means if you have a large win - it put a few thousand more balls in!

. I just don't understand why they state a % if the machine is "random" if it is "random" then the machine could in theory never ever meet its %? All seems a complete load of rubbish this "random", remove the "% payout" from a "random" machine, the two just do not go together. Only unless the "random number generator" removes the numbers each time giving a 1 in 1000 chance of the jackpot, sorry I just really can't understand.

[Matty.N]

. I just don't understand why they state a % if the machine is "random" if it is "random" then the machine could in theory never ever meet its %? All seems a complete load of rubbish this "random", remove the "% payout" from a "random" machine, the two just do not go together. Only unless the "random number generator" removes the numbers each time giving a 1 in 1000 chance of the jackpot, sorry I just really can't understand.

 

It works because like rolling the die, none of the outcomes are weighted and each press is considered a roll. So in theory it will meet that target. If you roll a die 60 times you stand a very chance of getting each number roughly 10 times meeting the 1/6 chance. Can you rig the roll of a die? (aside from weighted/damaged ones).

 

Bear in mind it will likely have 10s of thousands of combinations (probably more) so any proof will take hundreds of millions of spins.

[spa]

Just to add onto fruitman. If the machine was paying way above it could make the 6 sided dice 8 sided with 2 more loses and still be random and still be aiming to get back to 94%. Likewise if it was paying under, adding another number 5 and still be totally random and aiming for target %.

 

I swear this is how section 16s used to 'run'. Had a run on these many times and a random can't run, unless there are more winning outcomes than normal. Then it would indeed appear to run.

[fruitman69]

Hi Gary, like others I really cant see why you cant get your head round it.

 

With nails or my explanation its quite self explanatory 

 

With all randoms the more games played the less the variance will be on actual % experienced.

 

If it did use your method of having so many options and once its picked each one it starts again would indeed stick to % but would also become predictable in that unless the JP came out near the end of a cycle it could not come out again until at least the cycle is started again, so most the time once a JP as hit people would avoid.

 

Heres some more detail as on the gambling commissions website:-

 

 

Gaming machines (fruit machines, slot machines) offer different prize amounts depending on their category. 

 

All gaming machines are required to clearly display the percentage return-to-player figure (% RTP), or the odds of winning a prize, from use of the machine.

  

 It is important to realise however, that the % RTP is an average achieved over a significant number of game plays* and not each time the gaming machine is played. For example, if a gaming machine displays an 85% RTP, you should not expect to win an average of 85 pence for every £1 you stake during a playing session.

  

 *Average % RTP is generally measured over 10,000 or 100,000 games or greater for compensated (1) and random (2) machines respectively, dependent upon their category. 

 

Gaming machine (fruit machine, slot machine) minimum percentage payout 

 

There is no statutory minimum percentage payout for a gaming machine. The technical standards for legacy gaming machines however, do put a lower limit of 70% as a % RTP. 

 

(1) Compensated machines vary the chance of winning a prize as a result of the outcome from previous play. Where such a machine is below its target %RTP it may become more generous dependent upon design and vice versa, though the prize distribution is still determined by chance.

  

 (2) Random machines rely purely on statistical probabilities to achieve their target percentage return to player. The odds of achieving a win remain constant, and are not affected by previous wins or losses.

  

 Gaming machines must make information available about their category; % RTP; and whether they are compensated or random.

 

 

The biggest thing you haven't mentioned is as far as I know   a B3 cat  machine can be either  compensated OR random and you need to look at the info screen to see which it is.

 

So even tho its the same CAT can play very different.

Edited by fruitman69, 07 May 2015 - 04:18 PM.

[GaryGoldman74]

Hi Gary, like others I really cant see why you cant get your head round it.
 
With nails or my explanation its quite self explanatory 
 
With all randoms the more games played the less the variance will be on actual % experienced.
 
If it did use your method of having so many options and once its picked each one it starts again would indeed stick to % but would also become predictable in that unless the JP came out near the end of a cycle it could not come out again until at least the cycle is started again, so most the time once a JP as hit people would avoid.
 
Heres some more detail as on the gambling commissions website:-
 
 
Gaming machines (fruit machines, slot machines) offer different prize amounts depending on their category. 
 
All gaming machines are required to clearly display the percentage return-to-player figure (% RTP), or the odds of winning a prize, from use of the machine.
  
 It is important to realise however, that the % RTP is an average achieved over a significant number of game plays* and not each time the gaming machine is played. For example, if a gaming machine displays an 85% RTP, you should not expect to win an average of 85 pence for every £1 you stake during a playing session.
  
 *Average % RTP is generally measured over 10,000 or 100,000 games or greater for compensated (1) and random (2) machines respectively, dependent upon their category. 
 
Gaming machine (fruit machine, slot machine) minimum percentage payout 
 
There is no statutory minimum percentage payout for a gaming machine. The technical standards for legacy gaming machines however, do put a lower limit of 70% as a % RTP. 
 
(1) Compensated machines vary the chance of winning a prize as a result of the outcome from previous play. Where such a machine is below its target %RTP it may become more generous dependent upon design and vice versa, though the prize distribution is still determined by chance.
  
 (2) Random machines rely purely on statistical probabilities to achieve their target percentage return to player. The odds of achieving a win remain constant, and are not affected by previous wins or losses.
  
 Gaming machines must make information available about their category; % RTP; and whether they are compensated or random.
 
 
The biggest thing you haven't mentioned is as far as I know   a B3 cat  machine can be either  compensated OR random and you need to look at the info screen to see which it is.
 
So even tho its the same CAT can play very different.

I understand the compensated aspect but if the machine is totally random then let's say (just randomly thinking of values) the B3 takes £3000 for a £500 then obviously that's nowhere near %. Then perhaps it takes another £2000 for a £500, but I understand the whole "% is met over 1000s of spins" so surely the machine must have to go mental in the long run with multi £500s as it would be way under% from being "random" and the only way it could get to target % is by paying big wins in the very long run? So let's say a B3 has taken £10000 in the long run and if the "% is hit over 1000s of spins" the machine must go mental and pay off that 94%? Basically the only difference between B3 and AWP is AWP is always Influenced by previous play and in theory a non compensated pure random B3 could pay multi £500 if the % had not been hit for a very long time? Or what is the average a B3 takes before giving the £500 but again it's "random" ?! Sorry for this! Just like say a B3 takes £12000, when does it have to meet it's %? CONFUSED!

[ady]

i prefer the bingo balls in a bag explanation. lots of losing balls, lots of 50p win balls etc and 1 £500 ball. all in though their are hundreds and thousands of balls, so yes the jackpot is available every spin. to make it worse, barcrest machines are compensated, which basically means if you have a large win - it put a few thousand more balls in!

 

I really find this reply the best...NOT because it's the way I explain it, but it was what made me understand......

 

As I keep telling a mate of mine....it will make no odds you covering-up or not looking hahahaha I did that as a Nine-year-old.

[fruitman69]

I understand the compensated aspect but if the machine is totally random then let's say (just randomly thinking of values) the B3 takes £3000 for a £500 then obviously that's nowhere near %. Then perhaps it takes another £2000 for a £500, but I understand the whole "% is met over 1000s of spins" so surely the machine must have to go mental in the long run with multi £500s as it would be way under% from being "random" and the only way it could get to target % is by paying big wins in the very long run? So let's say a B3 has taken £10000 in the long run and if the "% is hit over 1000s of spins" the machine must go mental and pay off that 94%? Basically the only difference between B3 and AWP is AWP is always Influenced by previous play and in theory a non compensated pure random B3 could pay multi £500 if the % had not been hit for a very long time? Or what is the average a B3 takes before giving the £500 but again it's "random" ?! Sorry for this! Just like say a B3 takes £12000, when does it have to meet it's %? CONFUSED!

 

There is no way of calculating the statistical variance on any random slot without the actual par sheet, which is highly unlikely we will ever get.

 

You keep saying  IF it got way behind it would have to get mental to "catch" up  but even if its behind the chance of a JP is the same as what it was on a true random game, likewise it "could" still keep paying when i worked at gala many years ago we had a slot that was in minus 4 weeks in a row  but then was bad for 2 whole weeks!  People noticed it had had a lot in with little out so kept piling it in, yet they didnt know that it was effectively well above statistical variance and in true random fashion within a few weeks it had levelled back to its 94%

 

People think that randoms can't streak but they can its just more lucky and unlucky streaks and the big difference is its not a controlled streak like compensated  more a case of people hitting lucky times on the RNG.

 

However i have seen a random game  give  4 JPs in an afternoon once and it had 2 JPs the day before and despite people warning people that it had dropped multiple JP they still played and got lucky!

 

 

In answer to your last bit  £12,000 in is nothing on £2 play thats only 6000 spins  the gambling commissions website clearly states " Average % RTP is generally measured over 10,000 or 100,000 games or greater for compensated (1) and random (2) machines respectively, dependent upon their category. "    so your looking at  200k  worth of plays as a minimum.  but even after that i doubt they would be too concerned even if the %  was 5% either side of target.

 

There is a mathematical formula to work out something called the confidence factor which again varies due to the par sheet  im sure i read somewhere that a random machine can still be + or - 1%  on a million plays and thats considered normal variance.

 

However the smaller the JP the less it will stray.   Eg a £500  £2 play game is only a 250x stake which compared to USA randoms is a small win.

Edited by fruitman69, 07 May 2015 - 06:20 PM.

[stevedude2]

I understand the compensated aspect but if the machine is totally random then let's say (just randomly thinking of values) the B3 takes £3000 for a £500 then obviously that's nowhere near %. Then perhaps it takes another £2000 for a £500, but I understand the whole "% is met over 1000s of spins" so surely the machine must have to go mental in the long run with multi £500s as it would be way under% from being "random" and the only way it could get to target % is by paying big wins in the very long run? So let's say a B3 has taken £10000 in the long run and if the "% is hit over 1000s of spins" the machine must go mental and pay off that 94%? Basically the only difference between B3 and AWP is AWP is always Influenced by previous play and in theory a non compensated pure random B3 could pay multi £500 if the % had not been hit for a very long time? Or what is the average a B3 takes before giving the £500 but again it's "random" ?! Sorry for this! Just like say a B3 takes £12000, when does it have to meet it's %? CONFUSED!

The expected return to player percentage of a random game is calculated by adding up the result of each possible combination and then dividing it by the cost of playing each of those combinations.

 

Easy example is single-zero roulette.  If you bet £10 on every number it will cost you £370.  But when you get paid for hitting one of those numbers you'll get £360 back (£350 plus your original bet) because you get paid as if there are 36 numbers on the wheel, not 37.  This means you stand to lose 1/37 of your bet every time you play roulette, which is known as the house edge.  1 divided by 37 is 0.027, meaning the house edge on roulette is 2.7% and the expected return to player percentage is 97.3%.

 

The 'expected' bit is to highlight the fact that even though there is mathematical proof of the exact return of the game, there is nothing that governs how long it will take to meet that value, because the game of roulette is random.  The fewer the number of outcomes in a random game, the quicker it will meet its expected return.  It could overpay or underpay on its way to that value, but it will get there because you cannot argue with the maths and the fact that each game or spin is independent and not skewed by previous outcomes.

 

A random slot game will undoubtedly have a great many more outcomes than the 37 you see on roulette of course, meaning it will take longer to achieve its expected return.

[fruitman69]

 

I really find this reply the best...NOT because it's the way I explain it, but it was what made me understand......

 

As I keep telling a mate of mine....it will make no odds you covering-up or not looking hahahaha I did that as a Nine-year-old.

 

Not often i disagree with you ady but on this occasion i do.

 

Nails answer is good, however it relates to a compensated machine, but the OP was asking how random B3 work so i wouldn't have marked it as the "best" answer for that reason, but thats just my opinion.

[fruitman69]

The expected return to player percentage of a random game is calculated by adding up the result of each possible combination and then dividing it by the cost of playing each of those combinations.

 

Easy example is single-zero roulette.  If you bet £10 on every number it will cost you £370.  But when you get paid for hitting one of those numbers you'll get £360 back (£350 plus your original bet) because you get paid as if there are 36 numbers on the wheel, not 37.  This means you stand to lose 1/37 of your bet every time you play roulette, which is known as the house edge.  1 divided by 37 is 0.027, meaning the house edge on roulette is 2.7% and the expected return to player percentage is 97.3%.

 

The 'expected' bit is to highlight the fact that even though there is mathematical proof of the exact return of the game, there is nothing that governs how long it will take to meet that value, because the game of roulette is random.  The fewer the number of outcomes in a random game, the quicker it will meet its expected return.  It could overpay or underpay on its way to that value, but it will get there because you cannot argue with the maths and the fact that each game or spin is independent and not skewed by previous outcomes.

 

A random slot game will undoubtedly have a great many more outcomes than the 37 you see on roulette of course, meaning it will take longer to achieve its expected return.

 

 

I would consider that this is the best answer so far

[GaryGoldman74]

The expected return to player percentage of a random game is calculated by adding up the result of each possible combination and then dividing it by the cost of playing each of those combinations.
 
Easy example is single-zero roulette.  If you bet £10 on every number it will cost you £370.  But when you get paid for hitting one of those numbers you'll get £360 back (£350 plus your original bet) because you get paid as if there are 36 numbers on the wheel, not 37.  This means you stand to lose 1/37 of your bet every time you play roulette, which is known as the house edge.  1 divided by 37 is 0.027, meaning the house edge on roulette is 2.7% and the expected return to player percentage is 97.3%.
 
The 'expected' bit is to highlight the fact that even though there is mathematical proof of the exact return of the game, there is nothing that governs how long it will take to meet that value, because the game of roulette is random.  The fewer the number of outcomes in a random game, the quicker it will meet its expected return.  It could overpay or underpay on its way to that value, but it will get there because you cannot argue with the maths and the fact that each game or spin is independent and not skewed by previous outcomes.
 
A random slot game will undoubtedly have a great many more outcomes than the 37 you see on roulette of course, meaning it will take longer to achieve its expected return.

Chose this answer as the best, thanks big time bud! So basically the "94 or 92 %" stated on a RANDOM machine doesn't really mean anything at all? Even if a £500 has taken £5000 without the jackpot but a lucky player gambles £20 and wins the £500 that's the "%" hit?

[fruitman69]

Putting it very simply, the % indicated on the machine is the "expected" % over a very large number of plays.

 

In reality, An average player, on an average session will mostly likely be way above or way below the stated %

 

Also for example, you should not chase your losses on a random machine as the Jackpot has no extra chance of coming out than it did before you started even if you have lost.

 

 

These machines are more a case of player luck.

[theoak]

Chose this answer as the best, thanks big time bud! So basically the "94 or 92 %" stated on a RANDOM machine doesn't really mean anything at all? Even if a £500 has taken £5000 without the jackpot but a lucky player gambles £20 and wins the £500 that's the "%" hit?

What is the odds of a head in a toin coss? 50%, right? 

 

If you bet on heads every time, your average return is 50%.

 

If the coin has a (incredibly unlikely) streak of 100 heads in a row, would you then expect it to have 100 tails in quick succession to compensate for it and to meet its "payout percentage"? 

 

I would hope the answer is "no", becuase the 100 heads have literally no bearing on future results, just like with roulette.  This is how any system can show an expected payout percentage, whilst also being random, whilst ALSO having 1000s of statistically unlikely results while the actual percentage flutuates above and below the predicted 50% figure.