RTG - RTP Trial run for interested forum members

[binshakindown]

So with calculating RTP does it matter how many spins you actually do, even when the total amount won goes into your accrued balance and not back into your bank roll?

I do see the point about it not being an accurate representation of RTP, because of the possibility of a different server for real verses tourney modes, and the possible downplaying of actual ending accrued balances, among a few other things.

[Jasminebed]

I thought along the same lines as Chayton, but I forgot about remaining balances. Or rounds ending during free spins.

It may be true that tourney mode is not the same as real mode. Since RTG operators can change game payouts, it might be that all tourneys are at a set percentage for all operators, but individual casinos might have different RTPs on the game in real mode.

I think the Randoms somehow end up skewed on these too, compared with real play. I'm not just sure how, but I don't think most players make $100 spins as a matter of course. I've played late in games where randoms for games that usually go before 10K are over 40, or come in just after someone hit it and it's just barely over 1K.

I think it's more an exercise in futility than gathering any real data.

But if you decide to go for it, I'll certainly be interested in reading your results.

I do think that when you look at return versus start balance like tourney play, it does give an accurate picture of RTP, regardless of bet size. But any remaining balance in the start balance needs to be considered too.

[Simmo!]

Since RTG operators can change game payouts....

Just for clarification and historical accuracy on the above statement: the unofficial "official" line (!) is that a request needs to be put in by an operator to RTG with a justification to change to one of the 3 set percentage options offered within the software (91.5%, 95% or 97.5%) and it affects all Slot Games only.

I'm not sure how Tournament Mode is controlled, whether it ties in with "Real Play" or whether it has it's own settings.

So with calculating RTP does it matter how many spins you actually do, even when the total amount won goes into your accrued balance and not back into your bank roll?

I think - personally - to get anything like a true representation you would probably need 5 - 10 million spins.

[4 of a kind]

The only true way to figure RTP settings is to know the total exact amount the machine took in and returned. Plus most machines take varying denominations and RTP's could be different for each denomination. No one knows for sure how any of this actually works online.

With regulated land based casinos you could follow the RTP's by checking their monthly reports recorded by their gaming commissions like I posted here: http://www.casinomeister.com/forums/...tml#post442146

[quber]

You can't calculate the RTP with those 2 figures alone!
You need to know how much you wagered.

RTP = ((Amount Wagered) + Amount of gain or - Amount of loss)) / (Amount wagered) x 100

e.g. If you started with $100, played 500 spins @ $1 each and ended up with $75 (i.e. a $25 loss):
RTP = (500-25)/500 = 0.95 x 100 = 95%

In other words: You bet $500 and won back $475 = 95% RTP.

And like others have already said, to get any sort of accurate figure you do need to do 1,000s of spins.

KK

It can be done with those figures alone as all winnings go to a seperate pot and are not added to the balance. If you start with 5k and end up with zero then the total wager is 5k as none of the money from winnings is being wagered.

[KasinoKing]

It can be done with those figures alone as all winnings go to a seperate pot and are not added to the balance. If you start with 5k and end up with zero then the total wager is 5k as none of the money from winnings is being wagered.

Yes, that's true - if you lose (use) the entire starting balance in that way then that is the "amount wagered" in my calculation.
I didn't realise that's how those tourneys work!

KK

[pmutts]

Thing is, I have always wanted to beat the HOUSE. Blackjack is my game of passion (and hate). I have tried BJ tourney's and I just hate the whole concept. It's just not my thing. Now Slot Tournaments or VP tourneys I believe I could find enjoyable to an extent.

Fun yes, but I'm from the mindset that if you get a Royal Flush in a touney, it is a waste if all you are going to win is a couple of hundred bucks--max. Jasminebed got me into the Screenshot tournament last Saturday for Halloween at 3Dice. It was a blast. Plus, it kept me diverted at the Rangers blew their second shot at the series. Third times a charm--right? GO SAINTS! My heart is with the Saints, not the Cowboys. I grew up in Louisiana.

[dubleup]

So I remember learning a method in statistics which served to filter junk data from the data set in question. We started with typical distributions of data. Here, the data values will generally follow a bell curve pattern, whereby a majority of the data points are somewhat clumped together and within so many standard deviations of the average value. Data that falls outside a calculated number of standard deviations from the mean in either direction are called 'outliers' I believe. These are discarded to make the dataset more useful based on this particular method from statistics class.

So, you can find the mean value of all the tourney results, calculate the associated standard deviation, then apply the formula to calculate the greatest number of standard deviations from the mean for which a data point is still considered 'valid'. Continuing with this method (I wish I could remember the name)you would throw out those values that fall outside the calculated distance from the mean, and in either direction. The values that remain are your 'filtered' dataset... Phew!

For example, lets say average winnings is 1000 with standard deviation of 200; and you put these numbers into the formula and calculate the cutoff for valid data to be 1.5 standard deviatioins... In this case, all winnings data less than 1000-(200*1.5) = 700 would be thrown out, and all datapoints having winnings above 1300 would be thrown out.

Anyway, if you think about it, the values that get eliminated are either very high, which increases the likelihood that the players in this category were left with money in their remaining balance at tourneys end; or the score is so pitifully low that it is only right to throw it and that poor player with his few remaining scraps of dignity completely out of the calculations so that he may wallow in his self-pity in peace..LOL... just kiddin. To avoid skewing the data, the accepted formula (I think developed by Chebychev? or Poisson? ) should be used to ensure the most useful of data. This would require throwing out both high and low values that lie outside the calculated distance from the mean.

For all who are stressing accuracy accuracy accuracy.... welcome to the world of probability! He he... sure sounds funny when its put that way doesn't it? Often, the best we as humans can do for real-world data collection such as that proposed by binshaken; consists of employing methods to ensure we have REASONABLY accurate data, then make prudent determinations based upon that data. I can gaurantee there is no gaurantee of accuracy when it comes to data collection such as that proposed... it doesn't exist. The only way possible is to get the data from the one pulling the strings, and that apparently is not an option here. Solutions, we need solutions!

[binshakindown]

It's been awhile (since early Nov) but here are my results of the last 3 tourneys since then:
3512/5000=70.2%
877/2000= 43.8%
533/2000= 26.7%

What I get out of this is my RTP is looking sad. It also tells me to stick with the free tourney's for a bit.

[Nifty29]

It's been awhile (since early Nov) but here are my results of the last 3 tourneys since then:
3512/5000=70.2%
877/2000= 43.8%
533/2000= 26.7%

What I get out of this is my RTP is looking sad. It also tells me to stick with the free tourney's for a bit.

The problem with this data is that it doesn't provide the number of spins, and this is very important.

If you started with 5000 coins and were betting 100 a spin, that is only 50 spins which is waaaay to small to even begin to calculate the true RTP of the games concerned.

For instance, if you were in fact betting 100 per spin, a win of only 15xbet would have propelled you into the 100%+ RTP category which would shed a completely different...and equally flawed....light on the subject.