- Joined
- Mar 25, 2012
- Location
- IOM
Whilst I applaud the bravery of questioning Enzo's maths, I do have to question the sanity of it.
No offence dunover
No offence dunover
You do not have permission to view link Log in or register now.
Whilst I applaud the bravery of questioning Enzo's maths, I do have to question the sanity of it.
No offence dunover
Therefore the player, in reality would go far in excess of 1253 spins.
Remember, they are random and need millions of plays to establish any meaningful figure (and TRTP) and to 'grab' a random group of 100 consecutive spins from this huge sequence is simply pointless.
Now the bag of balls. The odds are the same (0.7x) but would you prefer to be asked to pick 20 whites out of a bag of 100 containing 70 white, or a bag of 10,000 containing 7,000 white? (There is some Canadian mathematician who discussed and tried a similar proposition with some unexpected results if Enzo or anyone can recall his name, he also did work on lottery numbers at the start of a new lottery).
All the savvy people did exactly what I tried - picked a HV slot and played that. Like I said, if you crunch the figures this could potentially cost the casino a loss. But Enzo knows it won't...
Like I said in the OP, it's a very clever offer and respect to Enzo or whoever for coming up with it. One of those things that makes JPP unique.
Some would, some wont and the average would be 1253 spins. There's no difference between theory and reality ..
Whether or not a sample size is pointless depends on the event you're trying to measure. Determining sample sizes is no exact science. A general rule of thumb is to take 30 times the frequency of the rarest event you're trying to gather stats on. For example .. if you want to verify something that you expect to happen 1 in 1000 times .. then 30.000 is a good rule-of-thumb sample size (assuming uniform normal distribution).
Since this particular event (zero-wins vs non zero-wins) typically sits around the 0.5 range it is one of the few things of a slot you can measure with a small sample size.
Assuming the picked ball is put back each time that makes no difference at all. (it makes a difference if it wouldn't get put back.).
Ok, please don't put words in my mouth. You also seem to be missing the obvious fact that I'm completely unrelated to JPP .. and quite honestly couldn't care less if they made an EV+ bonus or a really bad EV- one. My only interest in this discussion is the math - since they were so easy to do in this particular case I thought I'd chime in and show how to get a feel for what to expect. If you read my post again you might notice that I've picked my samples simply at the extreme ends of the spectrum. My post was intended to explain .. not to convince anyone of anything.
It's curious how you repeatedly describe what I call a math explanation as leading people on. Formulas have no opinions. Maybe I should leave my garden of reason and put on a tin-foil hat to avoid leading people on ..
On the fact you don't rep JPP sorry my bad. I am merely pointing out a simple formula (which indeed it is) is not representative of the vagaries of random slots with numerous win ranges and features. It's almost like trying to write a summary of 20 different novels with one paragraph. Plus even the highest variance slots which are natural choices for a promo such as this have value pools which include a very large and thus frequent amount of small (< 1x stake) wins. The model we looked at has a simple (win=any)/(0) assumption. I don't feel there is any need for 'tin foil' references because somebody questions a simplification which when overlaid with years of slot play and untold amounts of spins on many different slots bears little relation to the outcome most will experience.
I say follow the money. If JPP had crunched info/report data they had on the various WMS slots and come to the conclusion that some games would effectively end up paying >100% to savvy players then the 'insurance' wouldn't be offered, especially without limits. That's my 'garden of reason'.....
Ok dunover, I'm going to follow up on this one last time just to try and help out. First off, there is no simplification going on. We're looking at the actual win frequency which just happens to be a number that is quite easy to reliably approximate on a small sample. So lets assume we find a 0.3 win frequency slot. My mathematical estimate of 1253 spins to see 20 consecutive zero spins is the best possible estimate. It's the estimate that will be correct the largest number of times. If you base your estimate on experience rather than math - and its not 1253 - then its going to be wrong more often. If a thousand people play that 0.3 WF slot untill they hit 20 consecutive losses and write down the number of spins it took them then 1253 is going to be the number that gets written down most often.
I'd happily bet you it isn't - that, based on your model. would be the average number of spins needed.
There's no such thing as experience on random. It's random. Past events do not influence future ones so it doesn't matter how many past events you saw. They still don't influence future ones. And on your second point, +EV bonuses (intentional and not intentional) definitely do exist. In fact I give those out all the time to our own regulars. So I'm afraid you can't start from the assumption that a bonus can't be EV and hence my math can't be the best possible representation of the real world - its simply a wrong assumption - and if someone tells you hey I got a slot here with a 3/10 win frequency .. whats the number of spins I'll need - in your experience - to hit 20 consecutive zero spins - then I'd definitely say 1253
Don't shoot the math guy - I'm just trying to help
Enzo
Appreciated but that's what I said if you think about it, the slots being random and the past performance not affecting future outcomes therefore any 'average' or sample figure produced from 100 spins is meaningless in context of the insurance offer.
dunover said:As for second bold quote in your reply about 'cannot be experience on random' well I find that ironic considering you suggested playing 100 rounds on a random game to establish a figure between 20-80 to base the maths on. So based on your logic we can never get a precise figure to perform the calculation on, yes? I see the dilemma though, because we need somewhere to start, after all. It is also worth pointing out that despite 'no experience in random' we have all accepted Jufo and kktmd's data on various slots based on previous experience of hundreds of thousands of spins which have produced 'facts' for example 'I'R gives a bonus round about every 124 spins' etc. etc.
P.S. I know I'm debating on the hallowed ground trodden by an accredited rep, but it's not sacrosanct - I'm just a person who doesn't take all I'm told at face value and unquestioningly when I think I see a flaw or contradiction to debate.
The only thing I can see from a player's perspective is the casino can sit and work out the math and know whether or not they're going to go broke with this offer because they're counting on a very large sample. I'm sure that number 20 wasn't picked out of a hat. It's obviously possible to hit 20 zeros in a row but if they've done their math correctly they might end up paying a few (un)lucky players quite early, some will never see the bonus and the rest will probably average out to a % bonus they probably would have offered anyway.
From a player's perspective it's no different than playing a slot game and trying to win. (I mean seriously, who sits around hoping the lose 20 times in a row?) Random is still random and if one player sits down with his deposit he could conceivably pull this off in his first 30 spins OR he could sit here spinning until the cows come home and never see 20 consecutive zeros.
If we simply followed the math when we gambled there would be no point in playing. Pick any game you like, the math says you're going to lose.
The only difference in this bonus and any other is the psychological head game it plays with people trying to achieve it. I'm not sure if I want to be getting excited after 15 or 16 losing spins and then feel let down because I finally managed to win something. The worse your game play the more you'll want to keep trying.
"I hit nothing 18 times. I was so close! I want to try again."
"19 losing spins! CRAP! I hit a free spin bonus sonofabitch!"
I think we're confusing the terms 'measuring' and 'experience' here. If you mean measuring when you say experience - then you're absolutely right. Measuring is absolutely a valid way to assess behavior. Personally my interpretation of experience is more that of a conclusion drawn on past events - I don't use that term to describe a conclusion made with formulas and measurements. Measurements are definitely a useful tool here .. especially if you have prior knowledge about what you are measuring. For example in this case we know its a slot game so we know the outcomes follow a uniform random distribution. One of the properties of such a distribution is that any subset of it is also a uniform random distribution. Because of that we know that the division between zero and non zero spins is a uniform random distribution.
From those distributions we know that if they have frequent events like a 1/2 or a 1/3 or a 1/4 event that is uniformly distributed, they can be reasonably measured on a sample sizes as small as 60, 90 or 120 samples. So 100 samples really is a pretty good sample (I btw suggested 200 or 300 to improve accuracy) to asses the win/lose frequency of a slot. If you were to assess the RTP it would be hugely insufficient since very rare events still constitute large amounts of the payout so you'd need to adjust your sample size to reliably include those rare events to.
I made a quick simulation that did 100 'spins' with a uniform distributed random 0.7 odd on zero : here's the outcome on how many of those 100 spins were zero
71,70,67,72,68,74,70,69,63,74,
error : 1,0,3,2,2,4,0,1,3,4
average error : (1+0+3+2+2+4+0+1+3+4)/10 = 2
as you can see, two samples are 4 off and the average error is 2 .. but in general I think its quite clear that 100 spins is enough to get a ballpark figure .. of course more is always better but for these types of events that stabilizes really fast .. you should expect to be only a couple off for these types of odds and samples of 100 spins.
Cheers,
Enzo
As they have now changed the terms and conditions twice and have now limited the amount of the bonus to £150 a day I think maybe they were making a loss?
thats very naughty to change terms in promo , border line rogue lines !!! caution here
They just put some restricted slots on and limited the maximum bonus - they could have pulled the whole promotion but didn't. Ladbrokes have done that before. I am sure they honoured those who had already played that day.
Jackpot Party are trying a ton of new promotions. Now that they have WMS competition they are getting innovative.
I really don't think they are behaving like a Rogue.
Ahh, the old "Technical problems" excuse.
I've been having "technical problems" since I stopped getting enough fiber. Still, I have nobody to blame but myself.