The game is not the same !

[Vman96]

What am I missing?   

You're not missing anything. Manipulating hot and cold streaks significantly will make the machine fall outside of the legally required randomness criterion fairly quickly. IGT could program machines that way if they wish, but I don't think it's in IGT's best interest to violate multiple state laws, even though about one-third of this forum tends to disagree with me. Also if it was actually legal to do this, you wouldn't keep seeing casinos downgrading paytables over time. Why downgrade the paytable if you are allowed to give a non-random machine?

[BillyJoe]

[QUOTE=williejoe]billyjoe, any and every number in a segment of the random number stream generated by a prng in a vp machine will translate into a card within the 52 card deck ( or it can be programmed to generated one of the possible 2,598.960 five card hands ). the prng stream is random in terms not being able to predict the next number ( or numbers ) that will be calculated by the algorithm from any previous numbers (cards dealt )----unless you know the algorithm and where exactly you are within that algorithms number steam-----if you do know these last two items, one can know what the next card(s) to be dealt will be. every number stream generated by the prng is extremly long ( 2 to the 2000 power----a truly huge number ) before it repeats-----but there are segments of this number stream that will yield hands that are hotter or colder than other segments of this stream. QUOTE]
OK - then if I understand what you are saying, as an over-simplified example, you believe there are segments in the number stream that would have no Aces, and segments where there are many Aces.
 
Is that it ?  

[williejoe]

it depends on how large a number segment you are looking at----if it is a very small segment ( relative to how fast the processor is running ) then there will be small segments that may not contain every possible card------but what is much more likely is that since the minimum processor speed ( cycles per seconds ) by regulation i think is 100 Hz----- it is much more reasonable to think of a given number segment in terms of how many aces ( or 2,3,4s ) will that number segment contain vs some other number segment. given the mathematical requirements of randomness, it would be very unlikely ( and in violation of navada laws ) that a number stream segment of any meaningful length ( again, compared to the processor speed ) would not contain all 52 cards----it is the fequency and distribution of these numbers within the segment that impacts hot and cold play. and if we examine a much larger number number segment-----again due to the requirements of randomess-----all cards/hands should have equal probability of being selected by the algorithm

[williejoe]

having written the above post regarding number stream segments and processor speed, it dawned on me that another way to "smooth out" the displayed results of any dprng algorithm ( like the mersenne twister ) is to just have a higher processor speed. although the navada regs state that the processsor speed must be a minimum of 100Hz , they can run much faster than this---thousands or ten of thousands of number ( card ) selections per second. also, we are now told that in most vp machines the processors now run continuously------ eg., they do not stop between deals, and futher more, five cards are selected at a time and not ten as in much older machines. the net impact of this is that the processor moves through a much larger segment of the algorithms number steam in a given time period ( and spends a much shorter time period in a segment that we would consider hot or cold )-----and the larger the number stream segment from which cards are selected, the less variation ( cold and hot segments ) we will experience. so for those of us trying to logically unravel the mysteries of what has change in vp besides the paytables----well for one thing it is probably the higher speed and the continuously running of the processors

[olds442jetaway]

720,000 hands in round numbers played so far this year at the casino. Only 3 Royals...not counting a couple on multi line games that were not on the deal line. Most hands were single line games. Again in round numbers there should have been 17 or 18 royals. This pattern began in late 2010 and if anything, the space between Royals has gotten wider!!! I know I am not the only one out there experiencing this....

[williejoe]

olds442jetway, i've been taking pics of all my 4As/wk and royals since 2006. i don't have good records on how much play i had in each year----but i believe that since i have shifted a lot of my craps play to vp play over the last several years----my guess is that i have played more vp hands in the last three to four years than previously------and yet i have seen a fall off of the number of big payout hands over the past few years. but i will have to say that i did have a spike in the 4As/wk due to a lucky few days in vegas in 2012 ( hit seven in three days play ). other than this anomoly, i have seen a dimunition of big hands over the past eight years.

[BillyJoe]

having written the above post regarding number stream segments and processor speed, it dawned on me that another way to "smooth out" the displayed results of any dprng algorithm ( like the mersenne twister ) is to just have a higher processor speed. although the navada regs state that the processsor speed must be a minimum of 100Hz , they can run much faster than this---thousands or ten of thousands of number ( card ) selections per second. also, we are now told that in most vp machines the processors now run continuously------ eg., they do not stop between deals, and futher more, five cards are selected at a time and not ten as in much older machines. the net impact of this is that the processor moves through a much larger segment of the algorithms number steam in a given time period ( and spends a much shorter time period in a segment that we would consider hot or cold )-----and the larger the number stream segment from which cards are selected, the less variation ( cold and hot segments ) we will experience. so for those of us trying to logically unravel the mysteries of what has change in vp besides the paytables----well for one thing it is probably the higher speed and the continuously running of the processors
I understand your point about the processor speed, but I can't relate it to a meaningful result. If I hold 4 to an RF in single line, I should have a 1 in 47 chance of catching that 5th RF card. PERIOD. That is, of course, assuming the card is there to be drawn, and there is an equal distribution of cards to choose from. Of course, just because I don't catch it on this opportunity, these are 'independant trials', so my chances on my next opportunity are the same - 1 in 47.  
 
Unless the machine analyzed my held cards, it would not know what card to withhold. Mixing the cards faster should not diminish my drawing odds.

[williejoe]

in the long term you are correct-----in the shorter term you are a little off------more accurately what will determine what card you get next will depend on the exact timing of when you hit the deal button and where you are within the prng number stream----when you hit the deal button, it is the dprng that will calculate ( select ) the output that will be translated into what card you will get. remember, that why it's tecnically called a deterministic pseudo random number generator------it mimics something that is very close to random and for all practical purposes is random for the long term-----but in reality this number stream can be printed out ( it's so long that would not be practical ), examined and analyzed----which is done by igt. and given you are in the same place in the algorithm's number stream and you hit the button at the same time you will always----100% of the time-------get the same card-----that is just the way an algrorithm works. and further----if you examined the a very, very long portion of the algorithm's number you would observed that it would be highly likely that from number stream segment to segment that the probablitiy of lets say for the sake of argument-----five hundred processor calculations----the one in forty-seven odds of hitting the royal will differ------but again looking a much larger portion of the number stream the odds will approach the matematical one in forty seven.

[BillyJoe]

in the long term you are correct-----in the shorter term you are a little off------more accurately what will determine what card you get next will depend on the exact timing of when you hit the deal button and where you are within the prng number stream----when you hit the deal button, it is the dprng that will calculate ( select ) the output that will be translated into what card you will get. remember, that why it's tecnically called a deterministic pseudo random number generator------it mimics something that is very close to random and for all practical purposes is random for the long term-----but in reality this number stream can be printed out ( it's so long that would not be practical ), examined and analyzed----which is done by igt. and given you are in the same place in the algorithm's number stream and you hit the button and the same time you will always----100% of the time-------get the same card-----that is just the way an algrorithm works.
And if I wait an extra second to hit the button, the number that would translate to the card that I desire will have flown by hundreds of times... right?
 
If that is true, then whether it flows by hundreds of times, or thousands of times on a faster processor, it should still be available for me to select. That asumes, of course, that ALL the numbers available do not translate to the 7 of hearts...

[williejoe]

well, the issue here is not wether a given number is available to be selected within the number stream-----it almost certainly is given any significant length of number segment------it's an issue of the "regional" probability that that number will be selected amongst all the other numbers within the number stream segment you find yourself in when you hit the deal button. and it is true that as time progresses the little 'ol dprng cranks out thousands of numbers ( representing cards )----and the card you want goes flying by with some regularity------but even with a processor running (or shuffeling ) thousands of cards per second you will remain in a realitive very small segment of the full number stream of the algrorithm before it starts to repeat----for example for the mersenne twister that non-repeating number stream is about ten with 600 zeros behind it----a very, very large number. so even the "short" number stream we are talking about is very large and even a fairly fast processor speed would spend some time in it. one thing we do know is that the navada regulations require that running the processor for 10,000 outcomes the chip must meet a 95% confidence limit using the chi-sqare test for randomness. this is encouraging in terms of assuring short term randomness at slower processor speeds----but continuous shuffling and faster processor speeds makes it easier to meet this randomness requirement since numbers are being selected over a much longer segment of the number stream-----and this is one reason why i personally believe that faster processor speeds and continuous shuffling "smooths" the output ( hands dealt ) of the processor. this doesn't make the game unfair or compromise randomness----but it does change the "personality" of the game and i think this is the change that many long-term, experienced vp players are saying they see.