Thread Rating:
zeros, is a 50/50 bet. If you ignore the ties in
baccarat, is banker/player a 50/50 bet, or is
it slightly skewed to banker?
On an RNG, like Bodog has, is the algorithm
on bac set to account for the slight banker
edge, if there is one?
My understanding is that the Draw/Stand Rules gives the Banker hand it's advantage over the Player hand.Quote: EvenBobOn the Wiz of Odds site, red/black, ignoring the
zeros, is a 50/50 bet. If you ignore the ties in
baccarat, is banker/player a 50/50 bet, or is
it slightly skewed to banker?
see Google Book preview
Stewart N. Ethier
The Doctrine of Chances
2010
Part 2 Applications
Chapter 19 Baccarat: goes over the rules and shows the whys of each.
The algorithm is to shuffle the decks as I understand it. The RNG is used to give random numbers to each card.Quote: EvenBobOn an RNG, like Bodog has, is the algorithm
on bac set to account for the slight banker
edge, if there is one?
Then the game is played out just as in a B&M casino.
Some online casinos shuffle after every hand because it is so very easy to do.
Fisher–Yates shuffle
Quote: guido111
The algorithm is to shuffle the decks as I understand it. The RNG is used to give random numbers to each card.
So the outcome would be slightly different than
the RNG for red/black in roulette. Would there
be a difference that could be spotted.
The RNG for Roulette has to only choose 1 out of 37 or 38 numbers. Simple stuff. Many good ways to accomplish the task.Quote: EvenBobSo the outcome would be slightly different than
the RNG for red/black in roulette.
8 deck Baccarat it has to select 416 random numbers. Simple stuff.
I guess it could be viewed as a 38 card deck for Roulette, then the deck gets shuffled and the first card is the spin... but that really is not needed. Too much work. The RNG handles everything for Roulette.
It is the software's code that determines what then happens to the number that the RNG selected.
I am not a top notch programmer by any means but I can easily program any gambling software to do exactly what I want.
If a Roulette player bets 37 numbers out of 38, I could program the code to ALWAYS make the one number not chosen come up without doing anything to the RNG. Very Simple stuff. But the player would know something was really fishy. From a coding standpoint, very simple to do and again, has nothing to do with the RNG.
Video Poker machines were cheating way back in the 80s. Had nothing to do with the RNG, it was all after that fact.
American Coin video poker cheats makes some good reading.
I do not think so.Quote: EvenBobWould there be a difference that could be spotted.
Quote: guido1118 deck Baccarat it has to select 416 random numbers. Simple stuff.
But with a real deck, they don't shuffle after every
hand. They deal down the shoe, which has to have
a different outcome than shuffling after every hand.
Yeah, things are different if you shuffle every hand, but things are different if, in a B&M casino, you track the cards.
I.E. If you notice that nearly all of the 4's have been dealt, then it certainly makes long odds for a 4 to appear.
Does that change the odds or edge of the game? No.
Quote: EvenBobBut with a real deck, they don't shuffle after every
hand. They deal down the shoe, which has to have
a different outcome than shuffling after every hand.
I think you misunderstand what an RNG is used for. The RNG in a card game is used to shuffle a virtual deck of cards. Each card in the 8-deck shoe is put into a random position such that each card appears in every location with identical probability (1/416). Once that's done, the RNG is out of the picture. It does not "select an outcome" or "determine a winner". The shuffling algorithm for an 8-deck shoe for baccarat is identical to an 8-deck shoe for blackjack -- just like a physical card shuffler or human shuffling method. The shuffler (RNG or human) doesn't care what the game is. The game rules, not the RNG, determine the probability of winning.
In roulette, the RNG is used to pick which value from 0, 00, 1..36 will appear, again with identical probability (this time 1/38). The question of red or black is determined afterwards, not by the RNG, but by the game rules that say "7 is red, 13 is black, 0 is green", etc. Similarly, for dice, the RNG just picks two values between 1 and 6, and the rules of the game say "if the total is 12, anyone with money on the 12 bet gets 30-to-1."
Quote: MathExtremistI think you misunderstand what an RNG is used for. The RNG in a card game is used to shuffle a virtual deck of cards.
I was playing in fun mode on Dublin last night
at the live bac table. The dealer shuffled the
biggest banker trend I've ever seen. 52
bankers and 14 player hands. The table was
soon crammed with players, everybody betting
banker and winning big stacks of chips. It
went on till the end of the shoe. You'll never
see this from an RNG that shuffles after every
hand.
Pardon me for asking, but it just does not select one of 36 numbers? Isn't there a formula that selects a number from among billions.
Then applies some mathematical formula. dividing, remainders etc, to change the big number to one of 38 simple numbers? Or am I wrong again ??
Quote: EvenBobI was playing in fun mode on Dublin last night
at the live bac table. The dealer shuffled the
biggest banker trend I've ever seen. 52
bankers and 14 player hands. The table was
soon crammed with players, everybody betting
banker and winning big stacks of chips. It
went on till the end of the shoe. You'll never
see this from an RNG that shuffles after every
hand.
My understanding in Baccarat is that there is very, very low correlation between banker/player results and the composition of what is left of the deck. That may be a misunderstanding though.... and the fact it can change (even minutely) is a difference between a CSM RNG and a full deck RNG (that sort of correlation is why folks count blackjack after all).
The chance of a 52/14 split over 66 trials which split 50/50 is just a little over 1 in a million (I'm ignoring ties). It'll be less for a Banker win.
Quote: thecesspit
The chance of a 52/14 split over 66 trials which split 50/50 is just a little over 1 in a million (I'm ignoring ties). It'll be less for a Banker win.
There were ties, it was probably 70 trials. Still, it was
quite a shoe. Bac players live for trends. And because
you can piggyback bets in bac, there was a row of
players standing and betting in the sitting players
boxes.
14 or less Players is about 1 in 68,000.
(I'm using a binomial function to work these out).
Quite the time for all concerned :)
Sooner or later, you will.Quote: EvenBobYou'll never
see this from an RNG that shuffles after every
hand.
Just as, sooner or later, you finally saw it happen with a live shoe.
Chalk it up to "Shit happens."
Actually, the RND function selects a fraction in the range 0<=x<1 (or it might be 0<x<=1). How it exactly does that differs depending on the software/hardware being used.Quote: buzzpaffIn roulette, the RNG is used to pick which value from 0, 00, 1..36 will appear, again with identical probability (this time 1/38).
Pardon me for asking, but it just does not select one of 36 numbers? Isn't there a formula that selects a number from among billions.
Then applies some mathematical formula. dividing, remainders etc, to change the big number to one of 38 simple numbers? Or am I wrong again ??
Once that fraction is obtained, it is multiplied by the range you're looking for, and truncates it to an integer.
Then, in the case of picking a random card, it maps out the numbers 1-52 to a rank and suit. In the case of roulette, maps 1-36 to 1-36, 37 to 0 and 38 to 00.
In the case of shuffling, there are a couple ways to do. In one method, it randomly the cards, just keeping track to make sure it doesn't duplicate a card, until it has a full shuffled deck (or "X" decks). In another method it picks "slots". If the slot is empty, it puts the next sequential card into the slot.
It's really not all that mysterious.
Quote: buzzpaffIn roulette, the RNG is used to pick which value from 0, 00, 1..36 will appear, again with identical probability (this time 1/38).
Pardon me for asking, but it just does not select one of 36 numbers? Isn't there a formula that selects a number from among billions.
Then applies some mathematical formula. dividing, remainders etc, to change the big number to one of 38 simple numbers? Or am I wrong again ??
What a computer-implemented RNG actually does is generate a string of random bits. How that string is interpreted -- as a large number, as a floating point value between 0 and 1, etc. -- doesn't actually matter much. After the generation of the random bit string, a scaling algorithm is applied to convert that random data into a number in the appropriate range.
There are good and bad ways of both generating random data and scaling to desired ranges. Doing either badly means trouble.
Quote: DJTeddyBearIn the case of shuffling, there are a couple ways to do. In one method, it randomly the cards, just keeping track to make sure it doesn't duplicate a card, until it has a full shuffled deck (or "X" decks). In another method it picks "slots". If the slot is empty, it puts the next sequential card into the slot.
Another method, which I believe is well used is the Fisher-Yates shuffle - take a deck of cards, and walk through the deck from top to bottom. For each card randomly swap it with a card at any place below it or leave it in place. Repeat for complete deck.
This is a very efficient shuffle, as it works "in place" meaning you don't require a second array to hold the data. And you only need to generate as many random numbers as the number of cards in the deck. No need to check if a place is full or not.
Wikipedia entry on the Fisher-Yates
The caveat is the RNG must be unbiased, but that's always the case.