Random Shuffling Code Snippet

[iCountNTrack]

The following is a snippet of the random shuffle I use in my BJ programs. I user the Mersenne Twister but obviously you can use any RNG with it.

void shuffle(std::vector<short> & shoe)
{
	short N = shoe.size();  // number of cards in shoe
	short i, j, max, temp;
	for (i = N; i > 1; --i)
	{
		max = i / 2;
		j = floor(max + myRNG(max)); // random index j
		temp = shoe[j];
		shoe[j] = shoe[i - 1];
		shoe[i - 1] = temp;
	}
}

[Alvaro]

Thank you!

[k_c]

The shuffle I use is my adaptation of the Knuth Shuffle.

specificShoe[] is the shoe array with cards numbered 0 to (52*decks-1). I use short integers as I never planned to sim more than 8 decks.

totCards is number of cards in a full shoe = 52*decks.

unsigned long getRandNum(const unsigned long &max) is a function returning a random number (in the range 1 to max) output by a random number generator of your choice.

void shuffle(short specificShoe[], const long &totCards) {
 unsigned long i = (unsigned long)totCards;
 while (i > 1) {
  unsigned long j = getRandNum(i);
  short temp = specificShoe[i - 1];
  specificShoe[i - 1] = specificShoe[j];
  specificShoe[j] = temp;
  i--;
 }
}

k_c

[RS]

So basically you're going from card #52 to card #1. You're choosing a random card and swapping the two? Also, do you rearrange the deck before shuffling, or do you take it out of the discard rack and shuffle that? This is code you've written yourselves?

[k_c]

So basically you're going from card #52 to card #1. You're choosing a random card and swapping the two? Also, do you rearrange the deck before shuffling, or do you take it out of the discard rack and shuffle that? This is code you've written yourselves?

1. Start with a full shoe (totCards=52*decks)
2. Array elements are numbered 0 to (52*decks-1)
3. Swap last element with randomly selected element in the range 0 to totCards-1 - note last element may not change because it may be swapped with itself
4. Last element is now filled and eliminated
5. Decrement totCards by 1 and repeat from remaining elements until 1 card remains - note element 0 may not have changed if none of the previous swaps didn't randomly choose element 0; chance of element 0 not ever being selected = 1/2*2/3*3/4*.....*(totCards-1)/52*decks = 1/(52*decks)

This algorithm made sense to me because you initially randomly get 1 card out of all of the cards in a full shoe, then 1 card out of successively fewer and fewer cards, which is what would happen if each card was successively randomly drawn from a full shoe.

Re your question: It doesn't matter the initial order of the cards. Each successive shuffle reorders them.

k_c

[duanedibley]

Why don't you just use std::random_shuffle?

[Xennus]

Because he isn't using C++

std::random_shuffle is basically what's being implemented in with the above code. It works in pretty much the same way. So I'd rather implement it than use the library code so that I can use it in my C, C++, and Obj-C codebases.

[Norm]

If random shuffle uses a compiler supplied RNG, it's useless. Which I imagine is why ICNT called myRNG. That is, his own.

[duanedibley]

You can pass your own RNG to random_shuffle as an optional third parameter.

[rimar697]

i need help i have $10,000 Bankroll 6deck S17 minimum bet $25 how much positive T/C and how much maximum bet can i bet.thanks

[Three]

Not enough info. Go to Bj resources in the tool bar along the top of the screen. Click on CVCX online in the next tool bar. Input all the appropriate info and you have a ramp, its RoR and the hourly, SD, N0 etc.

[rimar697]

norm,what the difference CVBJ and CVCX ?

[Norm]

norm,what the difference CVBJ and CVCX ?

One id for practice and one is a high-speed simulator, calculator.