A clone of the original Poker Hand Simulator, rewritten in Ruby. A GUI is soon to-be-made using RubyFX or Shoes!
In a poker game, how many times on average do you need to draw a hand of cards before you finally get a straight flush? The Poker Hand Simulator was created as a statistical experiment to answer this question. There were a few bugs with the Javascript version, so I've recoded the program in Ruby, which is an Object-Oriented language.
The Poker Hand Simulator draws n=7 cards (such as in Texas Hold'em) from a shuffled deck of 52. The program then determines the best 5-card hand from the draw. The value of 'n' can be changed in the file run_phs.rb next to the comment "number of cards you want to draw." It does this repeatedly until it draws a straight flush, and tallies up the results. The program can only be used from the command line right now, but a GUI overlay will come eventually.
First, make sure you have Ruby installed (version 1.9 or later). Download the files from the repository and from command line, run the simulator using ``` ruby run_phs.rb ``` *An executable of this ruby script will also be coming soon.*Say the program draws a random hand [Qs, 3c, 9d, Qd, 5c, 10d, 2s]. It will determine that the best five-card hand will consist of the pair of Queens, and the three highest kickers, being 10, 9, and 5. The output will be:
You drew: [Qs, 3c, 9d, Qd, 5c, 10d, 2s]
Best hand is 5c, 9d, 10d, Qs, Qd
Pair of Queens
Since the hand was not a straight flush, the program will continue drawing and identifying new hands until it does find one. In the end, it will tally the results as such:
Final hand strength tally is:
High-Cards: 563
Pairs: 1459
Two-Pairs: 778
Triples: 138
Straights: 166
Flushes: 102
Full Houses: 93
Quads: 3
Straight Flushes: 1
Total number of draws before hitting a straight flush = 3303
Another interesting thing is that this program can be used to test if a random number generator isn't working as intended. The number of draws before hitting a straight flush will more or less follow a geometric distribution, with an expected value of approximately 3215. Repeated major deviations from this number may signify that something is wrong!
There are a few things remaining to be done. Firstly, an exe would probably be more convenient to run than the raw ruby source code, so that's next on the list. After that, a GUI will be built (using [Ruby Shoes](http://shoesrb.com/) maybe?).Finally, the algorithm that finds the best 7-card hand basically uses brute force. Since the major variants of poker use 7-12 cards, it didn't seem necessary to speed up the program that much. In addition, most high-value hands become quite common as you increase the number of cards you draw.
This results in O(n-choose-5) computation time if you vary 'n'. So far, the program is working for up to n=21 in a reasonable amount of time. In any case, I'll try to speed up this program using some kind of branch-and-bound method if I feel like it.
| Rank | Example |
|---|---|
| High Card | [As, 3d, 5s, 10h, 8d] = Ace High |
| Pair | [As, Ac, 5s, 10h, 8c] = Pair of Aces |
| Two-Pair | [As, Ac, 5s, 5h, 8c] = Two Pair: Aces and Fives |
| Three of a Kind | [As, Ac, Ad, 5h, 8c] = Three of a Kind: Trip Aces |
| Straight | [7c, 8d, 9c, 10s, Jh] = Straight: Jack High |
| Flush | [2h, 5h, 6h, 10h, Jh] = Flush: Jack High |
| Full House | [As, Ac, Ad, 5h, 5c] = Full House: Aces over Fives |
| Four of a Kind | [Ks, Kh, Kd, Kc, 5h] = Four of A Kind: Quad Kings |
| Straight Flush | [7c, 8c, 9c, 10c, Jc] = Straight Flush: Jack High |