Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Assume a simplified poker deck (to cut down the # of combinations):

  • 4 suits (h, c, d, s)
  • 5 ranks (A, K, Q, 7, 2)
  • 20 cards

Assume a basic 2-player setup:

  • Hand 1 (2 cards)
  • Hand 2 (2 cards)
  • Flop (3 cards)

There are 16,279,200 combinations of the above setup: choose(20, 2) * choose(18, 2) * choose(16, 3).

Take the following combination:

  • Hand 1 = Ad, As
  • Hand 2 = Kc, Kh
  • Flop = Qd, 7s, 2c

Question: what is the most efficient way to calculate the number of combinations (within the 16,279,200) that are suit-isomorphic with this setup?

For example, the following setup is suit-isomorphic:

  • Hand 1 = As, Ac
  • Hand 2 = Kh, Kd
  • Flop = Qs, 7c, 2h

while this example is not:

  • Hand 1 = Ad, As
  • Hand 2 = Kc, Kh
  • Flop = Qc, 7s, 2d

nor is this example:

  • Hand 1 = Kc, Kh
  • Hand 2 = Ad, As
  • Flop = Qd, 7s, 2c

===Edit ... explaining what I mean by suit isomorphism===

The only isomorphic transformation allowed is one that shifts suits, not ranks. The first example failed example isn't "suit" isomorphic because f(d) isn't a one-to-one map: on the Ace in Hand 1, f(d) = d. On the Queen in the flop, f(d) = c. The second example failed because the ranks within Hand 1 and Hand 2 were swapped.

===Edit ... a more general approach to the question===

The solution shouldn't rely on a specific structure of the problem. For example, in the example I gave, the solution is 24 (i.e. the permutations of the 4 suit symbols). But what happens in the following cases.

  • I might start eliminating cards. For example, I might pull the 2c out of the deck.
  • I might not use all 4 symbols.
share|improve this question
    
I'm not clear about what you mean by "suit isomorphic"...for example, what is the status of: Hand 1 = Ac,As; Hand 2 = Kh,Kd; Flop = Qd, 7s, 2c? And what about Hand 1 = Ac,Kh; Hand 2 =Ad,Ks; Flop = Ah, Kc, Kd? –  amWhy May 5 '11 at 20:58
    
1) Not isomorphic. Neither of the aces (which should be treated as a set ... order doesn't matter) have the same suit as the Queen. So f(d) = d and f(d) = either s or c. My understanding of isomorphism is that it is a one-to-one map, not a one-to-many. –  MikeRand May 5 '11 at 21:36
    
2) Not isomorphic. The ranks are different. The only transformation allowed is on the suits (h, d, c, s). –  MikeRand May 5 '11 at 21:36
    
got it; I wasn't clear if you were providing an "example" or if your examples were defining the isosmorphism...Yes, an isomorphism requires a one-to-one map, but not all one-to-one maps are isomorphic. i.e. being injective is a necessary condition for an isomorphism, but not at all sufficient. –  amWhy May 5 '11 at 22:03
    
Is isomorphism even the right term for what I'm describing? Unfortunately I have no formal math training, so the best I can understand is that I've designed some k-partite (maybe k = 4) graph but am only testing for isomorphism across one of the sets of vertices (i.e suits). –  MikeRand May 5 '11 at 23:03

1 Answer 1

up vote 3 down vote accepted

You are asking for an orbit of a group action, where the group is the group of all permutations of the suits, and the action is the action on the setups where a permutation acts on each card individually. The orbit can just be constructed by taking the 24 possibilities and seeing which ones are "allowed" and which are different. 24 is a tiny number.

Sometimes though permuting the suits may not actually change the setup. Consider the example:

  • Hand 1: 3♣, 3♠
  • Hand 2: 3♢, 3♡
  • Flop: 2♡, 4♡, 5♡

Then the permutation that switches ♣ and ♠ but leaves ♢ and ♡ alone actually ends up leaving the setup alone: hand 1 switches the order of the two cards, but that doesn't change the hand; hand 2 and the flop stay exactly the same.

Assuming you don't throw any cards out, the number of isomorphic setups is always a divisor of 24, namely the index of the "stabilizer" of the setup (see the "orbit-stabilizer theorem").

If you start throwing out cards, then basically your problem has no structure. This is not a big problem, because 24 is a tiny number. Just try all 24 and see which are "allowed".

share|improve this answer
    
Very interesting, never even came across orbits. Quick question: by throwing out cards, do you mean (in your example) if I were to throw the 5d out, for example? –  MikeRand May 6 '11 at 10:26
    
Exactly. Then if there was a 5h in the setup, you'd have to disallow a heart/diamond switch. The set of allowed permutations of the suits might no longer be a group, and then it really is honestly just checking all 24 ways, AFAIK. –  Jack Schmidt May 6 '11 at 12:32

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.