The 1-player game has 1,069,585,617,930,903,622,760,859,648,000 (1.07 nonillion) branches in the tree of possible outcomes.
The 2-player game has 5,121,301,942,761,316,925,474,565,470,398,818,041,856,000,000 (5.12 quattuordecillion) branches in the tree of possible outcomes.
The 3-player game has 8,261,952,859,233,476,421,844,194,443,440,293,355,459,268,562,124,800,000,000 (8.26 octodecillion) branches in the tree of possible outcomes.
The current world's fastest supercomputer (of which there is only one) can't even put a dent in these numbers. You would need 11.5 trillion of them to analyze the 1-player game (in one second), 55 octillion of them to analyze the 2-player game, and 89 duodecillion of them to analyze the 3-player game. This would be reduced somewhat if you were analyzing a specific starting hand, and somewhat more if you were in a later position, but this is obviously not enough.
ZCore13
I have analyzed 8 card poker (split 8 card hand into 3 hands that are 3,3,2 cards) using spreadsheets, also developed strategy rules using simulators But I admit I have not tried to do 13 card (5,5,3) Chinese poker.
I spent more time than I care to admit coming up with alternative ways to try to analyze the deal and the first draw. The problem is that there are so many combinations that every little detail of every possible implementation becomes an enormous bottleneck. I resorted to doing much of the coding in assembly.
You don't just get 13 cards. You start with 5 and place them anywhere you like. Then for each of the four draws, you get 3 cards and must choose 2 of them, discarding the other. So you see 17 cards in all, plus whatever cards your opponents play, if applicable.
I tried a few things. For the deal, knowing the 5 cards, I would then deal 8 random available cards and "maximize" the 13-card combination as if in Fantasyland. Then note the position each of the 5 deal cards ended up in. Repeat this many times, and generally one placement emerges as being the most common. This is obviously not optimal, but it gives an idea of what might be a good starting placement. Many starting hands have 2 or 3 placements vying for "most common" status.
There are also many hidden nuances. For example, the kickers to 3 of a kind (or 4 of a kind, or the pair in a full house) are irrelevant. Such hands cannot tie, so it's better to ignore the kickers so as to maximize something else. There were other situations too that I can't remember right now, but the choice of how to "maximize" a 13-card combination when you have several choices all worth the same number of points, has a significant impact on which 5-card placement emerges as being the most common.
Quote: Zcore13OFC is my favorite poker game. To me, the pineapple version dumbs it down a little. I wish the closest casino that spreads it wasn't 300 miles away from me.
Hey Zcore13, did you ever see my casino game version of OFCP, I would love to get your thoughts on it, you can play it HERE
You make bets depending on how likely you think you will make an open hand, you need 1 pair to win and the dealer either opens or fouls your hand does not play against the dealer. If the dealer is open you get paid from the pay table, if the dealer fouls you get 2-1 on your ante bet and the middle and back bets are pushed when you have an open hand. If you foul you lose your ante bet and your middle and back bets get pushed when the dealer is also foul. If the dealer is open and you're foul you lose all bets.
As to this post, well I did do some initial maths but then succumbed to writing a game simulator but, and this is a BIG BUT, my game does not allow the player to set out the community cards, they are dealt in a specific order to specific locations on the table so this did make the maths slightly easier (lol....easier!!!!)
Perhaps the approach to this is not looking at all possible combinations but breaking it down into the stages of the game. You start by being dealt 5 cards from a possible 52 and as we know there are 2,598,960 possible combinations from which you need to work out the best combination of card positions from a 5 card back row to a 3 card front row. You need to determine what five cards you have so if you had four cards of one suit you put the for cards on the back row and the last card if less than a 6 may then go on the front row with 7+ going in the middle row. You then need to take a random 3 cards from the remaining 47 which is 16,215 combos and figure out which 2 card work best with the 5 cards and their placement. Then you have 3 cards from 44, 3 cards from 41 and finally 3 cards from 38. The more cards placed the easier it should be to figure out the best placement. That's how I'd approach it and I'd write a simulator to do it.
Quote: UKMarkHey Zcore13, did you ever see my casino game version of OFCP, I would love to get your thoughts on it, you can play it HERE
You make bets depending on how likely you think you will make an open hand, you need 1 pair to win and the dealer either opens or fouls your hand does not play against the dealer. If the dealer is open you get paid from the pay table, if the dealer fouls you get 2-1 on your ante bet and the middle and back bets are pushed when you have an open hand. If you foul you lose your ante bet and your middle and back bets get pushed when the dealer is also foul. If the dealer is open and you're foul you lose all bets.
Just checked it out. I'd say it's a basic variation of OFCP. With no choice of card placement, which is a HUGE factor and skill in playing OFCP, it's more similar to Pai Gow Poker than OFCP. That being said, I'd play it if the house advantage was fair.
On a side note, I think the game is not paying correctly, unless I'm missing something. The page is protected somehow, so I can't take a screen shot, but I had the following hand:
My Card_______________Board_______________Dealer Card
Kh.................................... 5s 3c...................................10d
8c................................2h 9d 6c 8d..............................6d
Jd................................Qc 8h 4h Jh............................. Qs
It gave me the win on the bottom hand with a pair of Jacks
A loss in the middle hand
A loss on the Ante hand.
ZCore13
Quote: JB
The current world's fastest supercomputer (of which there is only one) can't even put a dent in these numbers. You would need 11.5 trillion of them to analyze the 1-player game (in one second),
So am I correct that it would take one computer 11.5 trillion seconds to analyze it? (That's 364,662 years for those of you playing at home.)