## Caribbean stud poker – wizard of odds gas 2015

####

In Caribbean Stud Poker the player has the choice to make a side bet of \$1 which pays for hands of a flush or better.The specific payoff tables vary from place to place but always feature a progressive jackpot, paying 100% of the jackpot meter for a royal flush and 10% for a straight flush. In the very unlikely event that two players had a royal flush in the same hand the first one to the dealer’s right would win the jackpot and the second would win whatever the jackpot is reseeded to, usually \$10,000. The reason for this is the dealer pays players from right to left. In the event that two players received a straight flush at the same time, the first one to the dealer’s right would get 10% of the meter and the second would get 10% of what was left after the first player was paid. I have heard (but can not confirm) that once in the same hand one player got a royal flush and another player got a straight flush. The player with the royal was closer to the dealer’s right and thus got the full jackpot amount. The meter was then reseeded to \$10,000 and the player with the straight flush got only 10% or that, or \$1,000.

A manager at Casino Niagara kindly explained how the jackpot meter works. For every dollar bet, 71 cents goes into the jackpot and the casino keeps the other 29 cents. A dealer at one of the Connecticut casinos said the contribution rate at his casino was only 65%. This rate of contribution can vary from place to place. All payoffs are paid right out of the meter. Every time somebody hits a royal flush the house contributes \$10,000 (called the seed) to the next jackpot. The house edge is just under the cut per bet because the casino puts up the initial seed to start a new jackpot after somebody wins the previous one. At the Casino Niagara the house can expect to receive 18.84times as much money from the 29% cut as it pays to seed new jackpots.

The next two tables show 11 side bet pay tables I have seen or heard about. The pay table number and pay table itself are listed in the top rows. The third row from the bottom is how much the flat wins contribute to the return. The second row from the bottom is how much the progressive wins contribute to the return for each \$10,000 in the meter. The bottom row is the break even point, or how high the meter would need to reach for the return to be 100%.

Casino Canberra: I have heard the Canberra Casino follows pay table 7 but also pays \$50 for a "deadman’s hand" consisting of AA88x, where x is any other card.For the side bet to have no house edge in this game the meter would need to reach \$149,389.47. For a \$5 minimum game to have no house edge the meter would need to reach\$238,716.85, and for a \$10 game the meter would need to be \$328,044.23. Player Collusion

I’ve been asked lots of times if an advantage can be gained by sharing information with other players. Although the rules forbid this in the land of casinos, there are multi-player Internet casinos where this could be done very easily by phone. However, don’t get your hopes up. According to the paper "An Analysis of Caribbean Stud Poker" by Peter Griffin and John M. Gwynn Jr., which appears in the book Finding the Edge, in the perfect situation of having seven colluding players, it would be possible to narrow down the dealer’s unseen cards to just 16 possible cards. Using a computer to analyze all 1,820 possible 4-card sets out of 16, the player would have an advantage of 2.3%. In a six player game the house would still have an edge of 0.4%. An article on this topic by Scott McIntosh at Review Poker Rooms also explores this topic and comes to similar conclusions.

Stealing a look at other player’s cards can cut down the house edge marginally. If you have a borderline ace/king hand it would help to see if the other player’s cards match the dealer’s up card. The more that match, the more inclined you should be to raise. While the Internet would be a perfect forum to share information in a multi-player table, the most number of seats I have ever seen is three.