PaddyBoy: Q for J.morgan re;caribbean stud

[PaddyBoy]

A few casinos in Europe are offering a highly beatable variant of Caribbean Stud Poker. The pay out schedule is identical to the American game. The game offer two differences from the American version:

1.After the deal the player can change one or two cards for the price of one or two antes respectively. Afterwards the player decides to raise or fold his hand.

2.Players are allowed to show all their cards to each other and share information

What would be the optimal strategy for this game and what would be the player's edge?

I know a Griffin/Gwinn paper concluded that with conventional American rules sharing info about the matching cards to the dealer's upcard in a 7 players situation gave player edge of +0.7% and perfect play strategy with 7 players sharing info about their hands gave the player +2.3%.

[Rambler]

Griffin & Gwinn wrote about this? Where can I find this papaer?

[Shark]

> Griffin & Gwinn wrote about this? Where
> can I find this papaer?

In a book named : Finding the Edge: Mathematical and Quantitive Analysis of Gambling. It's a University of Nevada at Reno publication. I will link to Amazon. It includes the table of contents. Many valuable papers.

I believe this book merely computes the edge with perfect card information. It doesn't present a full optimal strategy. Also PB plays a superior game where you can exchange cards for an ante and two antes.

I hope that helps, Shark

Finding the Edge

[Don Schlesinger]

> Griffin & Gwinn wrote about this? Where
> can I find this paper?

If you're looking for any references to Gwynn, make sure you spell his name right!

Don

[boiko]

...contacting Cal State University, Sacramento, Ca.

Try to contact either John Gwynn or Armand Seri, of the computer science or math departments. They were fellow professors of Peter Griffin. I know they both did some blackjack work with Peter. As you probably know, Peter passed away a few years ago. I believe one of the two fellow professors also passed on recently. Whether the remaining professor still teaches or not, I don't know.

> Griffin & Gwinn wrote about this? Where
> can I find this papaer?

[Don Schlesinger]

> Try to contact either John Gwynn or Armand
> Seri, of the computer science or math
> departments.

Contacting John will be problematic. :-)

> They were fellow professors of
> Peter Griffin. I know they both did some
> blackjack work with Peter. As you probably
> know, Peter passed away a few years ago. I
> believe one of the two fellow professors
> also passed on recently.

That would be Gwynn.

> Whether the remaining professor still teaches or not, I don't know.

Easy to look up. I haven't read anything from Armand Seri in ages.

Don

[J Morgan]

> A few casinos in Europe are offering a
> highly beatable variant of Caribbean Stud
> Poker. The pay out schedule is identical to
> the American game. The game offer two
> differences from the American version:

> 1.After the deal the player can change one
> or two cards for the price of one or two
> antes respectively. Afterwards the player
> decides to raise or fold his hand.

> 2.Players are allowed to show all their
> cards to each other and share information

> What would be the optimal strategy for this
> game and what would be the player's edge?

> I know a Griffin/Gwinn paper concluded that
> with conventional American rules sharing
> info about the matching cards to the
> dealer's upcard in a 7 players situation
> gave player edge of +0.7% and perfect play
> strategy with 7 players sharing info about
> their hands gave the player +2.3%.

Sorry, I can't help you here. I haven't done any analysis of that game, and to do a thorough analysis could take hundreds of hours. If this topic makes it to my second edition (I doubt it), I'll let you know, but it's WAY down the list.

As for the American version, G & G did indeed conclude that a 2.3% edge was theoretically possible with optimal play of seven "super-colluders," but that number is truly irrelevant. Computing a strategy and computing the edge when using that strategy are two completely different things. Notice that G & G do not provide the strategy. To do so would probably fill an entire book.

As for the 0.7% figure, I don't know where that comes from. I did not see it in the version of their paper published in the Finding the Edge book.

They do mention that six players is not enough. To have a crew of seven players work together to produce a 0.7% edge (assuming that number is legitimate) would be ridiculous, a complete waste of time and talent.