Pai Gow

Blair Robins

unread,

Jun 12, 1995, 3:00:00 AM6/12/95

to

igu...@uiuc.edu (Scotty A. Johnson) wrote:
>
>Hi all.
>
>I am looking for any information regarding the rules and rankings in Pai

Gow
>(not Pai Gow Poker).
>
>Thanks for any help you can offer.
>
>Scotty AJ

Unfortunately, there are only 3 books in the English language about Pai
Gow. They are:
1) Pai Gow - Chinese Dominos by Michael Musante
2) How to Play Pai Gow by George Allen
3) Pai Gow Without Tears by Bill Zender

These 3 books may be obtained from:
Gamblers General Store
800 South Main St.
Las Vegas, NV 89101
USA
Phone: (702) 382-9903
Toll-Free: (800) 322-2447
Fax: (702) 366-0329

These books are also available from Gamblers Book Club in Las Vegas. I do

not currently have their address and phone information, so I suggest
calling Nevada information at (702) 555-1212 if you wish to obtain it.

There is another book entitled "Bet on it" by Mary Jane and Greg Edwards.

This book describes just about every game available in Nevada and has a
pretty good section on Pai Gow, but it is not nearly as thorough as the
three books above and it does lacks much important information on the
game.

Another source of information is the gaming guides that are distributed
by the casinos. The "standard" Las Vegas Pai Gow gaming guide
(distributed by several casinos including Treasure Island, Golden Nugget,

San Remo, etc.) is useful only as a reference guide if you have already
learned the basics of Pai Gow well and need to refer to something to
remind you of the domino rankings. This guide does not come anywhere near

explaining the game thoroughly enough so that you would be able to play
after reading it. The best casino gaming guides that I have seen on Pai
Gow are from the MGM Grand and the Las Vegas Hilton. While these guides
are much better than the one discussed above, they still do not include
some important information about the game and are weak on explaining the

strategy of play. If you do wish to obtain one of these gaming guides, go

to the Pai Gow pit and ask for the Pai Gow gaming guide, otherwise you
will probably get their general gaming guide which just briefly describes

all of their games.

Pai Gow is actually not a very hard game to play once you learn the
rankings of the dominos and are comfortable with them. Learning the
rankings, however, does present a challenge to the non-chinese beginner
because we have no frame of reference from which to draw upon to assist
us. Instead, we have to learn the rankings from scratch and memorize
everything, just like when we were kids learning our multiplication
tables. But, once you learn the rankings and know them well, the actual
rules and concepts in Pai Gow are not difficult at all.

I teach classes on casino gambling and I include a class on Pai Gow and
Pai Gow Poker. Most of my students have found Pai Gow to be interesting
and are glad they had an opportunity to be exposed to the game. I am very

glad that I took the time and effort to learn the game because it has
given me many hours of enjoyment. Pai Gow is a lot of fun to play and as

many times as you may have played the game, it seems like something
different or surprising still crops up from time to time.

If anyone reading this has anything to say about Pai Gow, please reply
here publicly or send me an e-mail and I will respond in kind. Although I

am particularly interested in discussing the game with experienced
chinese players, I welcome the opportunity to help novices increase their

understanding of the game.

Bernard James Luger III

unread,

Jun 12, 1995, 3:00:00 AM6/12/95

to

>>Hi all.
>>
>>I am looking for any information regarding the rules and rankings in Pai
>
>Gow
>>(not Pai Gow Poker).
>>
>>Thanks for any help you can offer.
>>
>>Scotty AJ

Hi Scotty,

I love Pai Gow. And, just becuase I'm crazy and needed something better to
do, I sat down and wrote a program that calculated co-optimal stategy for
the game. This was *not* a simulation, it was a direct calculation. And, of
course, I accounted for the differences in betting/payoff structure (e.g.,
in Nevada, you pay a commission on your winners; in CA, you pay a flat rate).
So, after six months of programming and refining and ... well, you get the
idea, I came up with the right way to set any set of four tiles against another
opponent who also plays optimally. And I also programmed in some "house
ways" (ways typical casinos set their hands - this is not 'optimal', but it's
usually good). The grand sum nut of it all is this: as a player, you can't
beat the banker, whether or not you're playing in Nevada or CA, unless the
banker is using a horrid (and I mean horrid) strategy. However, as banker,
you can definetely win in CA (if your opponents are betting enough) and
you can usually win in Nevada (provided you have several additional opponents
besides the house).

I've been debating the value of posting optimal strategy here. Please either
encourage or discourage me.

Quad Q's,

Bernie
--
===============================================================================
Bernie Luger, III Chemistry, Stanford University (415) 723 - 4333
===============================================================================
bjl...@leland.stanford.edu "One man can make a difference
bjl...@d31rz0.stanford.edu and every one must try."
bjl...@chemistry.stanford.edu

Blair Robins

unread,

Jun 13, 1995, 3:00:00 AM6/13/95

to

>I've been debating the value of posting optimal strategy here. Please
either
>encourage or discourage me.
>

Please DO post the optimal strategies!

This should engender more discussion about the game and also be very
informative. I would like to compare my strategies with your computed
optimal ones.

Bernard James Luger III

unread,

Jun 15, 1995, 3:00:00 AM6/15/95

to

Hi gang,

Alright, I've been encouraged. For this part, I'm posting the optimal pair
splitting strategies (these do not vary that much for comission or
non-commission games). For additional insight, I'm comparing to Harrah's
'House Way', which I obtained from a dealer in training.

This is *not* a pai gow tutorial, and I highly recommend the pai gow program
in the rec.gambling archive at Berkeley (Macintosh) to learn the game.

Future posts (when I get the time) will be about setting 9/Wong/Gong and
balancing/front-loading, plus whatever else anyone asks.

May you always set Heaven and Earth,

Bernie
-------------------------------------------------------------------------------
Explanation of the table: The Pair column is the pair you hold in your hand.
The Banker and Player columns tell you when to split that pair, using the other
tiles in your hand as a guide. This table is exact for the no commission game,
and is almost perfect for the commission game.

Harrah's House Way: Same as optimal, except:

They split Heaven or Earth to make 6-8 or better;
They split Sevens as (2, 12) w/ 2, 10, 11, 12;
They never split Long, Big Head Six, or the Mixed Five.

How much is Harrah's giving up? Not much; their 5% commission overwhelms the
small differences caused by splitting incorrectly. But, as a player or banker,
these rules will save you about 0.1% over the Harrah's strategy. Not much, but
hey, it's money!

Bernie Luger III

unread,

Jun 16, 1995, 3:00:00 AM6/16/95

to

On Fri, 16 Jun 1995, Bill Goldman wrote:

> I don't really see the harm in posting it, but you won't really gain
> from it either. rec.gamblers don't bet much at Pai Gow unless they
> already have optimal strategy anyway. You may even get some flames
> from people complaining about your pair splitting strategy.

Let them flame all they want; it's a co-optimal strategy, analytically
calculated. There is no better way to play against an opponent who
either a) is also using co-optimal strategy or b) you don't know what
they're using as a strategy. I guess it should be worth noting two
things at this point: 1) You have an advantage playing co-optimally
against *any* opponent who is not using the co-optimal strategy
('advantage' here means that your expectation is greater). 2) Against a
known strategy, the co-optimal strategy may not be the best.

In other words, co-optimal is always better than the house way, but it
may not be the *best* way to play against any given house way. That
having been said, since few players follow a rigid strategy, co-optimal
play is the right way to go. Against the casino, however, it is better
to play against their house way directly.

> I am curious though as to whether there is a rigid optimal strategy, or
> whether you need to use a little game theory (like split 2s with a 4-6
> 19/32 of the time).

Well, now that you mention it, *yes*, game theory does apply when you
have an opponent who is also playing co-optimally against you; there are
certain hands that if you know your opponent would set them one way, you
would set certain other hands another way. However, these hands do not
involve pairs, and account for only 0.1% of your expectation. I'll
eventually get around to talking about these hands, though that's a ways
off. I'll be lucky to have time to post about 9/wong/gong setting next
week.

> Also I've thought about computing strategies based upon the knowledge
> that your opponenets hand has a given domino in it. If you can play
> 32 card monty well, it may be worth tracking a 12 or a 3 or something.
> It only has a 1/8 chance of coming into play, but .1%s add up in the
> long run. Have you done any work in this area?

Yes, I've looked at this as well. But, it's not usually worth it. For
example, by knowing the dealer has a 12 in his hand, you can change the
way you set your tiles for about a 2% advantage. This advantage,
however, is swamped by the inherent advantage the dealer now has (over
23%!) just by having that tile in his hand. On the other hand, if you know
two tiles, you can start to do some real damage.

Well, enough for now.

May you always find Heaven and Earth,

Bernie