Dear creators of PokerTH,

thank you for your fine and free software! I especially appreciate the
idea of a platform-independent AI-game for OFFline training and playing.
I learned of it via wikipedia reference and, about the same time, via a
bonus CD coming with the Poker special edition of print magazine CHIP.

While the version number still is a zero-something, the project looks
ambitious (to say the least . I guess whether my grateful two cents
below may be of some help on the way up to PokerTH v1.0 ?

Crawling the net --at various servers, plz use key words-- I discovered:
: Poker Calculator 1.1.4.
: Freeware - (c) 2002 Janne Raevaara.
: Poker Calculator is a program for calculating poker show down odds.

The software is Java-based, the source is freeware. Perhaps the math
algorithms can be adapted to the project. Java and C++ are conceptual
relatives, I've read, but I'm not in that business.

I want to argue in favor of adding realtime percentages to PokerTH,
similar to the data provided with legal plug-ins at some poker sites.

(1) It may be optional. Game host and players agree whether it's ok
or not. A LED in each players' window could indicate the use of help.

(2) It may be seen as a matter of fairness. It's more contesting to
play against informed opponents. And for math disableds like me (at
least when under pressure) it would fit as a nice learning tool.

(3) The computer-AI might also profit, but I have no idea how much it
already depends on math. I'm just curious of Mona Lisa's nice play .

(4) To keep things as simple but helpful as possible, data should be
plain_worst_case nofold'em showdown percentages, compared to the given
real pot odds. (Of course, each player should have access to her/his
individual odds based on their hole cards and the board showing.)

(5) Maybe you see few sense in #4, plz consider this: PokerTH creates
a wonderful world of both sit'n'go and final table game dynamics with
increasing blinds. It's not only the super-aggressive players who come
in with the murkiest hands at that stages. So the plain nofold'em data
might be a sound base of decision.

(6) The training effect of having seen the "numbers" of lots of hands
may improve my real play. Perhaps I adapt sort of card reading ability
by simply being told often enough that --given a low backdoor draw--
the chance of winning is slim, even IF it hits on the river...

(7) Final topic: a table size of up to ten players? That would make for
a perfect STT simulation, filled up with computer opponents as needed.
A tableful of smart AIs is fine when sitting lost in an offline spot.


I wish all the best to you and your project.
Regards and a happy & successful 2008!