Camp 2011 - Version 1.4
Chaos Communication Camp 2011
Project Flow Control
| Speakers | |
|---|---|
|
Torbjörn Lofterud |
| Schedule | |
|---|---|
| Day | Day 4 - 2011-08-13 |
| Room | Baikonur |
| Start time | 15:00 |
| Duration | 01:00 |
| Info | |
| ID | 4424 |
| Event type | Lecture |
| Track | Science |
| Language used for presentation | English |
| Feedback | |
|---|---|
|
Did you attend this event? Give Feedback |
Poker bots
Developing and running autonomous pokerbots at online casinos
For a few years I was part of a team that developed and ran autonomous poker playing robots on commercial Internet poker sites; playing poker with real money against real people in real time. The project failed...
At first glance, Texas hold'em poker does seem like a fairly simple game. Developing poker playing software can be done and has been done, and there are rumors on the Internet about poker playing robots winning millions online. There are even some commercial poker playing software available on the Internet.
But building a functional poker bot have two major parts; firstly integration with the online poker site, and secondly developing software capable of winning against human opponents in Texas hold'em poker. Contrary to popular belief, the first part is easy and the second part is hard.
Texas hold'em provides a programming challenge extraordinaire because its an imperfect information game paired with lots of randomness and psychology. Only small pieces of information is available at a given time, and the available information is biased and often deceptive. The complete game-tree Texas hold'em poker is so large that its infeasible to calculate even offline, and impossible to do in real-time, a feat necessary for online game-play.