Bot complaint

[Robk]

[ QUOTE ]
I've mastered NL, it's not that hard.


[/ QUOTE ]

that's why you spend your time in the 2/4 at party?

[ QUOTE ]
Are you referring to the Paradise bots from four years ago?


[/ QUOTE ]

no he's referring to the univ. of alberta's current bot, which defeated many good human players and held its own against gautam rao. the paper is posted on their website, you can read all about.

[ QUOTE ]
I missed a couple issues of the newsletter Better Bots, was this in there?


[/ QUOTE ]

mike caro did extensive research on this, and wrote about. he even programmed a NL bot called orac (which had very limited success under certain structures).

[ QUOTE ]
One person using winholdem, or someone who had seen it in action, posted that a line of code that would fold AA preflop if the pot wasn't larger than a certain value. It's tough to rule out any bug in a program.

[/ QUOTE ]

but you're referring to a successful program, not just any program.

[ QUOTE ]
I realize that you're trying to be helpful, but when it comes to gambling and MY money I worry about alot of things other don't.

[/ QUOTE ]

strange comment, given you seem remarkably uninformed about this issue.

[Robk]

[ QUOTE ]
Poker is a simple decision tree. Four betting rounds and you try to make the best of 3-4 choices at each point. Not really writing is it?


[/ QUOTE ]

this is such a ridiculous statement that im beginning to suspect you of trolling.

[BusterStacks]

yea this post goes to show that experience does not equal wisdom.

[Bytestream]

I could be the most talented and proficient software engineer on the planet, but if I didn't know anything about amortizing interest rates, I couln't develop a simple mortgage calculator.

Likewise, if I wasn't a consistent winning poker player, there is no way I could develop a program that could do it for me.

The basis step in proving that I could actually write a winning poker program is to prove that I can actually win at poker. No bragging, that's all.

[Bytestream]

Really? What is the 5th choice you can make?

[crazy canuck]

this is such a ridiculous statement that im beginning to suspect you of trolling.

By the way the author is on the U of A team who created Poki.

[Alobar]

I dunno why you are taking flack for this post. I'm glad there are intelligent players out there who are aware of the problem and take appropriate steps to further research their fears.

I also think that everyone who is giving you [censored] is missing the scariest point of this thread. The fact that party isn't doing [censored] about a top level players (i.e. a winning player who generates them tens of thousands of dollars of rake a month, and directly pays them thousands a month) concers. The retarded form letter from your intial email is absolutely absurd, and the second response isnt much better.

I have zero doubt there are bots out there (and that there are some that are winning), but the hope that party poker was at least doing something to thrwart them at least let me sleep at night. Its becoming more and more obvious that placing hope in a site that proves again and again how incompetant and stupid they are, is hope misplaced.

It also amazes me the egos of some people that think they will always be better than a bot because poker is some how really hard or something. It borders on funny, especially some of the names who ruitinely ridicule "can bots win?" posts.

[Yeknom58]

I 100% agree with you on all points..the problem is that I highly doubt a winning bot will check the nuts but I get what your saying.

[MicroBob]

[ QUOTE ]
Chess is a game of complete information, meaning that the position of all the pieces is known to the player, or in your case the bot. The bot can then run through every single combination of sequences and determine the best approach.

[/ QUOTE ]



This is true of course and must be the reason why there are chess-bots that can beat most of the top humans....whereas even the best poker-bots still aren't THAT good.

I really don't know anything about programming....and only a little bit about bots. I've read about poker-bots here....and I've read about chess-bots and have played against some good one's and some lousy one's on the Internet Chess Club (where they are allowed as long as the programmer lets everyone know that they are playing against a bot).


Even with the 'complete information' aspect it still isn't easy to run through all the different possibilities for a chess-bot.

I don't have the exact numbers with me because I can't seem to locate my copy of the book 'The Even More Complete Chess Addict' (or something like that) which contains a lot of interesting chess trivia (interesting to me anyway).


I seem to remember (and I'm only going from memory here) some numbers:
Number of possible chess games consisting of 40 moves is approximated at = 10^10^78
that is....10 to the 10th power to the 78th power.

That's a REALLY REALLY big number.
But to put it in proper context....they contributed a couple of other numbers.

Total number of words EVER printed (or it might have been...ever spoken...can't remember) =
10^10^40

Total number of estimated atoms in the universe =
10^10^58


Yes....the number of possible chess moves and combinations really is a freaking big number. The ways to maneuver the little pieces around a 64-square board is greater than the estimated number of atoms in the universe!!!




Seems they should be able to program a poker-bot to run through all the different possibilities that their opponents could be holding and have it compute pot-odds and make the appropriate decision.

Teach it a little aggression and add in the capacity to play differently against each opponent based on how they've been playing (program it to recognize that someone playing VPIP-17, PFR-9 is more likely to be on a certain range of hands) and I can easily see these things being very successful at some point.

And I still don't understand why it can't be done currently....but, again, I don't know diddly-squat about programming.

[Bytestream]

[ QUOTE ]
It also amazes me the egos of some people that think they will always be better than a bot because poker is some how really hard or something. It borders on funny, especially some of the names who ruitinely ridicule "can bots win?" posts.

[/ QUOTE ]

Yeah, 2+2 books can teach "people" how to win, but applying the wisdom to a mathematical machine that would apply those principles flawlessly, calculating precise odds and having no emotional influence at every step of the way simply would not work.

Its easier to teach Joe Highschooldropout to check raise the nuts on the turn vs lag, but unreasonable that you can instruct a computer to perform this same operation?

I believe there are some psycological factors (defense mechanisms, reaction formation, displacement) influencing a lot of these attitudes towards bots.