Some Things I Learned About Automated Trading Systems From Writing a Poker Bot

Discussion in 'Automated Trading' started by pipscooper, Jul 6, 2005.

  1. Hmmm, anyone named Terry in this forum?
     
    #51     Jul 18, 2005
  2. ElCubano

    ElCubano

    i know what its reffered to..i was wondering if ud done it and how much better ur bots where against the remaining 2 players considering you have much more info at this point....
     
    #52     Jul 18, 2005
  3. No, although some may say anyone who runs a bot is cheating, I felt that collusion was cheating and didn't want any part of it. I would say that it would provide a significant advantage in two areas. First, you could pump the pot when one bot had a very good starting hand and you have a calling station or a loose player in the mix. Second you could better calculate your approximate odds by knowing an additional 2 cards, but the benefit of that would be far less than the benefit from pumping the pot IMO but over time that small edge would add up especially on a 6 max table.
     
    #53     Jul 18, 2005
  4. ElCubano

    ElCubano

    yes i agree.... how easy is it for the site to detect this???
     
    #54     Jul 18, 2005
  5. I'd be interested in knowing exactly what the disadvantage is, for other players, in collusion. The series Tilt basically made the problem of collusion its main plot.


    Except for knowing what those extra two cards are what is the advantage?
     
    #55     Jul 18, 2005
  6. ElCubano

    ElCubano

    in the way u bet into a caller....if lets say your BOT bets with nothing knowing BOT2 has a monster hand ( 2 positions away) then the middle guy calls BOT1 bet only to get raised by the BOT2 ( which is behind him ) and so on and so forth....raising the pot...
     
    #56     Jul 18, 2005
  7. I'd like to know if anyone used game theory to quantify the disadvantage of collusion? Would be an interesting research project.

    Knowing the money is there for the stealing, then one must assume that teams are common. The claim that the sites screen for teams must be questioned, knowing how hastily web sites in general are slammed together. Other than some basic, IP-based alert trigger, they are probably doing nothing.

    I mean Etrade cant do basic security and fraud detection on the ACH pulls that it gets (if one is to judge from the posts on ET). I don't think that poker sites really have professional, banking level fraud filtering and prevention applications running.
     
    #57     Jul 18, 2005
  8. Banjo

    Banjo

    That would be interseting puffy, if there's enough $$$ involved somebody's doing it.
     
    #58     Jul 18, 2005
  9. I would say that most of the time collusion is spotted by other players, reported to the site, investigated and eventually the colluders may or may not be caught. But Puffy is right that the site on its own probably isn't smart enough to track all the players and figure out who is colluding and who is playing normally.

    If the same 2-3 players are always in the pot together bidding it up, that is a possible sign of collusion (or of 2-3 maniacs at the same table).
     
    #59     Jul 18, 2005
  10. I believe that asking the right questions is very important in developing as a trader. One series of questions I asked after I had learned the basics of poker helped me change my paradigm about developing trading systems.

    What is the best starting hand in poker? Obviously it is AA. There are numerous statistical studies that prove that pocket aces wins more times than any other starting hand. If you could play a version of poker where you only had to put money in the pot when you had aces, would you have an edge at such a game? If the dealer kept the game going, and the other players didn't know that you only play aces could you win? So the question that really helped me change my paradigm is this: In trading, what is the equivalent of holding pocket aces? Think about that for a while. I hope it starts you down some profitable paths.
     
    #60     Jul 21, 2005