|
|
#141
02-28-2005, 02:09 AM
|
|||
|
|||
Re: A little confused....
In chess 100% of the information is on the board. The correct or incorrect play is obvious. In poker you have no way of knowing what cards your opposition has or how they will play those cards. Also players vary their play based on other players, position, chip count, beers digested or many other criteria. A good player will beat a bot to bankruptcy in the long run.
Thank you, Jim Kuhn Catfish4u [img]/images/graemlins/spade.gif[/img] [img]/images/graemlins/diamond.gif[/img] [img]/images/graemlins/club.gif[/img] [img]/images/graemlins/heart.gif[/img] 
|
|
#143
02-28-2005, 02:34 AM
|
|||
|
|||
|
Re: saabpo: Bot or Joke?
I have read through this long thread, including the sidetrack issues of comparing poker to chess bots, and the ethics or lack of it in using a bot if it were possible. Forgive me if I read through too fast and someone else has made the substance of the comment I am about to make. Also as to whether any given player could possibly be a bot, the # of hours a day it was observed playing would be the deciding evidence to me.
I think the whole issue here, if you want to determine whether this player is a bot is whether various human players actually play in the described manner. I think the answer to this question is yes based upon my own observations in tournaments. The telling clue here is the push/fold pre-flop actions. This is merely following Sklansky's 'system' given in Tournament Poker For Advanced Players. They push with AK, AQ and any pair, hoping not to run into a bigger hand. Unless I misunderstood lorinda's observed rules, then it would seem very possible to me that a human player was using such a system, admittedly the easiest to program a bot with. Again to me, if this player were to be observed playing over 12 hours a day, then only then would I conclude it was a bot. Finally, should you care if it was a bot? The counter-strategy to such a system is merely to call such a pre-flop push with any hands that are in maybe the top 40% of the top 15 excluding AQ, thus with AK and TT-AA, never worrying that TT is not good enough of an edge in the long run.
|
|
#144
02-28-2005, 02:39 AM
|
|||
|
|||
|
Re: saabpo: Bot or Joke?
the # of hours a day it was observed playing would be the deciding evidence to me
The speed of betting, while not entirely constant, would have been consistent enough to convince almost anybody after a few games. Lori
|
|
#145
02-28-2005, 03:14 PM
|
|||
|
|||
|
Re: saabpo: Bot or Joke?
i think it would be easy to make a poker bot.
basically to start off preflop, it would have certain hands to raise, certain hands to fold certain hands to call, certain hands to fold to a raise, certain hands to 3bet and cap, based on preflop action and a person's VPIP, the bot could then give you a good range of hands that person may be playing. so you have a set of algorithms to calculate your current pot odds, the hand you have etc. it would then make a decision as to whether to fold, bet, raise, check raise, etc. all based on pokertracker info on the people ahead and behind you and what your current hand is. simple weighting system would make the decision turn and river are essentially the same. this bot would play a VERY good ring game, taking down probably 2-3bb/100 hands just some redimentary programming to make it look human (random delay, sitting out hands every few hours for "washroom" breaks, only playing 8 hours a day. stuff like that.
|
|
#147
02-28-2005, 03:53 PM
|
|||
|
|||
|
Re: saabpo: Bot or Joke?
[ QUOTE ]
A bot could be created to do well at small-stakes No Limit. It's strategy would be similar to a nut peddler's. [/ QUOTE ] What do you mean by a nut peddler's strategy? I'm not familiar with that term. -Mike
|
|
#148
02-28-2005, 04:00 PM
|
|||
|
|||
|
Re: saabpo: Bot or Joke?
[ QUOTE ]
[ QUOTE ] A bot could be created to do well at small-stakes No Limit. It's strategy would be similar to a nut peddler's. [/ QUOTE ] What do you mean by a nut peddler's strategy? I'm not familiar with that term. -Mike [/ QUOTE ] Only play very good hands and continue when the flop hits you. I first saw it used when the $25 NL full tables were very juicy. All you had to do was nut-peddle and you would get paid off every time.
|
|
#150
02-28-2005, 04:34 PM
|
|||
|
|||
|
Re: A little confused....
[ QUOTE ]
In poker you have no way of knowing what cards your opposition has or how they will play those cards. Also players vary their play based on other players, position, chip count, beers digested or many other criteria. A good player will beat a bot to bankruptcy in the long run. [/ QUOTE ] Ummm, why couldn't a bot vary its play based on other players, position, chip count and many other criteria. When good bots are written, I am positive they will not do the same thing in every situation. E.g., AA UTG+1 after UTG folds. No relevant information about table: Bot: 92% raise, 8% call Two LAGs at table: Bot 70% raise, 30% call. Imagine playing a bot heads up that has 5 HU playing styles. All of them solid winning styles. Before every hand it randomly picks one. Humans have great advantages against simple bots. Bots could be constructed that negate those advantages. Basically, any opponent who never varies his play would be easy pickings for a good player. Why would a bot not vary its play? rabbit
|
 |
|
|
Linear Mode
