Poker Bot piece

Pretend that all owners of poker bots own the same bot. Let's call him FishEater1000. It makes 1bb/hr at 5/10 nl on PartyPoker. Now the creator of FishEater1000 creates a better bot, the FishEater2000 that is able to make an average of 2bb/hr at this game.

The programmer will certainly sell this bot to the public because he can increase his profits by selling FishEater2000. If his bankroll is $10 million then he will not want to have more than $5 million in play. (I am not going to spend the time calculating variance to figure out how much he can have in play that will be virtually no risk so I make a guess of half of his bankroll.) This allows him to have about 5,000 bots in play making him money. But additionally he can make extra capital by selling bots. Selling these bots increases his base capital which increases the amount of bots that he can play that will not overextend his bankroll.

The only real detriment to the programmer for selling bots will be whether these bots will take enough profit away from his other bots to make selling them unprofitable. Selling these bots would decrease the programmer's bot's win rates by decreasing the density of fish and by scaring fish away by their mere presence. However I think that these cons to selling would not be as bad as they seem. Even if they are, however, the programmer would still offer the bots to be sold, he would just sell them at a very high price.

The question becomes, then, not whether or not this programmer will choose to sell bots, but what price he will sell them for.

Even when top-notch poker bots have been created, they will still be sold to the public.

Bad bots will be sold to the public.

A winning bot, that can make $5/hr which is then ran on 40 tables 24/hrs day makes $4800/day. It's only limited by the number of poker sites. THEN it's only limited by the number of accounts on the poker sites.

If you had a bot that could make nearly $5k per day and you wanted to bust the tables for as much as you could short term, you'd probably be running 100 boxes on 100 isp accounts.

I can't see programmers selling this piece of kit for a couple hundred or even thousand dollars, when the footprint of the bot can then be picked up and banned from every poker site in the world.

[magiluke]

[ QUOTE ]
don't you think that when (i think in less then 10 years) a real good poker bot is made (say, winning at higher stakes limit) that isn't detected by poker site's there's no way the maker is going to tell anybody about it or even sell it...?

cya, lars

[/ QUOTE ]

I'm going to try really hard not to sound too sarcastic, so if I do, don't take it personally.

Don't you think that when someone is real good at poker that there's no way they will tell anybody what their methods are or even write books about it.

See, that sounds really sarcastic. I tried hard not to... Anyway, that's my final answer.

[Pirc Defense]

[ QUOTE ]
Actually, limit Hold-Em is sufficiently computationally complex that QC would be necessary for solving it.

[/ QUOTE ]

Incorrect.

[ QUOTE ]
[ QUOTE ]
Actually, limit Hold-Em is sufficiently computationally complex that QC would be necessary for solving it.

[/ QUOTE ]

Incorrect.

[/ QUOTE ]

http://www.cs.ualberta.ca/~games/poker/FAQ.html
Heads up limit holdem has 1,097,690,218,045,566,601 states.

That's approximately 2^(48) states which is computable.

However, finding the optimal strategy is on the order
of 2^(96) which is not practically computable.

[Sniper]

[ QUOTE ]
However, finding the optimal strategy is on the order
of 2^(96) which is not practically computable.

[/ QUOTE ]

It is worth noting that for a human or bot to be successful they don't have to play perfectly optimal strategy!

[Jorge10]

[ QUOTE ]
I can't see programmers selling this piece of kit for a couple hundred or even thousand dollars, when the footprint of the bot can then be picked up and banned from every poker site in the world.

[/ QUOTE ]

What about websites that use bots to take money from people. I read about this recently and pissed me off, check out the quote below.


[ QUOTE ]
Pokertropolis

Pokertropolis is blacklisted because they use poker bots (computerized players run by the house) at their tables. Their management denies this, but anyone who plays there will realize that they are not playing against human opponents.

For what it's worth, the bots do not appear to have any unfair advantage over the players. However, the fact that they are used and then lied about is cause for rather large concern about Pokertropolis.



[/ QUOTE ]

[bdypdx]

Hi Pirc,

I think that your analysis is pretty much right on.

Quite a few of the responders seem to think that they will easily spot and counter a bot. So perhaps, if they spot a bot, they can engage counter measures. But, how will they do that?

Say one has 4 full ring tables going. That's 36 opponents each with a PT icon associated. What is the "bot" icon superimposed by GT+ or Playerview? Is a bot a "rock", a "tight neutral", a "neutral passive", or perish the thought, a "good player"? So far there is no "suspected bot" icon.

I may have already played against a Winholdem bot and have no idea that I did so. Anyway, in post session analysis of PT data one will see the usual suspects as typical opponents, not bot vs human.

[Pirc Defense]

[ QUOTE ]
That's approximately 2^(48) states which is computable.

[/ QUOTE ]

Luckily, the computer does not have to compute each and every state every time it make a decision. A mere fraction of these states, in fact.

As has been noted, the processing power in today's mid-level systems is more than sufficient to power a poker bot.

I agree that the most compelling argument against bot software is that if it was fail-safe...why would they share it with everyone?? I parused the website for WinHoldem and I see no guarantee of winnings...no statistical graphs or charts showing "X"BB/hr winnings, nothing. Why not create a bot that took every bit of online information worldwide into account and put it in charge of your stock portfolio, to decide which companies would be a good buy/sell/hold, based on company history,news releases, quarterly reports, etc...You probably wouldnt trust a bot like that with all your $$. Its just too situation specific.

Which leads me to ask...Who do you feel has the edge, a bot at a table playing (supposed) perfect hold 'em, or a quality 2+2er sitting at the same table who knows hes playing a bot?