Bots and the future of online poker

[dismalstudent99]

RITT,

1. No one's talking about pure strategies. If you know someone who relies on a pure strategy, please give me his screen name!

There exists a mixed strategy NE, though we may not know what it is.... yet

BTW, computers can randomize better than humans, therefore they can implement mixed strategies more precisely (assuming they are programmed to do so).


2. Why do you assume botting technology has stopped developing (plateaud)? Is there evidence that programmers have "hit a wall"?

With all the attention poker is getting now, some good minds (and some careers) are on the problem.


3. Yes, I am interested in what people enjoy doing -- I want online poker to remain highly populated with players, especially with fish. Therefore, if many players object to bots and threaten to leave online poker, then we should do everything to eliminate the bots. What's "backwards" about that? And what's wrong with "backwards" anyways?

4. Since I pay attention to what interests people, I can confidently say that most people would not enjoy tinkering with a "baby bot" as much as you do. Would I want to play HALO 3 via programming ("Use Grenade 33%, Rockets 66%, Jump when evading 100%")? NO. I'll wear out my thumbs instead.

However, I understand that botting has an intellectual appeal... much like Marcel Proust has an intellectual appeal among the literati. But my point is that these are small groups of people who appreciate these things. When it comes to online poker, size does matter.


5. Talented bots would have a different effect on the poker economy than talented people. This is because the bots can be replicated and become wide-spread, whereas real poker talent seems to be a rare thing. It is this feature makes bots a threat to the game.

One SBRugby playing at 200/400 doesn't affect me. But 2 SBRugby-bots at my 0.10/0.25 would probably cripple me (and stop me from donating money to 0.25/0.50 players) Do you see the difference?


6. Competitive pressures will force major sites to take an anti-bot stance, especially if the bots are good.

Here's an extreme case.

Suppose FTP allows bots but PokerStars bans them.

Suppose I'm a breakeven, 100NL player on FTP. However, today I find out that I have to 6-max against:

2 "SBrugby bots"
1 good player
1 average player, and
1 fish.

My EV just went from 0 to negative.

Obviously, I'm switching to Stars, even if it means putting up with a bunch of irritating anti-botting measures.

It gets worse-- the table at FTP now has only 5 players. The good player can beat the 2 humans but not the 2 bots. Therefore, his EV drops to about zero.

Guess what he does? He leaves, also.

Suppose the remaining two humans are dumb enough to stick around FTP (they are average to fishy, afterall). It's two bots fighting for 2 humans (who start short-stacking), and often times clash with each other (for 0 EV).

Now, because of the high licensing fee of the SBbot, one of the botters decides to quit because FTP has too many bots per human.

(Note: Even the botters will hate the botters!)

So what remains at the FTP table? Three players: 1 bot and 2 humans. The botter makes a nominal profit (subtracting the licensing fee), and the 2 humans loose and never move up. (Note that even a truly talented player might get discouraged and never break through.)

The population at FTP just got reduced by 50%, and that's counting the bots!

Do you think FTP is gonna allow that to happen? HELL no. Not if they can't stop it....


7. However, if none of the poker sites can stop the bots, then there is one other thing a human player can do: Stop playing online. .

8. Are there really 2:1 bots online? (You're not counting PT as a "bot", are you

[CallMeIshmael]

[ QUOTE ]
1. No one's talking about pure strategies. If you know someone who relies on a pure strategy, please give me his screen name!

[/ QUOTE ]

It is VERY VERY easy to use a pure strategy and still be a great player. Especially at non-HU.

[RIIT]

[ QUOTE ]
[ QUOTE ]
1. No one's talking about pure strategies. If you know someone who relies on a pure strategy, please give me his screen name!

[/ QUOTE ]

It is VERY VERY easy to use a pure strategy and still be a great player. Especially at non-HU.

[/ QUOTE ]

I agree with this statement - human or bot doesn't matter. However, over time, if others profile your play, it then becomes possible for them to find an exploit.

RIIT

[dismalstudent99]

[ QUOTE ]


It is VERY VERY easy to use a pure strategy and still be a great player. Especially at non-HU.

[/ QUOTE ]


I see your point (I think). It would take a long time for opponents to catch on to your strategy, and even if they did, there is still a lot of uncertainty over the meaning of your actions.

[Hielko]

You can create a pure strategy that works as an mixed stategy without a lot of problems.

For example:

limp J [img]/images/graemlins/heart.gif[/img] T [img]/images/graemlins/heart.gif[/img] preflop, but raise J [img]/images/graemlins/club.gif[/img] T [img]/images/graemlins/club.gif[/img]

And you have even more options, f.e.:

Raise 7 [img]/images/graemlins/heart.gif[/img] 5 [img]/images/graemlins/spade.gif[/img] on a 8 [img]/images/graemlins/heart.gif[/img] 7 [img]/images/graemlins/spade.gif[/img] 5 [img]/images/graemlins/club.gif[/img] , but call when you have the same hand but the flop suits are different.

It's a pure stategy since you always do the same thing when you are in the same situation, but since the number of possible situations is HUGE you can do a lot of things in your strategy.

[goofyballer]

[ QUOTE ]
My point is this: programs aren't magic, if you can say what you want the program to do, then you can program it. But the problem is, you can not say what you want it to do. When you start to get specific, you will find that you have problems.

[/ QUOTE ]

QFT. All you guys who think writing poker bots are easy, go ahead and try to write down how you play a hand of poker, with the restriction that you cannot include any steps which involve intuition. You can use any piece of data you like, but any single fact that was not given to you as a piece of specific information or was computed in a previous step is off-limits.

Once you do that, hey, there's your poker bot! Go forth and rob our monies!

[RIIT]

[ QUOTE ]
[ QUOTE ]
My point is this: programs aren't magic, if you can say what you want the program to do, then you can program it. But the problem is, you can not say what you want it to do. When you start to get specific, you will find that you have problems.

[/ QUOTE ]

QFT. All you guys who think writing poker bots are easy, go ahead and try to write down how you play a hand of poker, with the restriction that you cannot include any steps which involve intuition. You can use any piece of data you like, but any single fact that was not given to you as a piece of specific information or was computed in a previous step is off-limits.

Once you do that, hey, there's your poker bot! Go forth and rob our monies!

[/ QUOTE ]

Well said.

RIIT

[RIIT]

[ QUOTE ]
RITT,

1. No one's talking about pure strategies.

[/ QUOTE ]

Well I was but that wasn't immediately clear because I was using a non-standard phrase.

[ QUOTE ]
If you know someone who relies on a pure strategy, please give me his screen name!

[/ QUOTE ]

I've used a pure strategy many times (but no you cannot have my screen name [img]/images/graemlins/smile.gif[/img] ) ... they work fine as long as the number of hands needed to fingerprint the strat. is intractable.

[ QUOTE ]
There exists a mixed strategy NE, though we may not know what it is.... yet

[/ QUOTE ]

Yes.

[ QUOTE ]
BTW, computers can randomize better than humans, therefore they can implement mixed strategies more precisely (assuming they are programmed to do so).

[/ QUOTE ]

Yes.

[ QUOTE ]
2. Why do you assume botting technology has stopped developing (plateaud)?

[/ QUOTE ]

I didn't say plateau. I was just trying to bring balance to the idea that an orders of magnitude "deep blue" future is pending for all of poker.

[ QUOTE ]
Is there evidence that programmers have "hit a wall"?

[/ QUOTE ]

I guess that depends on how we'd define "wall". Within the context of a highly controlled lab environment I'd say no - this would be the pure game theory type of environmental conditions. But what I've been trying to say (and maybe not very well) is that when you take what you've learned in a sterile lab and field test it in live conditions, things change - and the reason is that the live field data is influenced by factors beyond sterile lab conditions. If all sites bought into the idea that computer assisted players (which would include autoplayers) are a natural and normal part of the internet environment then that would impact some of the inferior poker client designs out there that impede poker ai research in live conditions (i.e. some sites really suck badly at delivering a clean accurate game state that is readable by your computer (and in some cases it's unreadable to humans as well)). So I'd say that the current backward thinking irrational site policies are the single biggest "wall" in regards to advancing poker ai. I maintain that advances in poker ai cannot kill the online game for various observable reasons (with the possible exception of HU cash holdem for the game theory reasons already discussed).

[ QUOTE ]
With all the attention poker is getting now, some good minds (and some careers) are on the problem.

[/ QUOTE ]

Yes. But let the record reflect that the UofA (and others) do consistently decline to wager real money on their research whereas the 2007 PBWC saw players from 4 different countries do exactly that within real world real money conditions.

[ QUOTE ]
3. Yes, I am interested in what people enjoy doing -- I want online poker to remain highly populated with players, especially with fish.

[/ QUOTE ]

Yes I do too but there is something perverted and dishonest about artificially twisting and contorting the natural environment of the online game in order to maintain the perception that a site is populated with "mythical fish".

[ QUOTE ]
Therefore, if many players object to bots and threaten to leave online poker, then we should do everything to eliminate the bots.

[/ QUOTE ]

This is and will always be pure and blatant speculation until at least 1 major site adopts a policy that agrees with the natural environment of the internet and accepts and embraces all forms of computer assistance as perfectly normal and natural - the core idea here is that player freedom of choice is a good and healthy thing. I maintain that max player freedom is good for online poker.

Even if what you say is true (and I do not believe that to be the case because fundamentally there is zero difference between a HUD/calc player clicking buttons and an autoplayer clicking buttons insofar as the impact upon the poker economy (and there are many HUD players and I'm fairly sure you will see them all leave if HUD's were disallowed), in other words a HUD/calc player is a "bot" - it simply doesn't matter who/what clicks the buttons and presses the keystrokes).

[ QUOTE ]
What's "backwards" about that?

[/ QUOTE ]

It's backward because it disagrees with the natural normal environment of the internet where computers outnumber humans 2 to 1 (or more) and players cannot physically see each other.

[ QUOTE ]
And what's wrong with "backwards" anyways?

[/ QUOTE ]

Backward thinking always leads to undersirable outcomes.

[ QUOTE ]
4. Since I pay attention to what interests people, I can confidently say that most people would not enjoy tinkering with a "baby bot" as much as you do.

[/ QUOTE ]

Your statement assumes that online poker should reflect the aspects of a democracy while ignoring the natural conditions of the environment regardless. And once again you're conveniently failing to include HUD/calc players as identical to bots insofar as they impact the poker economy. If you disagree then you bear the burden to demonstrate how/why a human clicking buttons according to a HUD/calc has a neutral impact on the poker economy but if a computer clicks the buttons then the economy is somehow adversely effected. What I want trumps what the population at large wants as long as what I do agrees with the natural internet environment and as long as everyone else has equal access to my tools and my information.

[ QUOTE ]
Would I want to play HALO 3 via programming ("Use Grenade 33%, Rockets 66%, Jump when evading 100%")? NO. I'll wear out my thumbs instead.

[/ QUOTE ]

This about your personal choice and freedom (as well as your opponents' personal choice); In the case where it was possible to use a HALO bot then your fate is to accept that as a natural element of the environment you've chosen. If you then lobby and whine against HALO botters you are the very near equivalent of a child. If you decide to play a game within an environment inundated with computers then it is irrational to expect to not encounter computer players. It is what it is.

[ QUOTE ]
However, I understand that botting has an intellectual appeal... much like Marcel Proust has an intellectual appeal among the literati.

[/ QUOTE ]

Yes it can be very intoxicating at times and frustrating as well.

[ QUOTE ]
But my point is that these are small groups of people who appreciate these things. When it comes to online poker, size does matter.

[/ QUOTE ]

Again your statement is spoken from a "majority rule" mindset. I can challenge this in 2 ways:
a) the natural internet environment trumps the forces of "majority rule" (and thankfully so)
b) your references to the size of the special interest group is misleading in the context of the impact on the poker economy.
You are singling out a group of people who tweak their own poker playing systems within commercial software tools or tools they've privately written themselves and then also use an autoplayer to transfer those decisions to the poker client. These aspects certainly single out a reduced population (note that this demographic is still in sync with the natural environment of the internet), but for the purposes of your ultimate point (which is the impact upon the poker economy), this group of auto players is identical in every way to an illiterate who buys a HUD/calc and then clicks the decisions himself. So the bottom line here (which is the core essence of this thread) is that it is the quality of the poker decisions available to online players that has the greatest overall impact on the online poker economy (how those decisions get communicated to the poker client is entirely irrelevant).

And this then leads to the detestable idea that we must then mangle the natural internet game environment in order to hinder the general population from accessing high quality poker decisions - because too few "mythical fish" making low quality poker decisions will kill online poker.

And this then leads to the "all winners == all bad" threads.

[ QUOTE ]
5. Talented bots would have a different effect on the poker economy than talented people. This is because the bots can be replicated and become wide-spread, whereas real poker talent seems to be a rare thing. It is this feature makes bots a threat to the game.

[/ QUOTE ]

Once again you're rebelling against the natural internet environment (where information spreads like light). Your statements assume that online poker can be killed if the light shines too bright or too fast. You're wrong in 2 ways:
a) for believing that can happen in the first place
b) for trying to save holdem if it can actually be killed

WEAK GAMES WILL DIE
(i.e. get solved)

Holdem is not a weak game. Shame on you for judging it so.
(note: I accept that there may be solvable fringes as in HU cash holdem)

[ QUOTE ]
One SBRugby playing at 200/400 doesn't affect me. But 2 SBRugby-bots at my 0.10/0.25 would probably cripple me (and stop me from donating money to 0.25/0.50 players) Do you see the difference?

[/ QUOTE ]

Are you saying that a 2-chair team has edge against 1-chair team in a 3-chair holdem game with blinds?

[ QUOTE ]
6. Competitive pressures will force major sites to take an anti-bot stance, especially if the bots are good.

Here's an extreme case.

Suppose FTP allows bots but PokerStars bans them.

Suppose I'm a breakeven, 100NL player on FTP. However, today I find out that I have to 6-max against:

2 "SBrugby bots"
1 good player
1 average player, and
1 fish.

My EV just went from 0 to negative.

Obviously, I'm switching to Stars, even if it means putting up with a bunch of irritating anti-botting measures.

It gets worse-- the table at FTP now has only 5 players. The good player can beat the 2 humans but not the 2 bots. Therefore, his EV drops to about zero.

Guess what he does? He leaves, also.

Suppose the remaining two humans are dumb enough to stick around FTP (they are average to fishy, afterall). It's two bots fighting for 2 humans (who start short-stacking), and often times clash with each other (for 0 EV).

Now, because of the high licensing fee of the SBbot, one of the botters decides to quit because FTP has too many bots per human.

(Note: Even the botters will hate the botters!)

So what remains at the FTP table? Three players: 1 bot and 2 humans. The botter makes a nominal profit (subtracting the licensing fee), and the 2 humans loose and never move up. (Note that even a truly talented player might get discouraged and never break through.)

The population at FTP just got reduced by 50%, and that's counting the bots!

Do you think FTP is gonna allow that to happen? HELL no. Not if they can't stop it....

[/ QUOTE ]

This is the same line of thought ad nauseum which is:

We must do everything in our power to hinder access to high quality poker decisions available to the "mythical fish" in order to keep holdem alive.


[ QUOTE ]
7. However, if none of the poker sites can stop the bots, then there is one other thing a human player can do: Stop playing online.

[/ QUOTE ]

Or begin to explore the world of multi-chair holdem which increases the complexity of the game beyond solvability. If single chair holdem can be killed/solved (and I don't believe this) then it will either pass into history or evolve additional complexity worthy of continued existence.


[ QUOTE ]
8. Are there really 2:1 bots online? (You're not counting PT as a "bot", are you

[/ QUOTE ]

I never said this. I assert that the number of available computers (within the domain of online poker) outnumbers the human players 2 to 1. And if you're willing to include virtural setups (like vmware) then the number goes much higher. Consider for a moment, the ratio of computers to humans within the live game - 0:1, now observe that every internet player must have at least 1 computer to engage the online game which means that, at a minimum, when you transition from the live game to the online game the number of available computers within the playing population increases from 0 to N where N is the number of players in the poker economy. And yet we still have people out there who think it's unnatural to encounter computer assisted players in online poker.

RIIT

[dismalstudent99]

In today's WSJ there's an interesting article ("Hannah Montana Battles the Bots," 10/05/07) about how Ticketmaster is suing RMG Technology, which makes ticket-buying bots.

I think there are similarities to what we might see online poker sites doing to bot makers, if the problem became bigger.

http://online.wsj.com/article/SB1191...ub_marketplace

Major points:
- Ticketmaster seeks injunction to prohibit selling the software.
- Ticketmaster is also suing a group of brokers who use RMG software, alleging violation of terms of agreement.
- RMG responds, "This may be the only time in the history of litigation that any seller sued its customers for paying them too much money."
- RMG questions the definition of "bot"
- RMG's "bot" can decode the sqiggly letters and numbers ("captcha") used for verification, and it can fake multiple IP addresses for its numberous ticket requests.