Two Plus Two Publishing LLC Two Plus Two Publishing LLC
 

Go Back   Two Plus Two Poker Forums > >

Notices

Poker Theory General poker theory

Reply
 
Thread Tools Display Modes
Old 09-03-2017, 06:20 PM   #26
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

Quote:
Originally Posted by whosnext View Post
You have said this on multiple occasions.

It is not true.

It reflects a misunderstanding of what Nash Equilibrium entails, especially in a poker setting.


There are certainly examples of NE in repeated multiplayer games, such as a repeated prisoners dilemma. I believe multi-player poker's equilibrium to be similar if you were to exhaust it. You either get nothing in the long run, or less than nothing for deviating from dominant strategy.

Last edited by leavesofliberty; 09-03-2017 at 06:33 PM.
leavesofliberty is offline   Reply With Quote
Old 09-03-2017, 06:23 PM   #27
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

Quote:
Originally Posted by heehaww View Post
If you let a chess engine think long enough, it will solve chess.
It's safe to say chess is closer to being solved than poker.
leavesofliberty is offline   Reply With Quote
Old 09-04-2017, 06:32 AM   #28
Yadoula8
adept
 
Join Date: Aug 2014
Location: Royal Leamington Spa
Posts: 986
Re: Master Game Theory on University

Quote:
Originally Posted by mrno1324 View Post
finding a strategy that beats any other strategy is...
In a tournament with a mixed field of players, even a GTO bot would get destroyed by a good exploitative player. So not even this is true.
Yadoula8 is offline   Reply With Quote
Old 09-04-2017, 06:40 AM   #29
Yadoula8
adept
 
Join Date: Aug 2014
Location: Royal Leamington Spa
Posts: 986
Re: Master Game Theory on University

Quote:
Originally Posted by RustyBrooks View Post
In the context of game theory, which is what this post is about, "solving" a game has a very specific meaning. Some games have solutions that satisfy everyone. Many games, like poker, do not. As many people have noted over the years the word "optimal" in GTO is an unfortunate choice because it does not meet the intuitive defintion of the word optimal. When I studied game theory, those kind of solutions were usually called "sub-optimal" and were found in games where optimal solutions weren't possible, such as the classic single-iteration of "The Prisoner's Dilemma"

Finding a perfect exploitative strategy to a specific fixed strategy is actually very simple and does not take nearly as much computational resources as finding GTO strategies. It's also not particularly interesting from a game theory perspective.
Alrite, so, Game Theory Optimal has been solved. Thanks Rusty. I love when you back me up... But you did make a teeny little mistake there... Finding a perfect exploitative strategy to beat a specific set strategy is pretty much impossible in a real life situation, because, you will never know what that set strategy is. Unless you can read minds. Perhaps that's what you meant. "Finding a perfect exploitative strategy to a specific fixed strategy is actually very simple when you have telepathic powers"
Yadoula8 is offline   Reply With Quote
Old 09-04-2017, 10:00 AM   #30
RustyBrooks
Carpal \'Tunnel
 
RustyBrooks's Avatar
 
Join Date: Feb 2006
Location: Austin, TX
Posts: 23,071
Re: Master Game Theory on University

Quote:
Originally Posted by Yadoula8 View Post
Alrite, so, Game Theory Optimal has been solved.
At a theoretical level, but not really a practical level (some games are more solved than others)

Quote:
"Finding a perfect exploitative strategy to a specific fixed strategy is actually very simple when you have telepathic powers"
Yes, you have to know what the fixed strategy is to find the counter strategy

Quote:
In a tournament with a mixed field of players, even a GTO bot would get destroyed by a good exploitative player. So not even this is true.
It would depend on the terms of the tournament somewhat, but in most human-scale tournaments, a GTO bot would have much better than "fair" odds to win (i.e. it would win more than 1/N with N players) but winning would not be assured.
RustyBrooks is offline   Reply With Quote
Old 09-04-2017, 02:44 PM   #31
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

Back to the OP. An alternative is to analyze a sport, preferably one that the professor enjoys. That is, if you want to kiss ass. You'll enjoy the project, though not as much as poker, and the professor will enjoy it more than if it were about poker.

Baseball.
leavesofliberty is offline   Reply With Quote
Old 09-05-2017, 09:00 AM   #32
just_grindin
Pooh-Bah
 
Join Date: Dec 2007
Posts: 4,261
Re: Master Game Theory on University

Quote:
Originally Posted by leavesofliberty View Post


There are certainly examples of NE in repeated multiplayer games, such as a repeated prisoners dilemma. I believe multi-player poker's equilibrium to be similar if you were to exhaust it. You either get nothing in the long run, or less than nothing for deviating from dominant strategy.
By definition a Nash Equilibrium is where neither player can deviate from his or her strategy to gain EV for their own strategy. With collusion you have players cooperating to change another player's strategy so it violates a Nash Equilibrium by definition.

It's certainly interesting and relevant to talk about cooperation and collusion in game theory, but I think what whosnext was trying to say is that it doesn't make sense to talk about those things in the context of Nash Equillibria.

Sent from my SM-G900R4 using Tapatalk
just_grindin is offline   Reply With Quote
Old 09-05-2017, 11:29 AM   #33
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

Quote:
Originally Posted by just_grindin View Post
By definition a Nash Equilibrium is where neither player can deviate from his or her strategy to gain EV for their own strategy. With collusion you have players cooperating to change another player's strategy so it violates a Nash Equilibrium by definition.

It's certainly interesting and relevant to talk about cooperation and collusion in game theory, but I think what whosnext was trying to say is that it doesn't make sense to talk about those things in the context of Nash Equillibria.

Sent from my SM-G900R4 using Tapatalk
Yes, yes, I understand all that, but what I am saying in the context of a repeated game it makes perfect sense, which you'd follow if you listened to the lecture. Cartels are one of the examples used in the videos regarding collusion and NE.

You can also read the section of N-Player games in Further Limit Hold'em by Newhall. There is also the classic example of checking down to eliminate someone in a tournament.
leavesofliberty is offline   Reply With Quote
Old 09-05-2017, 11:30 AM   #34
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

TBH, whosnext didn't actually make any points whatsoever, and I have no idea what his education background is, either.
leavesofliberty is offline   Reply With Quote
Old 09-05-2017, 12:10 PM   #35
Bob148
Carpal \'Tunnel
 
Join Date: May 2012
Location: far too young and clever
Posts: 8,624
Re: Master Game Theory on University

Tourneys and drug dealing are positive sum games.
Bob148 is online now   Reply With Quote
Old 09-05-2017, 01:38 PM   #36
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

The tourney itself is zero-sum, but the subgame with the hand to eliminate someone is not. Two can benefit to the expense of a third as described also by Newhall. I have studying to do though, so unless there's a point coming addressing what I've mentioned, and cited, I'm done.
leavesofliberty is offline   Reply With Quote
Old 09-05-2017, 04:13 PM   #37
Bob148
Carpal \'Tunnel
 
Join Date: May 2012
Location: far too young and clever
Posts: 8,624
Re: Master Game Theory on University

The situation you're describing for tourneys happens when you're in the money or close to it; this is not zero sum in terms of tourney ev.
Bob148 is online now   Reply With Quote
Old 09-05-2017, 04:16 PM   #38
Yadoula8
adept
 
Join Date: Aug 2014
Location: Royal Leamington Spa
Posts: 986
Re: Master Game Theory on University

Quote:
Originally Posted by RustyBrooks View Post
It would depend on the terms of the tournament somewhat, but in most human-scale tournaments, a GTO bot would have much better than "fair" odds to win (i.e. it would win more than 1/N with N players) but winning would not be assured.
Hmmm, this has got me thinking...

I wonder how good/bad GTO is in a tournament.

Your 1/N is just a pure guess. It would depend on the tournament structure a lot, as you say. It would also depend on how exploitable the player pool was. And funnily enough, the worse the general pool of players, the less chance the GTO player stands of beating a good exploitative player.

Lets imagine a GTO bot were playing against loads of super fish and a few good exploitative players, or we could go even further, and say that were deep too, and that only 1 player takes the prize. In this not massively unlikely situation the GTO bot would blatantly get destroyed! It wouldn't be unbeatable at all. The exploitative players will build stacks much faster than the GTO player. They will gain strength via their coin which will enable them to beat GTO way more often than GTO beat them.

Hmmm... This also means that GTO players will all make more money as more and more people use GTO. If there were less GTO players in the tournament the GTO players would stand less of a chance.

... Those of you who have studied game theory often talk about co-operation in games. As an honest poker player I thought this never applied to me. But I guess I was wrong. Is there another word for what these guys are doing?? The exploitative player is working with the fish without the fishes knowledge so as to beat GTO. And none of the GTO players know what they are doing, but still, they are all working together to take that advantage away from the exploitative players.

Bloody GTO, just keeps getting worse and worse.
Yadoula8 is offline   Reply With Quote
Old 09-05-2017, 08:51 PM   #39
just_grindin
Pooh-Bah
 
Join Date: Dec 2007
Posts: 4,261
Re: Master Game Theory on University

Quote:
Originally Posted by leavesofliberty View Post
Yes, yes, I understand all that, but what I am saying in the context of a repeated game it makes perfect sense, which you'd follow if you listened to the lecture. Cartels are one of the examples used in the videos regarding collusion and NE.

You can also read the section of N-Player games in Further Limit Hold'em by Newhall. There is also the classic example of checking down to eliminate someone in a tournament.
I watched the video and to me there is no mention of cooperation or collusion, but then again I have never seen cooperation or collusion rigorously defined in game theory.

Collusion seems to be implied in minmax strategies but that's not necessarily a part of a Nash Equilibrium.

I am not sure how cooperation would be defined in game theory. Something like an agreement between 2 players so that each improves their EV? I don't know if that even works or is relevant. Any way I would love to hear your thoughts on my random musings here.

Sent from my SM-G900R4 using Tapatalk
just_grindin is offline   Reply With Quote
Old 09-06-2017, 06:34 AM   #40
MuppetExpress
stranger
 
Join Date: Aug 2017
Posts: 5
Re: Master Game Theory on University

Quote:
Originally Posted by just_grindin View Post
I watched the video and to me there is no mention of cooperation or collusion, but then again I have never seen cooperation or collusion rigorously defined in game theory.
Collusion (or cooperation) is a set of players acting to maximise their combined pay-off. Secondly, in a Bayesian game, players may share their private information with some other players which can be termed cheating or collusion. Usually these definitions work well but I would always clearly state in what sense I intended to use the words at the beginning of an argument.
MuppetExpress is offline   Reply With Quote
Old 09-06-2017, 12:25 PM   #41
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

I'll try to flesh-out and outline what I'm trying to say as soon as I can (like Monday, bad timing, weekend planned, etc.). I am a mere student fwiw. I am glad to participate in the stimulating discussion we're having.
leavesofliberty is offline   Reply With Quote
Old 09-06-2017, 03:32 PM   #42
Bob148
Carpal \'Tunnel
 
Join Date: May 2012
Location: far too young and clever
Posts: 8,624
Re: Master Game Theory on University

I think some of us get what you're talking about.

You think that the collusion and recollusion and exploiting and coexploiting and reexploiting repeat ad nauseum until no other collusion can occur? Ok, that's fine. I think it ends exactly the way it begins; with no acts of coexploitation nor collusion.

I have nothing but concept in my mind to back this up. So here's some salt:

****
Bob148 is online now   Reply With Quote
Old 09-06-2017, 04:54 PM   #43
leavesofliberty
Pooh-Bah
 
Join Date: Jul 2010
Posts: 5,736
Re: Master Game Theory on University

Quote:
Originally Posted by MuppetExpress View Post
Collusion (or cooperation) is a set of players acting to maximise their combined pay-off. Secondly, in a Bayesian game, players may share their private information with some other players which can be termed cheating or collusion. Usually these definitions work well but I would always clearly state in what sense I intended to use the words at the beginning of an argument.
Yes, collusion I mean what would commonly be called cooperation. Newall wrote about it in his book FLH with sections on N-player games. When the pot gets a size, and there are other players in the pot, then people try to "knock each other out", and in doing so are co-operating. But, it's not sharing information, or crossbooking, etc. I have nothing original to add. So, rather than re-write, I'd encourage people to read it.

Also, poker is solved as the joke goes. It's 0 EV not to play. You'll never have to post a BB again!
leavesofliberty is offline   Reply With Quote
Old 09-20-2017, 11:07 AM   #44
Bob148
Carpal \'Tunnel
 
Join Date: May 2012
Location: far too young and clever
Posts: 8,624
Re: Master Game Theory on University

Quote:
Originally Posted by leavesofliberty View Post
I'll try to flesh-out and outline what I'm trying to say as soon as I can (like Monday, bad timing, weekend planned, etc.). I am a mere student fwiw. I am glad to participate in the stimulating discussion we're having.
Bumping in case you forgot about this.

I've read the section in Further Limit Holdem referenced upthread, but I think what you're missing is the possibility of counter exploitation by a blind, which will dissolve any alliances before they ever come to fruition.

Also, Nash even titled his work as "non cooperative games" or something like that, so it seems that the idea that cooperation is a part of nash equilibrium is inherently flawed.
Bob148 is online now   Reply With Quote
Old 09-20-2017, 11:31 AM   #45
robert_utk
adept
 
robert_utk's Avatar
 
Join Date: Jan 2005
Posts: 954
Master Game Theory on University

Collusion is a proven method for two players to profit versus a third player who is optimal. If the prisoners can communicate then there is no prisoners dilemma. This is not cheating, but the two players would play opposite solutions and crossbook the results.

FWIW this is why most criminals just lawyer up and don't fall for the hard sell from the interrogators in real life. The criminals have predetermined to take the "no fink" line.

Last edited by robert_utk; 09-20-2017 at 11:38 AM.
robert_utk is offline   Reply With Quote
Old 09-20-2017, 11:35 AM   #46
Bob148
Carpal \'Tunnel
 
Join Date: May 2012
Location: far too young and clever
Posts: 8,624
Re: Master Game Theory on University

Right, but the point is that Nash proved that there is a minimum ev in non cooperative games, not that there is a minimum ev in cooperative games, which there is not.
Bob148 is online now   Reply With Quote
Old 09-21-2017, 04:36 AM   #47
Fatboy54
veteran
 
Join Date: Jul 2011
Posts: 3,306
Re: Master Game Theory on University

Since humour does not always do well hereabouts, I'll ruin my input before I make it, but what follows is both semi-ironic and semi-serious...

Quote:
As many people have noted over the years the word "optimal" in GTO is an unfortunate choice because it does not meet the intuitive defintion of the word optimal.
And ditto "theory".

In fact the only accurately appropriate word in the expression "GTO" is "game".
Fatboy54 is offline   Reply With Quote
Old 09-21-2017, 12:31 PM   #48
a dewd
adept
 
a dewd's Avatar
 
Join Date: May 2016
Location: west of the easternmost pt up north
Posts: 729
Re: Master Game Theory on University

Quote:
Originally Posted by RubenHarm View Post
....
And do you think it is possible to keep my thesis poker related? What are some serious academic researches I could do considering poker and game theory?
There have been critical essays written on an inordinate amount of applications, including animals and insects playing a game. If you are not already a member, you may want to consider joining

https://www.academia.edu/

You will probably not live long enough to digest all the papers submitted on game theory.

Unfortunately...

Quote:
Originally Posted by MuppetExpress View Post
.....
</Engage disappointment>
Your thesis will be related to the areas your advisor understands, otherwise they can not advise you. If you do not follow an advisor's topics then you will most likely fail because your work is wrong, already done etc etc. I know that is disappointing and rather dull but it is sadly true.
that is very accurate. If your advisor happens to be a poker buff, bon appetit!
a dewd is offline   Reply With Quote

Reply
      

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Forum Jump


All times are GMT -4. The time now is 10:12 AM.


Powered by vBulletin®
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © 2008-2010, Two Plus Two Interactive
 
 
Poker Players - Streaming Live Online