Computers Conquer Texas Hold'em Poker for First Time - Page 13 - Poker News

Something that might not have been really clear is that both games are still using $1 chips, which could certainly be a bit odd in real games. IF you use $1 chips, there are more betting options in $25/$50 than in $1/$2. You can raise to $25, $26, $27, ..., $25000 which is a lot more options that $1, $2, ..., $1000

This might be really naive but why can't we generalize to 1bb intervals?

Sure, maybe a gto optimal betsize in one spot could be exactly 27.5bb but it will be practically the same as 28bb, right?

Quote

01-13-2015 , 08:59 AM

#303

ambientsilence

journeyman

Join Date: Oct 2014 Posts: 391

Quote:

Originally Posted by maddog2030

People need to stop comparing the number of states of a game to the number of atoms in the universe.

One random line in an ASCII coded text file has more possible states than the number of atoms in the universe (256^40~10^95). Of course it has fewer possible states than the number of possible states of the universe, by far. In particular it has fewer possible states than the number of possible states of your computer's memory.

Quote

01-13-2015 , 09:49 AM

#304

Vempele

adept

Join Date: Jul 2010 Posts: 841

But we still need multiple atoms per bit of data storage, and the storage requirement is a function of the size of the state space* so it remains a valid comparison.

*Technically you could solve chess with kilobytes if you're only interested in evaluating the starting position, but a full solution would involve storing the best move in every position reachable with at least one side playing perfectly.

Quote

01-13-2015 , 10:50 AM

#305

maddog2030

adept

Join Date: Feb 2005 Posts: 889

Quote:

Originally Posted by ambientsilence

You're reading too much into it. It was just to give a little perspective on the other large numbers, not a comparison between states and atoms.

Quote

01-13-2015 , 11:04 AM

#306

DrawNone

Pooh-Bah

Join Date: May 2013 Posts: 5,023

Quote:

Originally Posted by CanadaPete

Im sure a programmer could design a real simple bot that could beat any live NL Holdem game (outside of the toughest lineups and beat low stakes online.). Playing a decent strategy along with not tilting or deviating from sound strategies is a big edge even if some of the human players were still better.

it's far different from solving the game but much more realistic in the near term.

there's a thread somewhere on 2p2 about bots on iPoker crushing 1/2 NL.

Quote

01-13-2015 , 11:28 AM

#307

Loctus

Carpal \'Tunnel

Join Date: Nov 2008 Posts: 11,220

There was a bot ring crushing HS NLHE exposed a couple months back IIRC.

Quote

01-13-2015 , 11:47 AM

#308

maddog2030

adept

Join Date: Feb 2005 Posts: 889

Quote:

Originally Posted by nburch

Exactly. At least with LHE it's easier because given a minimum stack size it's all the same. Whereas in NL every stack size is a different game. Allowing pennies is a different game. Rake is a different game. Etc.

And many other caveats. This is why I said for entertainment purposes only

Quote

01-13-2015 , 12:21 PM

#309

ambientsilence

journeyman

Join Date: Oct 2014 Posts: 391

Quote:

Originally Posted by Vempele

But we still need multiple atoms per bit of data storage, and the storage requirement is a function of the size of the state space* so it remains a valid comparison.

You can compress the solution. And as you admit yourself, often the vast majority of the state space is irrelevant to finding a solution.

Quote:

*Technically you could solve chess with kilobytes if you're only interested in evaluating the starting position, but a full solution would involve storing the best move in every position reachable with at least one side playing perfectly.

Sure, but that's still many orders of magnitude reduction from looking at every legal position. There are states that are never looked at. Almost all of them in fact.

edti: this seems to be less true for poker, but in poker you have symmetry arguments, effectively reducing the size of the game.

Quote

01-13-2015 , 12:53 PM

#310

Bill Hickok

journeyman

Join Date: Nov 2013 Posts: 347

Well there are more particles than atoms so yes it does

Sent from my LG-D855 using 2+2 Forums

Quote

01-13-2015 , 01:04 PM

#311

Shuuya19

stranger

Join Date: Nov 2013 Posts: 2

Is this about the new Hellmuth machine at the Bellagio?

Quote

01-13-2015 , 02:50 PM

#312

Rant

old hand

Join Date: Oct 2006 Posts: 1,436

Quote:

Originally Posted by ezdonkey

PS. This probably belongs to that "future of online poker" topic but w/e. Do you know if e.g. Apple tablets work so that iOS can prevent you from running 3rd party software if you're playing PokerStars app? I mean, if that's the case then I think a likely scenario is that in future online poker might be only available on specific devices. I guess it would add longevity for the game as well if usage of real-time huds etc. would be prevented.

Camera + robot arm = still undetectable bot

Quote

01-13-2015 , 02:50 PM

#313

Rant

old hand

Join Date: Oct 2006 Posts: 1,436

and don't suggest web cams. Not hard to conceal a camera and a computer that tells a human what to do.

Quote

01-13-2015 , 03:22 PM

#314

Donkem

journeyman

Join Date: Feb 2014 Posts: 359

Quote:

Originally Posted by maddog2030

...Rake is a different game.

Rake doesnt make the game different. It simply makes it less profitable

Quote

01-13-2015 , 03:57 PM

#315

JudgeHoldem1848

banned

Join Date: Jun 2012 Posts: 2,474

Quote:

Originally Posted by Donkem

Rake doesnt make the game different. It simply makes it less profitable

It changes starting hand distributions so the strategies at different rake levels should be markedly different for both HUNL and HU Limit.

Quote

01-13-2015 , 04:27 PM

#316

Rant

old hand

Join Date: Oct 2006 Posts: 1,436

Quote:

Originally Posted by Donkem

Rake doesnt make the game different. It simply makes it less profitable

Dude, just stop. You are either trolling or dumb. Stop.

Quote

01-13-2015 , 04:27 PM

#317

Donkem

journeyman

Join Date: Feb 2014 Posts: 359

Quote:

Originally Posted by JudgeHoldem1848

It changes starting hand distributions so the strategies at different rake levels should be markedly different for both HUNL and HU Limit.

What? lol. if your strategy is no longer the best with rake increase, then it wasnt the best to begin with. If the government decides to tax your winnings do u need to change it again?! lol

Last edited by Donkem; 01-13-2015 at 04:32 PM.

Quote

01-13-2015 , 04:40 PM

#318

JudgeHoldem1848

banned

Join Date: Jun 2012 Posts: 2,474

Quote:

Originally Posted by Donkem

What? lol. if your strategy is no longer the best with rake increase, then it wasnt the best to begin with. If the government decides to tax your winnings do u need to change it again?! lol

Yes.

Quote

01-13-2015 , 06:21 PM

#319

Bill Hickok

journeyman

Join Date: Nov 2013 Posts: 347

Quote:

Originally Posted by JudgeHoldem1848

Yes.

You should play tighter with rake and looser with hourly charge I suspect.

Sent from my LG-D855 using 2+2 Forums

Quote

01-13-2015 , 06:26 PM

#320

bjsmith22

veteran

Join Date: Sep 2011 Posts: 2,235

Quote:

Originally Posted by Donkem

What? lol. if your strategy is no longer the best with rake increase, then it wasnt the best to begin with. If the government decides to tax your winnings do u need to change it again?! lol

If the slow pony gets there too late, what do you call the pony who never gets there?

Quote

01-13-2015 , 06:38 PM

#321

Donkem

journeyman

Join Date: Feb 2014 Posts: 359

Lol just a thought for u guys, this bot cepheus is supposed to play considering no rake right? So basically by saying the FL HU is a different game with rake, you r saying that the bot didnt solve the game. LOL. Well at least we agree on something there.

Quote

01-13-2015 , 06:42 PM

#322

nnstrt

banned

Join Date: Apr 2014 Posts: 408

what is the proof that this bot "conquer" limit holdem 1 on 1?

Quote

01-13-2015 , 06:51 PM

#323

Donkem

journeyman

Join Date: Feb 2014 Posts: 359

The strategy remains almost the same, I do admit Im wrong though, there are specific spots where it will be better to call rather than raise on a cash game, for instance if u r betting for value and u suspect that your opponent has the same hand then it's pointless to raise any further because u r just getting raked. Still this doesnt exactly prove that u should change your strategy, u just get less profitable thats all. This doesnt apply to SNGs though, the best strategy is the best strategy, the money paid in rake or taxes is not directly involved in gameplay there.

Last edited by Donkem; 01-13-2015 at 07:19 PM.

Quote

01-13-2015 , 07:03 PM

#324

ArtyMcFly

Carpal \'Tunnel

Join Date: Dec 2014 Posts: 13,256

Quote:

Originally Posted by Donkem

It essentially solved rakeless HUFLHE at a particular stack size.

Precisely the same CFR method could be used to solve a FLH game with rake or different stack sizes. The resulting hand ranges would be different, but the solution method is identical.

FWIW, optimal pre-flop ranges at 2000NL (with low rake) are markedly different (looser) to pre-flop ranges for 2NL (with high rake).

Quote

01-13-2015 , 07:26 PM

#325

Donkem

journeyman

Join Date: Feb 2014 Posts: 359

Quote:

Originally Posted by ArtyMcFly

FWIW, optimal pre-flop ranges at 2000NL (with low rake) are markedly different (looser) to pre-flop ranges for 2NL (with high rake).

A pro is more willing to fold to a cbet than a fish is, so it's not surprising that at lower stakes u get to stack the donks with a tighter range. I really don't see the rake as having a major role on PF ranges, if any. Just my opinion. There's probably not that many donks at 2NL, but still, the field is def weaker.

Quote

Page 13 of 16

First

3 8 9 10 11 12 13 14 15 16

Last

Post Reply Subscribe

...

Page 13 of 16

First

3 8 9 10 11 12 13 14 15 16

Last