Image Image Image




Post new topic Reply to topic  [ 7 posts ] 
Author Message
 Post subject: ICM-play for postflop?
PostPosted: Wed May 20, 2009 3:31 pm 
Offline
Senior member
User avatar

Posts: 232
Favourite Bot: Poki
Is there anything like this for postflop play? I am thinking about the situations where I am in BB and it has been limped to me so I am forced to see the flop


Top
 Profile E-mail  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Fri May 22, 2009 10:05 am 
Offline
PokerAI fellow
User avatar

Posts: 7731
Favourite Bot: V12
I don't think so ... at least I don't know of one in the public space.

_________________
indiana


Top
 Profile E-mail  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Fri May 22, 2009 11:01 am 
Offline
PokerAI fellow
User avatar

Posts: 1673
Favourite Bot: none
This should be possible to calculate aswell, but I am not aware of anyone who has done that yet.
You would need to bucket the flop into different categories that call for the same strategic action. This alone would be hard enough to do but then you could use the same algorithm which this time not only takes the holecards into account but also the flop. I'd just go with a rule or equity based approach, e.g. Push Tptk fold all other or calculate you equity against arandom hand or some limping range for SB and push if it is favorable for you. You can use ICM to calculate the $EV at each outcome your move to decide that.

If you are working on a SNG bot then please consider applying to the SNG-Testbed subforum via PMing me.

_________________
Cheers.


Top
 Profile  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Fri Jun 26, 2009 12:04 pm 
Offline
Senior member
User avatar

Posts: 138
Favourite Bot: Custom
Blinds 50/100


9 Players start with T2000 every one starts with 11.11 ev

sb limps we check and see the flop sb checks to us.



now if we win the pot our ev will be 11.6 and if we lose it will be 10.6

but to win the pot we will need to bet lets just say we bet half the pot (T100)

if we win after betting our ev is still 11.6 but if we lose after betting our ev will be 10.1


lets say our hand has 20% equity if checked down
80% of the time we win 10.6 = 8.48 20% of the time we win 11.6 = 2.32 total = 10.8

so if we check the hand down we have 10.8 ev

Lets say he will call with any pair and he will have a pair around 45% of the time.

65 % of the time we will win 11.6ev = 7.54
45% of the time he will call but we will have lets say 5% ev (the 5% ev does not really make that much differance)

If we win the pot after he calls we will have 12.09 ev
If we lose we have 10.1

so (10.1 * 0.95 ) 9.59 + (12.09*.05) = 0.60 = (9.59 + .60) = 10.19


but all of this will only happen 45% of the time so (10.19*0.45) = 4.58

so if we bet we have (7.54 ev + 4.58) = 12.1

if we check our ev is 10.8
if we bet our ev is 12.1

Betting is the better option

I think these numbers are correct. This is just an example of how it can be done but the hard part is
getting his call percentage and hand range at each stage this is where you will need a good opponent model.


You should be able to use icm for any desision pre or post flop.


Hope this helps


Top
 Profile  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Fri Jun 26, 2009 4:35 pm 
Offline
Junior member
User avatar

Posts: 32
Favourite Bot: None
In a SNG you shouldn't need to use ICM till push or fold phase, if you reach and really need a push or fold phase. ICM doesn't take the opponent or yours skill in count, so it's not accurate, against skilled players a simply ruled based with these maths will often loose, because the skilled player may induce you to a error... This is what i know... In low limits is VERY nice to use it, you simply ignore everything else, just do "quick" maths, train a bit to know what to do when to do ( rule basing yourself ) and play it... It's +EV on low limits, but i don't know how it will react at higher limits.


Top
 Profile E-mail  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Fri Jun 26, 2009 6:46 pm 
Offline
Senior member
User avatar

Posts: 138
Favourite Bot: Custom
Yes a rule based bot with ICM at the push and fold stage would be +ev and is .

I am wondering why you are saying ICM only applies at the push and fold stage. The reason people play tight in the first few levels is because ICM calculations say confrontation is bad for you early in the SNG.

Eg first hand player goes all in and shows you AK you look at your cards and see 22 ICM says you should fold because when you dbl your stack you dont dbl you $equity

I am also wondering what things an opponent could do to induce errors.

The icm calculations are only one part of it you need a good opponent modeling system for this to work. The question was can this be done I showed the calculations on how you can do it and with a very good modeling system this would probably beat most rule based system.


Top
 Profile  
 
 Post subject: Re: ICM-play for postflop?
PostPosted: Sat Jun 27, 2009 11:40 am 
Offline
PokerAI fellow
User avatar

Posts: 1673
Favourite Bot: none
luizsilva wrote:
In a SNG you shouldn't need to use ICM till push or fold phase, if you reach and really need a push or fold phase. ICM doesn't take the opponent or yours skill in count, so it's not accurate, against skilled players a simply ruled based with these maths will often loose, because the skilled player may induce you to a error... This is what i know... In low limits is VERY nice to use it, you simply ignore everything else, just do "quick" maths, train a bit to know what to do when to do ( rule basing yourself ) and play it... It's +EV on low limits, but i don't know how it will react at higher limits.


The error doesn't come from the ICM calculation itself because ICM is only a way to map your chips to equity in $. Another way would be to take your stack and multiply it by (total prizepool divided by total amount of chips).

The error you are talking about is coming from the assumed ranges. ICM alone doesnt suggest what hands to push and what to fold but is only a method to compensate the fact that cEV != $EV. That is very important to have understood.

_________________
Cheers.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 7 posts ] 


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to: